Estrogene effects in psychiatric disorders 9783211404850, 3-211-40485-6

Table of contents :
Content: Contents: Introduction and overview (N. Bergemann, A. Riecher-Roessler) * Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology (H. Vedder, C. Behl) * Estrogens and Schizophrenia (A. Riecher-Roessler) * Gender Differences in Schizophrenia (H. Hafner) * Puberty and Schizophrenia Onset (R. Z. Hayeems, M. V. Seeman) * Clinical Estrogen Trials in Patients with Schizophrenia (J. Kulkarni) * Hypoestrogenism and Estrogen Replacement Therapy in Women suffering from Schizophrenia (N. Bergemann, Ch. Mundt, P. Parzer, B. Runnebaum, F. Resch) * The Effect of Estrogens on Depression (L. S. Kahn, U. Halbreich) * Estrogens and other Hormones in the Treatment of Premenstrual Syndroms (L. Born, M. Steiner) * Estrogens and Perinatal Disorders (A. Gregoire) * Estrogen Therapy in Perimenopausal Affective Disorders (G. Stoppe, M. Doeren) * The Effects of Estrogens on Cognition and Alzheimer's Dementia (T. Edwin, U. Halbreich) * Effects of Hormone Therapy on Patterns of Brain Activation during Cognitive Activity: a Review of Neuroimaging Studies (P. M. Maki, S. M. Resnick) * Estrogens in Alzheimer's Disease: a Clinical and Neurobiological Perspective (P. Schoenknecht, J. Pantel, A. Hunt, M. Henze, T. Strowitzki, J. Schroeder) * Estrogen Therapy: Interface between Gynecology and Psychiatry (K. M.K. Ismail, G. V. Sunanda, P. M. Shaughn O'Brien)

Citation preview

W

Niels Bergemann Anita Riecher-Rössler (eds.) Estrogen Effects in Psychiatric Disorders

SpringerWienNewYork

Dr. Dr. Niels Bergemann Department of Psychiatry, University of Heidelberg, Germany

Prof. Dr. Anita Riecher-Rössler Psychiatric Outpatient Department, University of Basel, Switzerland

This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically those of translation, reprinting, re-use of illustrations, broadcasting, reproduction by photocopying machines or similar means, and storage in data banks. Product Liability: The publisher can give no guarantee for all the information contained in this book. The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.

© 2005 Springer-Verlag/Wien Printed in Austria

Springer WienNewYork is a part of Springer + Science Business Media springeronline.com

Typesetting: H. Meszarics • Satz & Layout • 1200 Wien, Austria Printing: Druckerei Theiss GmbH, 9431 St. Stefan, Austria Printed on acid-free and chlorine-free bleached paper SPIN: 10942017

With 40 Figures Library of Congress Control Number: 2004111514

ISBN 3-211-40485-6 SpringerWienNewYork

Contents Introduction and Overview (N. Bergemann, A. Riecher-Rössler) ...................... VII 1 Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology (H. Vedder, C. Behl) ............................................................

1

2 Estrogens and Schizophrenia (A. Riecher-Rössler) ........................................

31

3 Gender Differences in Schizophrenia (H. Häfner) .........................................

53

4 Puberty and Schizophrenia Onset (R. Z. Hayeems, M. V. Seeman) .............

95

5 Clinical Estrogen Trials in Patients with Schizophrenia (J. Kulkarni) ......... 107 6 Hypoestrogenism and Estrogen Replacement Therapy in Women Suffering from Schizophrenia (N. Bergemann, Ch. Mundt, P. Parzer, B. Runnebaum, F. Resch) ................................................................................ 123 7 The Effect of Estrogens on Depression (L. S. Kahn, U. Halbreich) .............. 145 8 Estrogens and Other Hormones in the Treatment of Premenstrual Syndromes (L. Born, M. Steiner) ..................................................................... 175 9 Estrogens and Perinatal Disorders (A. Gregoire) ............................................ 191 10 Estrogen Therapy in Perimenopausal Affective Disorders (G. Stoppe, M. Dören) ...................................................................................... 207 11 The Effects of Estrogens on Cognition and Alzheimer’s Dementia (T. Edwin, U. Halbreich) ................................................................................ 223 12 Effects of Hormone Therapy on Patterns of Brain Activation during Cognitive Activity: A Review of Neuroimaging Studies (P. M. Maki, S. M. Resnick) .................................................................................................. 239 13 Estrogens in Alzheimer’s Disease: A Clinical and Neurobiological Perspective (P. Schönknecht, J. Pantel, A. Hunt, M. Henze, T. Strowitzki, J. Schröder) ............................................................................... 253 14 Estrogen Therapy: Interface Between Gynecology and Psychiatry (K. M. K. Ismail, G. V. Sunanda, P. M. S. O’Brien) ....................................... 271 List of Contributors ............................................................................................... 289

V

Introduction and Overview N. Bergemann and A. Riecher-Rössler

The study of the effects of gonadal hormones in the brain was focused mainly on reproductive actions for a long time. Meanwhile, however, it is well known that gonadal hormones, in particular estrogens, also have neuroprotective and psychoprotective properties. They can obviously modulate many actions in the brain such as cognitive functions, pain regulation, motor coordination, epilepsy, as well as affective and psychotic disorders, to name just a few. In fact, during the past few years we have experienced a major change in our understanding of the endocrinologic aspects of psychiatric disorders. Endocrinologic irregularities in psychiatric patients are no longer viewed as pure epiphenomena but rather discussed as part of the pathomechanism of the disorders. How exactly estrogens affect various disorders is a fascinating and intriguing aspect of this emerging field of non-reproductive brain actions of gonadal hormones. Among the estrogens, especially estradiol appears to play an important and multimodal role in the brain. Which of estradiol’s many membrane, intracellular, and genomic actions matter most in psychiatric disorders, remains to be discovered. The aim of this volume is to summarize the role estrogens play in major psychiatric disorders, such as schizophrenia, depression, and dementia, and to provide a state-of-the-art overview of current knowledge but also of open questions. We hope that it will be a useful resource for clinicians and readers who are interested in contemporary research developments in this field. Helmut Vedder and Christian Behl (“Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology”) begin this volume with a chapter on the basic mechanisms of estrogens, also presenting an overview of the genomic and nongenomic effects of estrogens and of the modulatory effects of estrogens on the various neurotransmitter systems, particularly the glutaminergic, dopaminergic, serotonergic, and cholinergic systems. Furthermore, the neuroprotective effects of estrogens and their immunological activation in the central nervous system during neurodegeneration and aging are discussed.

VII

N. Bergemann and A. Riecher-Rössler

The following twelve chapters are divided into three parts. The first part focuses on the impact of estrogens in schizophrenia and contains five chapters on the role of estrogens in the development and course of schizophrenia. This section starts with chapter 2 by Anita Riecher-Rösssler (“Estrogens and Schizophrenia”), which gives an overview on the impact of estrogens in schizophrenia. She first presents a historical view of this issue and then summarizes the findings from various trials on the “estrogen protection hypothesis” and the “hypoestrogenism hypothesis” in schizophrenia. She closes her chapter with a discussion of the implications for therapy and prophylaxis. In chapter 3, Heinz Häfner gives a detailed review on “Gender Differences in Schizophrenia”, in particular concerning the differences in diagnosis, subtypes and symptoms, lifetime risk and age at onset, risk factors, illness behavior and illness-specific deficits, and both course and outcome. In his review he refers to the findings of the ABC study, which he conducted together with Anita Riecher-Rössler in a first-episode sample of schizophrenia. The data presented here are of major importance for the “estrogen protection hypothesis” in schizophrenia. Robin Z. Hayeems and Mary V. Seeman (chapter 4: “Puberty and Schizophrenia Onset”) then report on a study showing the inverse relationship between onset of menarche and schizophrenia onset in women. In chapter 5, Jayashri Kulkarni and co-workers (“Clinical Estrogen Trials in Patients with Schizophrenia”) present findings which show that add-on therapy with estradiol to the standard treatment of acute schizophrenia alleviates symptoms and shortens the acute hospital stay. Niels Bergemann and co-workers (chapter 6: “Hypoestrogenism and Estrogen Replacement Therapy in Women Suffering from Schizophrenia”) report on studies providing evidence for the “hypoestrogenism hypothesis” that show menstrual irregularities and pathologically low estradiol blood levels throughout the menstrual cycle as well as anovulation in the majority of women suffering from schizophrenia. Furthermore, the results of estrogen replacement as a prophylactic adjunct to antipsychotics in women with schizophrenia are reported. The second part of this book examines the role of estrogens in depression. Uriel Halbreich and Linda S. Kahn open this section with an overview on “The Effect of Estrogens on Depression.” They start with a summary of the gender differences in the prevalence of depression and regarding the response to psychotropic medication and then outline the role of estrogens in the treatment of major depres-

VIII

Introduction and Overview

sion and reproduction-related disorders in women. In the following chapter. 8, Leslie Born and Meir Steiner take a closer look at “Estrogens and Other Hormones in the Treatment of Premenstrual Syndromes.” Alain Gregoire (chapter 9: “Estrogens and Perinatal Disorders”) gives an overview of the evidence for antidepressive effects of estrogens in postpartum disorders and summarizes the implications this has for research and clinical practice. Gabriela Stoppe and Martina Dören (chapter 10) then critically discuss the current findings regarding “Estrogen Therapy in Perimenopausal Affective Disorders.” The third part of the book focuses on the important association between estrogens and cognition and/or Alzheimer’s disease. Tony Edwin and Uriel Halbreich give a concise overview on “The Effects of Estrogens on Cognition and Alzheimer’s Dementia” (chapter 11). Pauline M. Maki and Susan M. Resnick follow with a review of neuroimaging studies in this field (chapter 12: “Effects of Hormone Therapy on Patterns of Brain Activation during Cognitive Activity: A Review of Neuroimaging Studies”). They conclude that neuroimaging studies give evidence for a protective effect of estrogens on age-related changes in cognition and Alzheimer’s disease. Findings suggest that estrogens affect neural substrates of several cognitive functions. Peter Schönknecht and co-workers (chapter 13: “Estrogens in Alzheimer’s Disease: A Clinical and Neurobiological Perspective”) elaborate on the question of whether estrogen might have a mediating effect on cerebral β-amyloid metabolism in Alzheimer’s disease. Finally, the last chapter of this book by Khaled M. K. Ismail, G.V. Sunanda and P. M. Shaughn O’Brien deals with the “Interface between Gynecology and Psychiatry” regarding estrogen therapy. The present volume includes both reviews on the main topics and also chapters providing more detailed information on individual studies. Thus, some overlap may occur and different views and interpretations of empirical findings become evident. It is not intended to rule out such discrepancies – on the contrary: different views may stimulate the ongoing discussion. Overall, it is obvious that interest is growing in this area and that further research is required since clinical studies, especially intervention studies, are still rare. Results of larger-scale, controlled studies are needed before recommendations for routine clinical application of estrogens can be made for psychiatric patients. In addition,

IX

N. Bergemann and A. Riecher-Rössler: Introduction and Overviews

research needs to be undertaken on the best modality of hormone (replacement) therapy in psychiatric patients. Clearly, any decision for estrogen (replacement) therapy in women with psychiatric disorders must be made on the basis of an individual risk-benefit-assessment, evaluating the pros and cons. In this context, the results of the “Women’s Health Initiative” need to be considered, which should also include a critical evaluation of the study results and interpretations (Notman and Nadelson 2002; Schneider 2002; Writing Group for the Women’s Health Initiative Investigators 2002; Writing Group of the International Menopause Society Executive Committee 2004; Turgeon et al. 2004). We strongly believe that further research on the influence of estrogens in schizophrenia, affective disorders, and dementia will not only contribute to our understanding of the pathogenetic mechanisms underlying some aspects of psychiatric disorders, but will also have direct therapeutic benefits for many women with these disorders. Many people have contributed to this volume. In particular, we express our gratitude to our fellow authors. We also gratefully acknowledge the expert tutelage, guidance, and patience of the staff at SpringerWienNewYork, especially Raimund Petri-Wieder. Furthermore, we owe many thanks to Sherryl Sundell for editing the manuscripts. Last but not least, the sponsoring of this volume by AstraZenecaGmbH, Pfizer GmbH, and Lilly Deutschland GmbH through unrestricted grants is greatly appreciated.

References Notman MT, Nadelson CC (2002) The hormone replacement therapy controversy. Arch Women Ment Health 5: 33-35 Schneider HPG (2002) The view of the International Menopause Society on the Women’s Health Initiative. Climacteric 5: 211-216 Turgeon JL, McDonnel DP, Martin KA, Wise PM (2004) Hormone therapy: physiological complexity belies therapeutic simplicity. Science 304: 1269-1273 Writing Group for the Women’s Health Initiative Investigators (2002) Risks and benefits of estrogen plus progestin in healthy postmenopausal women. JAMA 288: 321-333 Writing Group of the International Menopause Society Executive Committee (2004) Guidelines for the hormone treatment of women in the menopause transition and beyond. Climacteric 7: 8-11

X

1 Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology Helmut Vedder and Christian Behl

Introduction: The Role of Estrogens in Human Neuropsychiatric Diseases A large number of studies, which often focus on the effects of estrogen replacement therapy (ERT) in women, have reported beneficial actions of these hormones on various neurobiological and neuropathological parameters in health and disease (Fig. 1). It is likely that postmenopausal ERT helps reduce the risk of Alzheimer’s disease (Kawas et al. 1997), improves cognitive and affective functions (Schmidt et al. 1996), including postmenopausal depressive symptoms (Halbreich 1997), and affects the course of other illnesses such as schizophrenia (Riecher-Rössler and Häfner 1993; Riecher-Rössler et al. 1994) and probably also stroke (Toung et al. 1998; Yang et al. 2000) and Parkinson’s disease (Blanchet et al. 1999; Saunders-Pullman et al. 1999). Unfortunately, these data have often not been clearly confirmed by appropriately designed double-blind studies due to inherent problems such as the necessity for long-term evaluation and controlling for confounding factors. Moreover, a recent treatment study under rigorously controlled conditions failed to show positive effects of estrogen treatment in patients already affected by Alzheimer’s disease (Mulnard et al. 2000). In contrast, a large number of preclinical neurobiological studies have unequivocally demonstrated neuroprotective effects of this group of hormones during numerous toxic states and under various different conditions (Behl and Holsboer 1999). Up to now, it is not at all clear which of the various effects of estrogens on cells of different systems (neuronal, vascular, immune, and others) contribute to what extent to the neuroprotective effects of these hormones. It is likely that several of these described actions represent similar cellu-

1

Helmut Vedder and Christian Behl

lar mechanisms or that the effects, for example, on the vascular and the neuronal systems promote neuroprotection. Although the dualism between “slow classical receptor-mediated genomic effects” and the rather rapidly occurring so-called “nongenomic” effects is still apparent, the existence of membrane receptors similar to the nuclear receptors has been suggested. It is possible that these may mediate the so-called “nongenomic” effects of these hormones (Weiss and Gurpide 1988; Fiorelli et al. 1996). On the other hand, estrogens are also able to induce alterations in the redox state or changes of the electrophysiological state of a cell and it is possible that these effects may result in changes in gene-expressions patterns. Therefore, the network of the multiple activities of estrogens must be studied and new categories defined to characterize and to evaluate the data on the neuroprotective actions of these hormones. Hopefully, this will be possible in the near future, to allow not only for a clearer understanding of the described effects but also to gain further perspective for the clinical use of estrogen compounds. In this chapter, we focus on the basic mechanisms of the actions of estrogens, which are likely to play an important role in the pathology of neuropsychiatric diseases and to contribute to the improvement of neuronal survival and function in general. Several of

Estrogen Replacement Therapy Cognition

Mild cognitive impairment

Parkinson’s disease

Schizophrenia

Stroke

Depression

Alzheimer’s disease Fig. 1. ERT affects the pathogenesis of various neuropsychiatric disorders

2

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology

these mechanisms of actions may be conclusively related to specific disease processes such as Alzheimer’s disease or to the toxic effects of free radicals in the course of stroke and other diseases. Moreover, estrogen-induced changes in neurotransmitter systems such as the NMDA excitatory system or the dopaminergic system suggest possible effects in diseases such as Parkinson’s disease and schizophrenia. In addition, even clinical studies support an interaction of estrogens with cholinergic functions and dysfunctions, since comedication with estrogens has been shown to induce advantageous effects in the treatment of memory deficits, including those occurring in conjunction with Alzheimer’s disease (Schneider at al. 1997). In a large number of preclinical and basic neurobiological studies, the neuroprotective potential of this group of hormones has been clearly shown (for recent reviews see: Behl and Holsboer 1999; Green and Simpkins 2000). Several effects, including the interactions of estrogens with the cellular Ca++ metabolism or with important factors of cellular metabolism such as the signaling cascades involving cAMP and MAP-kinase, must be further examined in the future with regard to their neuroprotective potential. Additional clinical studies still need to be conducted to clarify whether the protective effects of estrogens observed in vitro also apply to the in vivo situation, whether the basic neuroprotective effects are also detectable in the human system and which co-factors have to be controlled or modified to optimize treatment effects. In addition, we need to analyze whether novel estrogen derivatives lacking genomic activity can be designed and are more likely to be used in humans. Since the question of which particular functions contribute to the beneficial effects of estrogens is still open, the different mechanisms of actions under which neuroprotection occurs need to be analyzed and this knowledge applied in clinical treatment and trials.

Basic Mechanisms of Estrogen Action: Genomic Effects of Estrogens via the “Classical” Estrogen Receptors Estrogen receptors (ER) belong to the family of nuclear steroid receptors (Evans 1988) and, therefore, many of the effects of estrogens are mediated via the “classical genomic way of steroid hormone

3

Helmut Vedder and Christian Behl

Fig. 2. “Classical” genomic pathway of estrogen action as transcription factor

action” (Fig. 2). Estrogens readily enter and cross the cellular membrane due to their lipophilicity and bind to the intracellular “high affinity estrogen hormone receptors” (ER). The formation of the “hormone-receptor complex” then leads to the binding of the complex to “hormone-responsive elements” (HRE) at regulatory sequences of the genomic DNA after dissociation of HSP-90 shock protein and dimerization of receptors. This binding subsequently induces the regulation of the HRE-regulated genes, probably via transient induction of nuclear proto-oncogenes (Yamashita 1998). In addition, a variety of co-activator and co-repressor proteins that further affect estrogen-responsive genes, the cellular occurrence, and the state of the unbound receptor molecule represent important factors for the effects of estrogens (Shibata et al. 1997). Other co-factors and co-repressors are also involved in the regulation of gene expression by steroid hormones, leading to complex actions of various elements converging on the changes in gene transcription (Shibata et al. 1997; Klinge 2000). Overall, these data show that the genomic effects of estrogens not only regulate the activity of single genes, but also induce more complex changes in the metabolism of the re-

4

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology

sponsive cell via the induction of gene-expression patterns (Russell 1996). Up to now, at least two ERs, ER-α and ER-β, have been identified (for a recent review see: Gustafsson 2000). ER-α appears to be only weakly expressed in certain areas of the hippocampus, whereas ER-β is found more abundantly in this region (Shughrue et al. 1997). In addition to the hippocampus, estrogens may also affect other regions and various transmitter systems in the brain that express ERs, including the basal forebrain transmitter systems for acetylcholine and 5-HT and the dopamine and the norepinephrine system (Das and Chaudhuri 1995; Mudd et al. 1998; McEwen and Alves 1999). ER-induced gene expression via the classical pathway may participate in the neuroprotective actions of estrogens since the genes for the neurotrophin brain-derived neurotrophic factor (BDNF) (Singh et al. 1995; Sohrabji et al. 1995) are regulated by estrogens and have been demonstrated to exert neuroprotective properties. Some studies of the estrogen antagonist tamoxifen support a role of the classical ERs in neuroprotective paradigms (Singer et al. 1996; Green et al. 1997). Despite this, these results are by far not applicable to all studies in this area and other mechanisms, including nongenomic effects, also contribute – most likely to a much larger extent – to the neuroprotective effects of these hormones (Behl and Holsboer 1999; Moosmann and Behl 1999; Green and Simpkins 2000). In addition, most of the work up to now has been done on the ER-α. Therefore, further studies are required to clarify the role of other ERs in genomicallymediated neuroprotective actions of estrogens. Interestingly, the transcription of the bcl-2 gene, which is significantly involved in apoptotic cell death, is affected by estrogens via a cAMP response element in the promotor region (Dong et al. 1999), showing that the genomic effects of these hormones are also mediated by mechanisms other than the classical binding of receptors to the hormone-responsive DNA elements. This means that estrogens are also able to indirectly modulate gene transcription by influencing intracellular signaling pathways. Increasing knowledge about the selective expression and coexpression of ERs and their pharmacology and the elucidation of the mechanisms of the neurotrophic activity of estrogens via their genomic effects suggests that these hormones and their receptors could prove to be important novel targets in the search for neuroprotective agents.

5

Helmut Vedder and Christian Behl

Basic Mechanisms: The Importance of Nongenomic Effects for the Neuroprotective Actions of Estrogens As already discussed, increasing data have revealed that other mechanisms of action beside the “classical” genomic effects of estrogen hormones are also effective: • Under certain conditions, estrogen receptor antagonists are not, or only partly, effective in neutralizing the neuroprotective actions of these hormones (Green et al. 1997; Regan and Guo 1997; Moosmann and Behl 1999; Culmsee et al. 1999). • Moreover, protein (Goodman et al. 1996; Regan and Guo 1997; Sawada et al. 1998) or mRNA synthesis inhibitors (Goodman et al. 1996) do not interfere with the neuroprotective actions in several toxicity paradigms. • Structure-activity studies demonstrate different structural requirements for the effects on cellular ERs (Korenman 1969; Wiese et al. 1997) and the neuroprotective effects with regard to different cytotoxic paradigms, including free radical-induced cytotoxicity (Behl et al. 1997; Green et al. 1997; Moosman and Behl 1999) in different neuroprotection models. Moreover, cells such as the clonal mouse hippocampal cell line HT 22 that do not show a classical ER response are also protected by estrogens after cytotoxic challenge (Behl et al. 1995; Green et al. 1998; Vedder et al. 2000). • Cell-free cytotoxic paradigmatic approaches such as the induction of lipid peroxidation in cell or brain homogenates also point to beneficial actions of estrogens via this pathway under in vivo conditions, excluding the necessity for a functioning cellular metabolism as required for the genomic actions (Sergeev et al. 1974; Goodman et al. 1996; Vedder et al. 1999). These results clearly demonstrate that both genomic actions of estrogens, including those via the steroid receptors, and a variety of other nongenomic effects mediate and contribute to the structural and functional neuroprotective effects of these hormones.

6

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology

Modulatory Effects of Estrogens on Neurotransmitter Systems (Glutamate, Dopamine, Serotonin, Acetylcholine) and Neuronal Excitability In the central nervous system, a large number of cellular activities and functions are influenced by the excitatory status of the cell membrane and the effects of neurotransmitters that, to a large extent, modulate this state. Glutamate, the major excitatory transmitter, may cause neuronal cell death under certain conditions, such as cerebral ischemia after activation of N-methyl-D-aspartate (NMDA) and kainate/α-amino3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) binding sites (Choi 1992; Choi 1996). Estrogens have been examined with regard to their interference with this type of cytotoxic paradigm and both interactions with NMDA-evoked excitotoxicity and membrane effects have been demonstrated in this context (Weaver et al. 1997; Regan and Guo 1997). Additionally, genomic effects may also play a role in the inhibition of glutamate toxicity, since the antiestrogen tamoxifen was able to block the neuroprotective effects of the hormone – at least under certain experimental conditions (Singer et al. 1996). Interestingly, there are also data on the enhancement of the activity of NMDA effects by estradiol, most likely in a region-specific manner (Wooley et al. 1997; Foy et al. 1999; Cyr et al. 2000), with a focus on the hippocampal CA 1 region, the frontal cortex, and the nucleus accumbens (Cyr et al. 2000). Moreover, estradiol has been demonstrated to potentiate the depolarizing influences of AMPA, kainate, and quisqualate, but not NMDA (Wong and Moss 1992), pointing to short-term actions of estrogens on non-NMDA receptors. Unfortunately, the data are not conclusive yet; thus, these underlying effects may be influenced by other yet undefined factors. Inhibitory transmitters such as GABA or other excitabilitydecreasing substances may reduce the cytotoxicity of NMDA. Weiland et al. (1992) showed that estrogens induce the mRNA levels for glutamic acid decarboxylase (GAD), the GABA-synthesizing enzyme in the CA1 pyramidal cell layer of the hippocampus. This effect may also contribute to a reduction in excitatory amino acid-evoked cytotoxicity by increasing the inhibitory neurotransmission via a genomic effect on the GABA-ergic system. Direct membrane effects have been demonstrated with regard to Ca++ currents (Joels and Karst

7

Helmut Vedder and Christian Behl

1995; Mermelstein et al. 1996), probably related to the activation of G-proteins (Mermelstein et al. 1996). Recently, molecular biological data have demonstrated that both putative membrane and nuclear ERs for ER-β can be derived from a single transcript (Razandi et al. 1999), further supporting the concept of direct membrane effects of these steroids even via the same receptor, which mediates the effects of estrogen at the cell nucleus. Overall, the data on potential estrogen membrane receptors are presently not conclusive. Other results suggest that at least some of the membrane effects of estrogens might also be mediated by so-called caveolae, specific membrane structures (Toran-Allerand et al. 1999). The dopaminergic system is a transmitter system which is involved in addiction and reward processes and acts as a neuromodulator (Raevskii 1997). Effects of estrogens on the dopaminergic system include a decrease in dopaminergic neurotransmission and a subsequent increase in dopaminergic binding sites (Di Paolo 1994; Bosse and DiPaolo 1996), supporting a physiological and possibly also a pathophysiological role of this mechanism in schizophrenia. Moreover, estrogen treatment increases the concentrations of dopamine in the striatum (Becker 1999) and modulates the sensitivity and also the number of striatal D2 receptors (Lammers et al. 1999). These effects have also been supported by functional studies on apomorphine- and haloperidol-induced dopamine-mediated behavior such as stereotypy and catalepsy (Häfner et al. 1991, Gattaz et al. 1992). These data clearly show effects of these hormones on the dopaminergic system and support a role of estrogens in the modulation of the pathophysiology of diseases such as schizophrenia, Parkinson’s disease, and probably even addiction. Effects of estrogens on the serotonergic system have also been described, showing a significant increase in the density of 5-HT 2A binding sites in different brain areas (Fink et al. 1998) and inhibition of serotonin reuptake in synaptosomal preparations of the rat cortex (Michel et al. 1987). Other actions may include direct effects of the hormones on serotoninergic neurons as suggested by autoradiography studies on the binding sites in the midbrain region (Stumpf and Jennes 1984). Moreover, a study on hormonal responses in postmenopausal women with or without estrogen replacement even suggested a modulation of serotonin agonist-evoked effects by these hormones (Halbreich et al. 1995). Other effects of estrogens have been detected on the neuroten-

8

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology

sin/neuromedin gene (Watters and Dorsa 1998). This gene is induced by estrogens in the preoptic area. Since the promoter region lacks an estrogen-responsive element, it is likely that the changes in gene-expression are mediated by other mechanisms, such as interaction with the cAMP system.

Specific Interactions of Estrogens with Disease Processes: Effects of Estrogens on Cholinergic Functions The cholinergic system consists of neurons which synthesize the transmitter acetylcholine and innervate most parts of the neocortex, including the hippocampal formation. Cholinergic neurons are mostly localized in the basal forebrain. One of the areas involved in the innervation of the hippocampal formation is the nucleus basalis magnocellularis Meynert. In the course of Alzheimer’s disease, cholinergic functions decrease with a concomitant impairment of mnestic and cognitive functions. Interestingly, estrogens influence cholinergic functions by several mechanisms (Gibbs and Aggarwal 1998; McEwen and Alves 1999). The therapeutic relevance of these mechanisms is further supported by the additive effects of a co-medication with estrogens in the course of the treatment of Alzheimer patients with the acetylcholineesterase inhibitor tacrine (Schneider et al. 1996). The growth and the function of cholinergic neurons highly depends on the influence of nerve growth factor (NGF), a protein which is synthesized and secreted by neurons and glial cells. Interestingly, effects of estrogens have not only been detected with regard to cholinergic functions, but also with regard to the NGF system. Estrogen treatment promotes an upregulation of the acetylcholine-producing enzyme choline acetyltransferase (ChAT) and a subsequent downregulation of the receptors for NGF, p75, trkA, and the NGF protein itself (Gibbs et al. 1994). Sohrabji et al. (1994) even demonstrated a reciprocal regulation of the two systems in PC 12 cells: long-term treatment with NGF induced a six-fold increase in estrogen binding, whereas subsequent co-treatment with estrogen and NGF led to an induction of the NGF receptor mRNA for trkA and down-regulation of the p75 mRNA. This was also confirmed by another study, even with a specificity of the effects for females (Gibbs and Pfaff 1992). Taken together, the data show an induction of ChAT by estrogens and

9

Helmut Vedder and Christian Behl

a down-regulation of the NGF system. In addition to the functional actions on the cholinergic system, synergistic neuroprotective effects have been shown after treatment with estrogens and NGF in a paradigm of serum-derived PC 12 cells (Gollapudi and Oblinger 1999). Induction of cholinergic cell death by deposition and toxicity of ß-amyloid represents an important hallmark of Alzheimer’s disease (Hardy and Allsop 1991; Yankner 1996; Behl 1999). Interestingly, estrogens are able to effectively interact with ß-amyloid-induced cell death, most likely in a direct manner, by effectively interfering with free radical-mediated impairment of cellular functions (Behl et al. 1995, 1997; Behl and Holsboer 1999; Vedder et al. 2000).

Cross-Talk of Estrogens with Intracellular Factors and Signal Transduction Mechanisms cAMP, CREB A large number of peripheral influences on the cell converge at several intracellular transduction pathways. One of these pathways includes the second messenger cyclic-AMP (cAMP). Changes in the intracellular concentrations of this molecule induced by the binding of extracellular mediators to their corresponding membrane receptors affects gene transcription via the cAMP response element-binding protein (CREB) (Hagiwara et al. 1996). Estradiol has been shown to increase cAMP concentrations in different cellular systems (Minami et al. 1990; Gu and Moss 1996; Watters and Dorsa 1998, Kelly et al. 1999), including hypothalamic neuronal cells (Gunaga and Menon 1973). Moreover, the functional character of these changes has been demonstrated by an increase in the phosphorylation of the CREB protein. With regard to the underlying receptor systems, no clear-cut results were obtained by the use of receptor antagonists (Gu et al. 1996; Watters and Dorsa 1998). In a recent review, Green and Simpkins (2000) discussed the importance of the phosphorylation step of CREB as a convergence point not only for the cAMP pathway, but also for at least two other pathways, the MAP kinase pathway (Singh et al. 1999) and the CAM kinase pathway (Matsuno et al. 1997), which have also been shown to be activated after exposure to estradiol. They pointed out that an activation of the cAMP pathway is associated with a decreased susceptibility of neuronal cells to several types of apoptotic signals

10

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology

(D’Mello et al. 1993; Kew et al. 1996; Campard et al. 1997; Kobayashi and Shinozawa 1997), indicating a role of this mechanism in the prevention of cell death. Moreover, increased phosphorylation of CREB is related to an increased resistance to ischemic injury (Walton et al. 1996) and may activate the MAP kinase (Vossler et al. 1997). Interestingly, the transcription of the bcl-2 gene, which is strongly involved in apoptotic cell death, is also affected by estrogens (Garcia-Segura et al. 1998; Singer et al. 1998, Dong et al. 1999), most likely via a cAMP response element in the promotor region (Dong et al. 1999). Thereby, the cAMP pathway most likely contributes to genomic changes induced by estrogens. These effects may then induce or favor a state of decreased susceptibility of neuronal cells to damaging insults via cAMP increase, affording effective neuroprotection under certain conditions. MAP Kinase and Other Kinases Another important pathway for the intracellular transduction of signals is the mitogen-activated protein kinase (MAP ) pathway. After tyrosine phosphorylation of cellular proteins, the MAP kinases ERK1 (extracellular-signal-related kinase-1) and ERK-2 are activated and evoke further cellular effects, e. g., the cellular response to peptide growth factors such as nerve growth factor (NGF) or BDNF. These factors then act on nerve and other cells via the receptor molecules p75 or trkA. Subsequent cellular signaling takes place by intermediate proteins such as ras and b-raf and subsequently affects the phosphorylation state of a variety of proteins via MAP kinase activity. Actions of estrogens on this pathway have been shown by at least two groups and under various different conditions and support a membrane specificity of the effects (Watters et al. 1997; Singh et al. 1999). These data indicate that such actions occur quite rapidly in the minute range and are not blocked by the estrogen receptor antagonist ICI 182.780 –, at least under in vitro conditions with primary neocortical neurons (Singh et al. 1999). Further available data on this issue allow for the speculation that the activated ER induces and increases the B-raf kinase, leading to the formation of an multimeric complex of ERs, hsp90, and B-raf kinase (Singh et al. 1999; Toran-Allerand et al. 1999). We have recently been able to show that the activation of MAPK (ERK-1/ERK-2) by 17β-estradiol directly mediates the enhanced re-

11

Helmut Vedder and Christian Behl

lease of the nonamyloidogenic form of the amyloid precursor protein (APP) (Manthey et al. 2001). The metabolism of APP (amyloidogenic versus non-amyloidogenic processing) is believed to be one of the key events in the pathogenesis of Alzheimer’s disease. Interestingly, very recent data (Honda et al. 2000) show that phosphatidylinositol 3-kinase (PI3-K) – as a result of ER receptor activation by incubation with 10 nM of 17β-estradiol – may lead to a phosphorylation of Akt/phosphokinase B, thereby significantly contributing to the neuroprotective effects of estrogens in glutamate-treated cultured cortical neurons via this pathway. Bcl-2-Related Proteins Cellular apoptosis describes a specific type of cell death, which starts from the cellular nucleus, and is characterized to a significant extent by the degradation of the cellular DNA. During gel electrophoretic analysis of the cellular material, the “laddering phenomenon” can be detected, which indicates the specific effects of DNA-degrading enzymes. The DNA degradation is executed by caspases, enzymes that exist in inactive pro-forms and are activated by apoptotic stimuli (Wolozin and Behl 2000). The induction of apoptotic processes is controlled by specific inhibitors (bcl-XL, bcl-2) and promoters (bax, bad, bcl-Xs), which are members of the so-called Bcl-2 family of proteins. Not only does the occurrence of one or the other factors determine the fate of the cell, but their relative concentrations lead to a cellular state towards apoptosis, either under normal conditions or after challenge with toxic stimuli. Thereafter, more complex mechanisms take over and lead the cell into apoptosis (Merry and Korsmeyer 1997). Inhibitors of apoptotic cell death such as bcl-2 may be affected by classical hormone action via the ERs, since several of their genes contain HREs (Dong et al. 1999; Perillo et al. 2000). Functionally, an increase in the mRNA of these proteins has been demonstrated after treatment with estrogens in neuronal NT 2 cells (Singer et al. 1998) and in the hypothalamus of female rats (GarciaSegura et al. 1998). With regard to both clinical and mechanistic aspects, the group of Dubal (1999) reported changes of the mRNA for bcl-2 in an animal model of stroke after administration of estrogens. Further data point to complex interactions between the cellular levels of the classical

12

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology

receptors for estrogens (ER-α and ER-β), bcl-mRNA levels, and estrogen-induced changes in the ischemic penumbra (Dubal et al. 1999). Nuclear Transcription Factor κB (NF-κB) Nuclear Transcription Factor κB (NF-κB) is a redox-sensitive inducible transcription factor that positively regulates the expression of proimmune and proinflammatory genes, while glucocorticoids are potent suppressors of such responses (McKay and Cidlowski 1998). With an in vitro approach, Shyamala and Guiot (1992) showed that estrogens activate NF-κB-specific proteins and postulated an interaction between these proteins and ERs. Further studies with glucocorticoid receptors and NF-κB revealed that these effects are indeed mediated by physical interaction, resulting in a repression of NF-κB transactivation. Subsequently, this was also shown for estrogen, progesterone, and androgen receptors (McKay and Cidlowski 1998). Functionally, Dodel et al. (1999) demonstrated that estradiol attenuated the amyloid-β and LPS-induced translocation of NF-κB in cultured rat astrocytes, pointing to a possible clinical relevance of this mechanism in Alzheimer’s disease. Because NF-κB is also thought to be directly involved in the survival of nerve cells under conditions of oxidative stress (Lezoualc’h and Behl 1998), an interaction between NF-κB and estrogens might even directly affect cell survival at a very basic cellular level. Cellular Ca++ Levels Intracellular Ca++ levels are under strict control by various cellular mechanisms (Racay et al. 1996). Multiple factors affect these concentrations, leading to morphological and functional changes in the cell (Sola et al. 1999). Thereby, Ca++ also functions as an important messenger system in the cell, even affecting cellular survival and cell death under various conditions (Leist and Nicotera 1998). Interestingly, Ca++ concentrations seem to be able to directly influence cellular gene expression, including important genes such as the CREB gene (Bito et al. 1997; Hardingham and Bading 1998). A lack of control of the intracellular Ca++ concentrations mostly results in an increase or a profound dysregulation of cellular Ca++ concentrations and is induced under conditions of stroke or after treatment of cells with β-amyloid (Mattson et al. 1993). If the homeostatic mechanisms are not able to counteract this increase, the

13

Helmut Vedder and Christian Behl

loss of cellular functions due to the Ca++ dysregulation leads to cell death (Tymianski and Tator 1996; Leist and Nicotera 1998). Estrogens may affect cellular Ca++ levels by different mechanisms: direct membrane effects (Joels and Karst 1995; Mermelstein et al. 1996; Pozzo-Miller et al. 1999), interactions with different NMDA receptors (Wong and Moss 1992; Weaver at al. 1997; PozzoMiller et al. 1999), effects on intracellular mechanisms of Ca++ regulation via calmodulin (Hayashi et al. 1994), or direct effects on intracellular stores (Beyer and Raab 1998). Functional aspects with regard to neuroprotection have been shown in the attenuation of Ca++ increases by estrogens after challenge of motor neurons with glutamate (Kruman et al. 1999) as well as in a model of cerebral ischemia in the gerbil (Chen et al. 1998). It is still not known whether the actions of estrogens on this cellular parameter are a relevant factor in neuroprotection or just an epiphenomenon in the course of other cellular events.

Structure-Related Intrinsic Neuroprotective Effects of Estrogens Antioxidative Actions of 17β-Estradiol Free radicals are a group of molecules which are generated at several places in the course of cellular metabolism, paticularly in energyproviding reactions in the mitochondria. These molecules are highly reactive and interact with a large number of cellular compounds and cause structural and functional changes in the cell (Halliwell and Gutteridge 1990; Sies 1997; Gutteridge and Halliwell 2000). Usually, the intracellular amount of free radicals is controlled very strictly by a number of mechanisms, such as by different enzymes – superoxide dismutase, gluthatione peroxidase, and catalase – as well as by other components – ascorbic acid and vitamin E. Increases in free radical levels in the cell may either be evoked by impaired detoxification mechanisms or by an increased production under pathological conditions such as degenerative disease processes or stroke (Coyle and Puttfarcken 1993). Accumulating cell damage from the metabolism of free radicals has been connected to the aging process: this introduced the socalled “free radical hypothesis of aging” (Harman et al. 1976; Beck-

14

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology

man and Ames 1998). Since free radicals are highly likely to play a substantial role in the pathophysiology of neurodegenerative diseases, including Alzheimer’s, exacerbation of cellular damage in the course of these diseases has to be postulated as an important factor, aggravating the “physiological changes of aging” to the more severe disease process of dementias and other neurodegenerative disorders (Harman et al. 1976). This assumption is further supported by data showing that several products of oxidation reactions and mediators of oxidative stress can be found in association with the histopathological hallmarks of Alzheimer’s disease and vascular dementia, mainly in the “senile plaques”. Such oxidation end products include malondialdehyde, advanced glycation endproducts (AGEs), carbonyls, nitrotyrosine, and various other oxidized molecules (Pappolla et al. 1992; Beal 1995; Smith et al. 1996). The increased state of cellular oxidation as revealed by such an increase in the oxidation of proteins and lipids (Hajimohammadreza and Brammer 1996) may then lead to longterm changes such as alterations in enzyme activities and structural effects on membrane integrity and the oxidation of DNA with possible long-term mutagenic effects (Ames 1988). Estrogens and related molecules have been demonstrated to exhibit intrinsic antioxidant activity, since they protect neuronal cells under in vitro conditions against free radical-induced cell death. This is connected to specific properties of the molecule, e.g., the phenolic structure, since all components with such a structure act as efficient neuroprotectants under these conditions (Moosmann and Behl 1999) (Fig. 3). Modifications of this particular moiety by etherization (e.g., mestranol, methyl ether of ethinyl oestradiol) block the antioxidant activity of the molecule. Furthermore, various aromatic alcohols with intact phenolic groups but without ER-activating properties (e.g., dodecyl phenol) also possess intrinsic antioxidant neuroprotective activity and also prevent lipid peroxidation (Moosman and Behl 1999). Although the physiological relevance of the concentrations (nanomolar to millimolar) that are required for antioxidant neuroprotection in vitro are still a matter of debate, it is clear that the neuroprotective activity of estrogens can be structurally separated from their ER-activating classical properties as hormones. Moreover, several factors may contribute to the efficacy of this mechanism: First, the concentration of estrogens is highly variable,

15

Helmut Vedder and Christian Behl

Fig. 3. Estradiol is a phenolic free radical scavenger/antioxidant similar to vitamin E

depending on the menstrual cycle, and might reach lower nanomolar concentrations under in vivo conditions. Secondly, in general, steroid compounds, including ovarian steroids, are concentrated several-fold in the brain in relation to the plasma levels. Moreover, the turnover rate of brain sequestration of blood-borne sex steroids is high compared to other steroid compounds such as corticosteroids (Pardridge et al. 1980). Thirdly, drugs used for effective treatment might be able to reach pharmacological levels rather than physiological concentrations due to an effective administration regimen. And fourthly, other estrogen or phenolic compounds may be found, which require lower tissue concentrations for an effective interaction with free radical-induced cell damage such as the catecholestrogens (Trepker et al. 2003). Inhibition of Lipid Peroxidation The central nervous system contains a substantial amount of membranes and fatty acids (Halliwell and Gutteridge 1990; Halliwell

16

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology

1992). This implies an increased vulnerability of membrane lipid constituents in the CNS to oxidative injury, either directly by cellular free radicals (Coyle and Puttfarcken 1993; Olanow 1993; Bondy 1995; Frölich and Riederer 1995; Sies 1997) or via other indirect or exogenic mechanisms (Nohl 1993; Piotrowski et al. 1996). The increased vulnerability of neuronal membranes may even be due to the low amount of integral proteins compared to other tissues, as proposed recently (Moosmann and Behl 2000). In this recent study, it was shown that the peptide stretches enriched in phenolic amino acids (tyrosine) protect neurons against oxidative cell death. As pointed out above, pro-oxidative mechanisms such as trauma, ischemia, and other illnesses increase the cellular load of free radicals (Olanow 1993; Bondy 1995). This increase in free radicals subsequently affects the membranes in the brain, resulting in increased lipid peroxidation (LPO), a loss of cellular compartimentalization (Nohl 1993) and finally in cell death. Therefore, LPO respresents an important cellular endpoint for a large number of toxic events in the CNS (Halliwell and Gutteridge 1990; Gutteridge 1995). Different approaches showed interactions of estrogen hormones with the peroxidation of lipids in non-neuronal systems, including lipid fractions (Sugioka et al. 1987), microsomal liver preparations (Ruiz-Larrea et al. 1994), and blood constituents such as low density lipoproteins (Miller et al. 1996). Up to now, most of these studies have been conducted on systems outside the CNS. Due to the importance of LPO processes in the CNS and the clinical relevance of their interaction with estrogens, we recently characterized the effects of estrogens on iron-induced LPO in different CNS-relevant systems: Hippocampal HT 22 cells and living rat neocortical brain cells were used as CNS-derived in vitro systems to study the direct cellular effectiveness of estrogens. Immortalized hippocampal cells of the HT 22 cell line have been used to evaluate the neuroprotective effects of estrogens after challenge with glutamate, iron, and hydrogen peroxide (Morimoto and Koshland 1990; Behl et al. 1995; 1997). Primary cultures of rat brain cells represent model systems for the evaluation of mechanisms of neurodegeneration and steroid-mediated neuroprotection (Choi et al. 1990; Vedder et al. 1993). In the course of earlier studies, whole brain homogenates were used to characterize LPO in the rodent and in the human CNS, respectively. We examined the effects of estrogens on LPO in these systems and were able to show an inhibition of this important

17

Helmut Vedder and Christian Behl

pathophysiological mechanism in all different systems, including the human brain (Vedder et al. 1999) These data are in line with other nonhuman data from the literature showing the LPO-inhibiting effects of estrogens under different conditions, including the LPO evoked by β amyloid (Gridley et al. 1997) and iron sulfate (Goodman et al. 1996).

Effects of Estrogens on Immunological Activation in the CNS: Immunological Activation During Neurodegeneration and Aging Estrogens affect a large variety of immune functions (Grossman 1984; Cutolo et al. 1995), including macrophage functions (Miller and Hunt 1996). These effects become pathophysiologically relevant in autoimmune diseases such as multiple sclerosis (Kim et al. 1999; Jansson and Holmdahl 1998). Another emerging focus is the immunological response in Alzheimer’s disease (McGeer and McGeer 1999) and its possible relevance for treatment (McGeer and McGeer 1996). Therefore, estrogens may influence microglial functions such as the expression and the secretion of cytokines in the brain (Mor et al. 1999). Interestingly, such effects may be mediated by the ER-β, which was detected in microglial cells in the brain (Mor et al. 1999). Up to now, the role of this modulation of the inflammatory response by estrogens has not been elucidated. In addition, the local immune response has also been shown to induce tissue damage via oxidation reactions either directly through the inflammatory mediators or through secondary events (McGeer and McGeer 1999). Particularly, activated microglia detected in Alzheimer’s disease tissue may powerfully induce oxidative damage via the released inflammatory mediators (Mor et al. 1999).

Neuroprotection as a Vital Byproduct of the Homeostatic Actions of Estrogens – A Hypothesis Studies of several parts of the effector mechanisms of estrogens indicate that these hormones might not only directly affect the cellular responses during physiological and pathological states, but exert stabilizing and homeostatic effects: With regard to the antioxidative

18

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology

actions after a cellular load with hydrogen peroxide, a free-radical inducing agent, estrogens acted only at higher concentrations of this toxic substance, but showed no effects at lower concentrations (Vedder et al. 2000). In neocortical neurons in culture, the dose-response profile for 17β-estradiol regulation of the macromorphological features exhibited a bimodal dose-response relationship whereas the dose-response profile for 17β-estradiol regulation of the micromorphological features displayed the more characteristic dose-response of an inverted V-shaped function (Brinton et al. 1997). In hippocampal neurons, the modulation of electrophysiological changes on the NMDA response – the major part of an excitotoxic challenge – by estrogens comparatively depends on the extent of the increase in NMDA activation (Murphy and Segal 1996; Wooley and McEwen 1994). Therefore, estrogens may only act under pathological conditions, supporting the homeostatic mechanisms in the cell and yielding a state of “decreased vulnerability” to damaging conditions and agents. This assumption fits with the observation that estrogens only delay illness processes such as Alzheimer’s disease and schizophrenia but are not able to directly neutralize the basic pathophysiological mechanisms. At the cellular level, they act on basic mechanisms such as modulation of excitotoxicity, free-radical detoxification mechanisms, and changes in gene transcription. Co-factors such as CREB, GSH, and other intracellular compounds are required for the actions of estrogens. Moreover, the application of estrogens often increases the detoxifying and antiapoptotic actions of these factors. Future research will have to further examine and subsequently describe a possible unifying concept of these estrogen actions and to weigh the relative importance of the individual factors and mechanisms involved in the pathophysiological events of the different disease processes.

Outlook: From Preclinical to Clinical Neuroprotective Effects of Estrogens A number of clinical data support a beneficial function of estrogens in health and disease. During the past few years, the effects of estrogens on the brain have gained increasing interest and has led to the ongoing characterization of the neuroprotective effects of this class

19

Helmut Vedder and Christian Behl

of hormones. Presently, a certain gap in our knowledge still exists regarding the presumed clinical effects on neuropsychiatric diseases such as Alzheimer’s and Parkinson’s disease and schizophrenia, on one hand, and the large number of preclinical neurobiological findings, on the other. Overall, authors consider estrogens more as preventive compounds rather than as drugs. This implies the necessity of advance administration or administration during the course of the illness and appears feasible since age-related disorders such as Alzheimer’s disease develop over decades. Once the disease process has started, it may be difficult to halt the pathogenetic mechanisms by physiological modulators such as estrogen, even if high concentrations are used. With respect to the basic neurobiological research, estrogens have been demonstrated to act as powerful neuroprotectants under a large variety of toxic challenge paradigms such as iron and amyloid-β induced neurotoxicity. Moreover, effects on transmitter systems such as the cholinergic and the glutamatergic system and on neurocellular morphology and functions also point to neuroprotective endpoints of the actions of these hormones. By examining the underlying cellular mechanisms in more detail, changes in secondmessenger systems and a number of signaling cascades have been detected, including interactions with apoptosis-inducing and -inhibiting proteins and nuclear transcriptions factors such as NF-κB. The very basic effector mechanisms include, in addition to the classical genomic and the nongenomic effects, interactions with free radical detoxifying systems and the inhibition of the cellular LPO, an important pathophysiological process in the brain with its large amount of membranes. The process of discovering further neuroprotective effects and their underlying mechanisms is still in progress. Presently, it does not seem that a general scheme of the effects is unraveling, a circumstance which may contribute to the lack of transfer from the basic neurobiological knowledge to the clinical application of these neuroprotective effects. Clinical trials will have to show whether the concept of antioxidants as illness-preventive drugs or as a therapy principle for neurodegenerative disorders, including Alzheimer’s disease, holds promise for the future. Overall, it is clear that the neuroprotective effects of estrogens represent more a general concept rather than just one or a few effects of one substance. The general applicability of the actions of estro-

20

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology

gens on the damaged brain – and even on other tissues – is the advantage of the underlying concept. On the basis of this concept, further research should give a more unifying picture of the pathways by which neuronal cell damage is conveyed under several pathophysiological conditions and identify the important effector points where interference with these processes could result in optimal and therefore the most effective neuroprotection.

Acknowledgement The authors thank A. Tittmar for substantial help with the editing of the manuscript and A. Thum for valuable suggestions.

References Ames BN (1988) Measuring oxidative damage in humans: relation to cancer and ageing. IARC Sci Publ 89: 407-416 Beal MF (1995) Aging, energy, and oxidative stress in neurodegenerative diseases. Ann Neurol 38: 357-366 Becker JB (1999) Gender differences in dopaminergic function in striatum and nucleus accumbens. Pharmacol Biochem Behav 64: 803-812 Beckman KB, Ames BN (1998) The free radical theory of aging matures. Physiol Rev 78: 547-581 Behl C, Holsboer F (1999) The female sex hormone oestrogen as a neuroprotectant. Trends Pharmacol Sci 20: 441-444 Behl C, Skutella T, Lezoualc’h F, Post A, Widmann M, Newton CJ, Holsboer F (1997) Neuroprotection against oxidative stress by estrogens: structure-activity relationship. Mol Pharmacol 51: 535-541 Behl C, Widmann M, Trapp T, Holsboer F (1995) 17beta-estradiol protects neurons from oxidative stress-induced cell death in vitro. Biochem Biophys Res Commun 216: 473-482 Beyer C, Raab H (1998) Nongenomic effects of oestrogen: embryonic mouse midbrain neurones respond with a rapid release of calcium from intracellular stores. Eur J Neurosci 10: 255-262 Bito H, Deisseroth K, Tsien RW (1997) Ca2 ± dependent regulation in neuronal gene expression. Curr Opin Neurobiol 7: 419-429 Blanchet PJ, Fang J, Hyland K, Arnold LA, Mouradian MM, Chase TN (1999) Shortterm effects of high-dose 17beta-estradiol in postmenopausal PD patients: an crossover study. Neurology 53: 91-95 Bondy SC (1995) The relation of oxidative stress and hyperexcitation to neurological disease. Proc Soc Exp Biol Med 208: 337-345 Bosse R, DiPaolo T (1996) The modulation of brain dopamine and GABAA receptors by estradiol: a clue for CNS changes occurring at menopause. Cell Mol Neurobiol 16: 199-212

21

Helmut Vedder and Christian Behl Brinton RD, Proffitt P, Tran J, Luu R (1997a) Equilin, a principal component of the estrogen replacement therapy pre-marin, increases the growth of cortical neurons via an NMDA receptor-dependent mechanism. Exp Neurol 147: 211-220 Brinton RD, Tran J, Proffitt P, Montoya M (1997b) 17beta-estradiol enhances the outgrowth and survival of neocortical neurons is culture. Neurochem Res 22: 1339-1351 Campard PK, Crochemore C, Rerie F, Monnier D, Kock B, Loefiier JP (1997) PACAP type I receptor activation promotes cerebellar neuron survival through the CAMP/PKA s’tgnaling pathway. DNA Cell Biol 16: 323-333 Chen J, Adachi N, Liu K, Arai T (1998) The effects of 17beta-estradiol on ischemiainduced neuronal damage in the gerbil hippocampus. Neuroscience 87: 817-822 Choi DW (1992) Excitotoxic cell death. J Neurobiol 23: 1261-76 Choi DW (1996) Ischemia-induced neuronal apoptosis. Curr Opin Neurobiol 6: 667-672 Choi DW, Monyer H, Giffard RG, Goldberg MP, Christiane CW (1990) Acute brain injury, NMDA receptors, and hydrogen ions: observations in cortical cell cultures. Adv Exp Med Biol 268: 501-504 Coyle JT, Puttfarcken P (1993) Oxidative stress, glutamate, and neurodegenerative disorders. Science 262: 689-695 Culmsee C, Vedder H, Ravati A, Junker V, Otto D, Ahlemeyer B, Krieg J-C, Krieglstein J (1999) Neuroprotection by estrogens in a mouse model of focal cerebral ischemia and in cultured neurons: evidence for a receptor-independent antioxidative mechanism. J Cereb Blood Flow Metab 19: 1263-1269 Cutolo M, Sulli A, Seriolo B, Accardo S, Masi AT (1995) Estrogens, the immune response and autoimmunity. Clin Exp Rheumatol 13: 217-226 Cyr M, Calon F, Morissette M, Grandbois M, DiPaolo T, Callier S (2000a) Drugs with estrogen-like potency and brain activity: potential therapeutic application for the CNS. Cur Pharm Des 6: 1287-1312 Cyr M, Ghribi O, DiPaolo T (2000b) Regional and selective effects of oestradiol and progesterone on NMDA and AMPA receptors in the rat brain. J Neuroendocrinol 12: 445-452 Das A, Chaudhuri SK (1995) Effects of sex steroids on the concentrations of some brain neurotransmitters in male and female rats: some new observations. Indian J Physiol Pharmacol 39: 223-230 DiPaolo T (1994) Modulation of brain dopamine transmission by sex steroids. Rev Neurosci 5: 27-41 D’Mello SR, Galli C, Ciotti T, Calissano P (1993) Induction of apoptosis in cerebellar granule neurons by low potassium: inhibition of death by insulin-like growth factor I and cAMP. Proc Natl Acad Sci USA 90: 10,989-10,993 Dodel RC, Du Y, Bales KR, Gao F, Paul SM (1999) Sodium salicylate and 17betaestradiol attenuate nuclear transcription factor NF-κB translocation in cultured rat astro-glial cultures following exposure to amyloid A beta(1-40) and lipopolysaccharides. J Neurochem 73: 1453-1460 Dong L, Wang W, Wang F, Stoner M, Reed JC, Harigai M, Samudio L, Kladde MP, Vyhlidal C, Safe S (1999) Mechanisms of transcriptional activation of bcl-2 gene expression by 17beta-estradiol in breast cancer cells. J Biol Chem 274: 32099-32107

22

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology Dubal DB, Shughrue PJ, Wilson ME, Merchenthaler I, Wise PM (1999) Estradiol modulates bcl-2 in cerebral ischemia: a potential rote for estrogen receptors. J Neurosci 19: 6385-6393 Evans RM (1988) The steroid and thyroid hormone receptor superfamily. Science 240: 889-895 Fink G, Sumner BE, McQueen JK, Wilson H, Rosie R (1998) Sex steroid control of mood, mental state and memory. Clin Exp Pharmacol Physiol 25: 764-775 Fiorelli G, Gori F, Frediani U, Franceschelli F, Tanini A, Tosti-Guerra C, Benvenuti S, Gennari L, Becherini L, Brandi ML (1996) Membrane binding sites and nongenomic effects of estrogen in cultured human pre-osteoclastic cells. J Steroid Biochem Mol Biol 59: 233-240 Foy MR, Xu J, Xie X, Brinton RD, Thompson RF, Berger TW (1999) 17beta-estradiol enhances NMDA receptor-mediated EPSPs and long-term potentiation. J Neurophysiol 81: 925-929 Frölich L, Riederer P (1995) Free radical mechanisms in dementia of Alzheimer type and the potential for antioxidative treatment. Arzneimittelforschung. 45: 443-446 Garcia-Segura LM, Cardona-Gomez P, Naftolin F, Chowen JA (1998) Estradiol upregulates Bcl-2 expression in adult brain neurons. Neuroreport 9: 593-597 Gattaz WF, Behrens S, De Vry J, Häfner H (1992) Estradiol inhibits dopamine mediated behavior in rats – an animal model of sex-specific differences in schizophrenia. Fortschr Neurol Psychiatr 60: 8-16 Gibbs RB (1994a) Estrogen and nerve growth factor-related systems in brain: effects an basal forebrain cholinergic neurons and implications for learning and memory processes and aging. Ann NY Acad Sci 743: 165-196 Gibbs RB, Aggarwal P (1998) Estrogen and basal forebrain cholinergic neurons: implications for brain aging and Alzheimer’s disease-related cognitive decline. Horm Behav 34: 98-111 Gibbs RB, Pfaff DW (1992) Effects of estrogen and fimbria/fornix transection on p75NGFR and ChAT expression in the medial septum and diagonal band of Broca. Exp Neurol 116: 23-39 Gibbs RB, Wu D, Hersh LB, Pfaff DW (1994b) Effects of estrogen replacement on the relative levels of choline acetyltransferase, trkA, and nerve growth factor messenger RNAs in the basal forebrain and hippocampal formation of adult rats. Exp Neurol 129: 70-80 Gollapudi L, Oblinger MM (1999) Estrogen and NGF synergistically protect terminally diffentiated, ERalpha-transfected PC12 cells from apoptosis. J Neurosci Res 56: 471-481 Goodman Y, Bruce AJ, Cheng B, Mattson MP (1996) Estrogens attenuate and corticosterone exacerbates excitotoxicity, oxidative injury, and amyloid betapeptide toxicity in hippocampal neurons. J Neurochem 66: 1836-44 Green PS, Simpkins JW (2000) Neuroprotective effects of estrogens: potential mechanisms of action. Int J Devl Neuroscience 18: 347-358 Green PS, Bishop J, Simpkins JW (1997a) 17alpha-estradiol exerts neuroprotective effects on SK-N-SH cells. J Neurosci 17: 511-515

23

Helmut Vedder and Christian Behl Green PS, Gordon K, Simpkins JW (1997b) Phenolic A ring requirement for the neuroprotective effects of steroids. J Steroid Biochem Mol Biol 63: 229-235 Green PS, Gridley KE, Simpkins JW (1998) Nuclear estrogen receptor-independent neuroprotection by estratrienes: a novel interaction with glutathione. Neuroscience 84: 7-10 Gridley KE, Green PS, Simpkins JW (1997) Low concentrations of estradiol reduce beta-amyloid (25-35)-induced toxicity, lipid peroxidation and glucose utilization in human SK-N-SH neuroblastoma cells. Brain Res 778: 158-165 Grossman CJ (1984) Regulation of the immune system by sex steroids. Endocr Rev 5: 435-455 Gu G, Rojo AA, Zee MC, Jianhua Y, Simerly RB (1996a) Hormonal regulation of CREB phosphorylation in the anteroventral periventricular nucleus. J Neurosci 16: 3035-3044 Gu Q, Moss RL (1996b) 17β-estradiol potentiates kainate-induced currents via activation of the cAMP cascade. J Neurosci 16: 3620-3629 Gunaga KP, Menon KM (1973) Effect of catecholamines and ovarian hormones on cyclic AMP accumulation in rat hypothalamus. Biochem Biophys Res Commun 54: 440-448 Gustafsson JA (2000) An update on estrogen receptors. Semin Perinatol 24: 66-69 Gutteridge JM (1995) Lipid peroxidation and antioxidants as biomarkers of tissue damage. Clin Chem 41: 1819-1828 Gutteridge JM, Halliwell B (2000) Free radicals and antioxidants in the year 2000. A historical look to the future. Ann NY Acad Sci 899: 136-147 Häfner H, Behrens S, De Vry J, Gattaz WF (1991a) An animal model for the effects of estradiol on dopamine-mediated behavior: implications for sex differences in schizophrenia. Psychiatry Res 38: 125-134 Häfner H, Behrens S, De Vry J, Gattaz WF (1991b) Oestradiol anheances the vulnerability threshold for schizophrenia in women by an early effect on dopaminergic neurotransmission. Evidence from an epidemiological study and from animal experiments. Eur Arch Psychiatry Clin Neurosci 241: 65-68 Hagiwara M, Shimomura A, Yoshida K, Imaki J (1996) Gene expression and CREB phosphorylation induced by cAMP and Ca2+ in neuronal cells. Adv Pharmacol 36: 277-285 Hajimohammadreza I, Brammer M (1990) Brain membrane fluidity and lipid peroxidation in Alzheimer’s disease. Neurosci Lett 112: 333-337 Halbreich U, Rojansky N, Palter S, Tworek H, Hissin P, Wang K (1995) Estrogen augments serotonergic activity in postmenopausal women. Biol Psychiatry 37: 434-441 Halliwell B (1992) Reactive oxygen species and the central nervous system. J Neurochem 59: 1609-1623 Halliwell B, Gutteridge JM (1990) Role of free radicals and catytic metal ions in human disease: an overview. Methods Enzymol 186: 1-85 Hardingham GE, Bading H (1998) Nuclear calcium: a key regulator of gene expression. Biometals 11: 345-358 Hardy J, Allsop D (1991) Amyloid deposition as the central event in the etiology of Alzheimer’s disease. Trends Pharmacol Sci 12: 383-388

24

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology Harman D, Eddy DE, Noffsinger J (1976) Free radical theory of aging: inhibition of amyloidosis in mice by antioxidants; possible mechanism. J Am Geriatr Soc 24: 203-210 Hayashi T, Ishikawa T, Yamada K, Kuzuya M, Naito M, Hidaka H, Iguchi A (1994) Biphasic effect of estrogen on neuronal constitutive nitric oxide synthase via Ca(2+)-calmodulin dependent mechanism. Biochem Biophys Res Commun 203: 1013-1019 Honda K, Sawada H, Kihara T, Urushitani M, Nakamizo T, Akaike A, Shimohama S (2000) Phosphatidylinositol 3-kinase mediates neuroprotection by estrogen in cultured cortical neurons. J Neurosci Res 60: 321-327 Jansson L, Holmdahl R (1998) Estrogen-mediated immunosuppression in autoimmune diseases. Inflamm Res 47. 290-301 Joels M, Karst H (1995) Effects of estradiol and progesterone on voltage-gated calcium and potassium conductances in rat CA1 hippocampal neurons. J Neurosci 15: 4289-4297 Kawas C, Resnick S, Morrison A, Brookmeyer R, Corrada M, Zonderman A, Bacal C, Lingle DD, Metter E (1997) A prospective study of estrogen replacement therapy and the risk of developing Alzheimer’s disease: the Baltimore Longitudinal Study of Aging. Neurology 48: 1517-1521 Kelly MJ, Lagrange AH, Wagner EJ, Ronnekleiv OK (1999) Rapid effects of estrogen to modulate G protein-coupled receptors via activation of protein kinase A and protein kinase C pathways. Steroids 64: 64-75 Kew JN, Smith DW, Sofroniew MV (1996) Nerve growth factor withdrawal induces the apoptotic death of developing septal cholinergic in vitro: protection by cyclic AMP analogue and high potassium. Neuroscience 70: 329-339 Kim S, Liva SM, Dalal MA, Verity MA, Voskuhl RR (1999) Estriol ameliorates autoimmune demyelinating disease: implications for multiple sclerosis. Neurology 52: 1230-1238 Klinge CM (2000) Estrogen receptor interaction with co-acivators and co-repressors. Steroids 65: 227-251 Kobayashi Y, Shinozawa T (1997) Effect of dibutyryl cAMP and several reagents an apoptosis in PC12 cells induced by sialoglycopeptide from bovine brain. Brain Res 778: 309-317 Korenman SG (1969) Comparative binding affinity of estrogens and its relation to estrogenic potency. Steroids 13: 163-177 Kruman II, Pedersen WA, Springer JE, Mattson MP (1999) ALS-linked Cu/Zn-SOD mutation increases vulnerability of motor neurons to excitotoxicity by a mechanism involving increased oxidative stress and perturbed calcium homeostasis. Exp Neurol 160: 28-39 Lammers CH, D’Souza U, Qin ZH, Lee SH, Yajima S, Mouradian MM (1999) Regulation of striatal dopamine receptors by estrogen. Synapse 34: 222-227 Leist M, Nicotera P (1998) Calcium and neuronal death. Rev Physiol Biochem Pharmacol 132: 79-125 Lezoualc’h F, Behl C (1998) Transcription factor NF-kB: friend or foe of neurons? Mol Psychiatry 3: 15-20 Matsuno A, Takekoshi S, Sanno N, Utsunomiya H, Ohsugi Y, Saito N, Kanemitsu H, Tamura A, Nagashima T, Osamura RY, Watanabe K (1997) Modulation of

25

Helmut Vedder and Christian Behl protein kinases and microtubule-associated proteins and changes in ultrastructure in female rat pituitary cells: effects of estrogen and bromocriptine. J Histochem Cytochem 45: 805-813 Mattsson MP, Rydel RE, Lieberburg I, Smith-Swintoskky VL (1993) Altered calcium signaling and neuronal injury: stroke and Alzheimer’s disease as examples. Ann NY Acad Sci 28: 679-721 McEwen BS, Alves SE (1999) Estrogen actions in the central nervous system. Endocr Rev 20: 279-307 McGeer EG, McGeer PL (1999) Brain inflammation in Alzheimer disease and the therapeutic inplications. Curr Pharm Des 5: 821-836 McGeer PL, McGeer EG (1996) Anti-inflammatory drugs in the fight against Alzheimer’s disease. Ann NY Acad Sci 777: 213-220 McKay LI, Cidlowski JA (1998) Cross-talk between nuclear factor-κB and the steroid hormone receptors: mechanisms of mutual antagonism. Mol Endocrinol 12: 4556 Mermelstein PG, Becker JB, Surmeier DJ (1996) Estradiol reduces calcium currents in rat neostriatal neurons via a membrane receptor. J Neurosci 16: 595-604 Merry DE, Korsmeyer SJ (1997) Bcl-2 gene family in the nervous system. Annu Rev Neurosci 20: 245-267 Michel MC, Rother A, Hiemke C, Graf R (1987) Inhibition of synaptosomal highaffinity uptake of dopamine and serotonin by estrogen agonists and antagonists. Biochem Pharmacol 36: 3175-3180 Miller CP, Jirkovsky I, Hayhurst DA, Adelman SJ (1996) In vitro antioxidant effects of estrogens with a hindered 3-OH function on the copper-induced oxidation of low density lipoprotein. Steroids 61: 305-308 Miller L, Hunt JS (1996) Sex steroid hormones and macrophage function. Life Sci 59: 1-14 Minami T, Oomura Y, Nabekura J, Fukuda A (1990) 17beta-estradiol depolarization of hypothalamic neurons is mediated by cyclic AMP. Brain Res 519: 301-307 Moosmann B, Behl C (1999) The antioxidant neuroprotective effects of estrogens and phenolic compounds are from their estrogenic properties. Proc Natl Acad Sci USA 96: 8867-8872 Moosmann B, Behl C (2000) Cytoprotective antioxidant function of tyrosine and tryptophan residues in transmembrane proteins. Eur J Biochem 267: 5687-5692 Mor G, Nilsen J, Horvath T, Bechmann I, Brown S, Garcia-Segura LM, Naftolin F (1999) Estrogen and microglia: a regulatory system that affects the brain. J Neurobiol 40: 484-496 Morimoto BH, Koshland DE (1990) Induction and expression of long- and short-term neurosecretory potentiation in a neural cell line. Neuron 5: 875-880 Mudd LM, Torres J, Lopez TF, Montague J (1998) Effects of growth factors and estrogen on the development of septal cholinergic neurons from the rat. Brain Res Bull 45: 137-142 Mulnard RA, Cotman CW, Kawas C, van Dyck CH, Sano M, Doody R, Koss E, Pfeiffer E, Jin S, Gamst A, Grundman M, Thomas R, Thal LJ (2000) Estrogen replacement therapy for treatment of mild to moderate Alzheimer disease: a ran-

26

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology domized controlled trial. Alzheimer’s Disease Cooperative Study. JAMA 283: 1007-1015 Murphy DD, Segal M (1996) Regulation of dendritic spine density in cultured rat hippocampal neurons by steroid hormones. J Neurosci 16: 4059-4068 Nohl H (1993) Involvement of free radicals in ageing: a consequence or cause of senescence. Brit Med Bull 49: 653-667 Olanow CW (1993) A radical hypothesis for neurodegeneration. Trends Neurosci 16: 439-444 Pappolla MA, Omar RA, Kim KS, Robakis NK (1992) Immunhistochemical evidence of oxidative stress in Alzheimer’s disease. Am J Pathol 140: 621-628 Pardridge WM, Moeller TL, Mietus LJ, Oldendorf WH (1980) Blood-brain barrier transport and brain sequestration of steroid hormones. Am J Physiol 239: E96-E102 Perillo B, Sasso A, Abbondanza C, Palumbo G (2000) 17beta-estradiol inhibits apoptosis in MCF-7 cells, inducing bcl-2 expression via two estrogen-responsive elements present in the coding sequence. Mol Cell Biol 20: 2890-2901 Piotrowski J, Pietras T, Kurmanowska Z, Nowak D, Marczak J, Marks-Konczalik J, Mazerant P (1996) Effect of paraquat intoxication ond ambroxol treatment on hydrogen peroxide production and lipid peroxidation in selected organs of the rat. J Appl Toxicol 16: 501-507 Pozzo-Miller LD, Inoue T, Murphy DD (1999) Estradiol increases spine density and NMDA-dependent Ca2 ± transisents in spines of CA1 pyramidal neurons from hippocampal slices. J Neurophysiol 81: 1404-1411 Racay P, Kaplan P, Lehotsky J (1996) Control of Ca2+ homeostasis in neuronal cells. Gen Physiol Biophys 15: 193-210 Raevskii KS (1997) Brain dopamine receptors: structure, functional role, and modulation by psychotropic substances. Vopr Med Khim 43: 553-565 Razandi M, Pedram A, Greene GL, Levin ER (1999) Cell membrane and nuclear estrogen receptors (ERs) originate from a single transcript: studies of ERalpha and ERbeta expressed in Chinese hamster ovary cells. Mol Endocrinol 13: 307-319 Regan RF, Guo Y (1997) Estrogens attenuate neuronal injury due to hemoglobin, chemical hypoxia, and excitatory amino acids in murine cortical cultures. Brain Res 764: 1333-1340 Riecher-Rössler A, Häfner H (1993) Schizophrenia and oestrogens – is there an asssociation? Eur Arch Psych Clin Neurosci 242: 323-328 Riecher-Rössler A, Häfner H, Stummbaum M, Maurer K, Schmidt R (1994) Can estradiol modulate schizophrenic symptomatology? Schizophr Bull 20: 203-214 Ruiz-Larrea B, Leal A, Lacort M, de Groot H (1994) Antioxidant effects of estradiol and 2-hydroxyestradiol on iron-induced lipid peroxidation of rat liver microsomes. Steroids 59: 383-388 Russell SR (1996) Nuclear hormone receptors and the Drosophila ecdysone response. Biochem Soc Symp 62: 111-121 Saunders-Pullman R, Gordon-Elliott J, Parides M, Fahn S, Saunders HR, Bressman S (1999) The effect of estrogen replacement an early Parkinson’s disease. Neurology 52: 1417-1421

27

Helmut Vedder and Christian Behl Sawada H, Ibi M, Kihara T, Urushitani M, Akaike A, Shimohama S (1998) Estradiol protects mesencephalic dopaminergic neurons from oxidative stress-induced neuronal death. J Neurosci Res 54: 707-719 Schmidt R, Fazekas F, Reinhart B, Kapeller P, Fazekas G, Orfenbacher S, Eber B, Schumacher M, Freidl W (1996) Estrogen replacement therapy in older women: a neuropsychological and brain MRI study. J Am Geriatr Soc 44: 1307-1313 Schneider LS, Farlow MR, Henderson VW, Pogoda JM (1996) Effects of estrogen replacement therapy on response to tacrine in patients with Alzheimer’s disease. Neurology 46: 1580-1584 Sergeev PV, Vladimirov IA, Seifulla RD, Denisov IP, Rudnev IN (1974) The role of the chemical structure of steroid hormones in inhibiting peroxidative lipid oxidation in mitochondrial membranes. Vopr Med Khim 20: 359-362 Shibata H, Spencer TE, Onate SA, Jenster G, Tsai SY, Tsai MJ, O’Malley BW (1997) Role of co-activators and co-repressors in the mechanism of steroid/thyroid receptor action. Recent Prog Horm Res 52: 141-164 Shughrue PJ, Lane MV, Merchenthaler I (1997) Comparative distribution of estrogen receptor-alpha and -beta mRNA in the rat central nervous system. J Comp Neurol 388: 507-525 Shyamal G, Guiot MC (1992) Activation of κB-specific proteins by estradiol. Proc Natl Acad Sci USA 89: 10628-10632 Sies H (1997) Oxidative stress: oxidants and antioxidants. Exp Physiol 82: 291-295 Singer CA, Rogers KL, Dorsa DM (1998) Modulation of Bcl-2 expression: a potential component of estrogen protection in NT2 neurons. Neuroreport 9: 2565-2568 Singer CA, Rogers KL, Strickland TM, Dorsa DM (1996) Estrogen protects primary cortical neurons from glutamate toxicity. Neurosci Lett 212: 13-16 Singh M, Meyer EM, Simpkins JW (1995) The effect of ovariectomy and estradiol replacement on brain-derived neurotrophic factor mRNA expression in cortical an hippocampal brain regions of female Sprague Dawley rats. Endocrinology 136: 2320-2324 Singh M, Setalo G Jr, Guan X, Warren M, Toran-Allerand CD (1999) Estrogen-induced activation of mitogen-activated protein kinase in cerebral cortical explants: convergence of estrogen and neurotrophin signaling pathways. J Neurosci 19: 1179-1188 Smith MA, Perry G, Richey PL, Sayre LM, Anderson VE, Beal MF, Kowall N (1996) Oxidative damage in Alzheimer’s. Nature 382: 120-121 Sohrabji F, Greene LA, Miranda RC, Toran-Allerand CD (1994) Reciprocal regulation of estrogen and NGF receptors by their ligands in PC12 cells. J Neurobiol 25: 974-988 Sohrabji F, Miranda R, Torran-Allerand CD (1995) Identification of a putative estrogen response element in the gene encoding brain-derived neurotrophic factor. Proc Natl Acad Sci USA 92: 11110-11114 Sola C, Barron S, Tusell JM, Serratosa J (1999) The Ca2+/calmodulin signaling system in the neural response to excitability. Involvement of neuronal and glial cells. Prog Neurobiol 58: 207-232 Stumpf WE, Jennes L (1984) The A-B-C (Allocortex-Brainstem-Core) circuitry of endocrine-autonomic integration and regulation: a proposed hypothesis on the

28

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology anatomical-functional relationships between estradiol sites of action and peptidergic-aminergic neuronal systems. Peptides 5 (Suppl 1): 221-226 Sugioka K, Shimosegawa Y, Nakano M (1987) Estrogens as natural antioxidants of membrane phospholipid peroxidation. FEBS Lett 210: 37-39 Toran-Allerand CD, Singh M, Setalo G Jr (1999) Novel mechanisms of estrogen action in the brain: new players in an old story. Front Neuroendocrinol 20: 97-121 Toung TJK, Traystman RJ, Hurn PD (1998) Estrogen-mediated neuroprotection after experimental stroke in male rats. Smoke 29: 1666-1670 Tymianski M, Tator CH (1996) Normal and abnormal calcim homostasis in neurons: a basis for the pathophysiology of traumatic and ischemic central nervous system injury. Neurosurgery 38: 1176-1195 Vedder H, Anthes N, Stumm G, Würz C, Behl C, Krieg J-C (1999) Estrogen hormones reduce lipid peroxidation in cells and tissues of the central nervous system. J Neurochem 72: 2531-2538 Vedder H, Teepker M, Fischer S, Krieg J-C (2000) Chraracterization of the neuroprotective effects of estrogens on hydrogen peroxide-induced cell death in hippocampal HT22 cells: time and dose-dependency. Exp Clin Endocrinol Diabetes 108: 120-127 Vedder H, Weiß I, Holsboer F, Reul JMHM (1993) Glucocorticoid and mineralocorticoid receptors in rat neocortical and hippocampal brain cells in culture: characterization and regulatory studies. Brain Res 605: 18-24 Vossler MR, Yao H, York RD, Pan MG, Rim CS, Stork PJ (1997) cAMP activates MAP kinase and Elk-1 through a B-Raf- and Rap1-dependent pathway. Cell 89: 73-82 Walton M, Sirimanne E, Williams C, Gluckman P, Dragunow M (1996) The role of the cyclic AMP-responsive element binding Protein (CREB) in hypoxic-ischemic brain damalte und repair. Mol Brain Res 43: 21-29 Watters JJ, Dorsa DM (1998) Transcriptional effects of estrogen on neuronal neurotensin gene expression involve cAMP/protein kinase A-dependent signaling mechanisms. J Neurosci 18: 6672-6680 Watters JJ, Campbell JS, Cunnfngham MJ, Krebs EG, Dorsa DM (1997) Rapid membrane effects of steroids in neuroblastoma cells: effects of estrogen a mitogen actived Protein kinase signalling cascade und c-fos immediate early gene transcription. Endocrinology 138: 4030-4033 Weaver CE Jr, Park-Chung M, Gibbs TT, Farb DH (1997) 17beta-estradiol protects against NMDA-induced excitotoxicity by direct inhibition of NMDA receptors. Brain Res 761: 338-341 Weiland NG (1992) Glutamic acid decarboxylase messenger ribonucleic acid is regulated by estradiol and progesterone in the hippocampus. Endocrinology 131: 2697-2702 Weiss DJ, Gurpide E (1988) Non-genomic effects of estrogens and antiestrogens. J Steroid Biochem 31: 671-676 Wiese TE, Polin LA, Palomino E, Brooks SC (1997) Induction of the estrogen specific mitogenic response of MCF-7 cells by selected analogues of estradiol-17beta: a 3D QSAR study. J Med Chem 40: 3659-3669

29

Estrogens in Neuropsychiatric Disorders: From Physiology to Pathophysiology Wolozin B, Behl C (2000) Mechanisms of neurodegenerative disorders. Part 2: Control of cell death. Arch Neurol 57: 801-804 Wong M, Moss RL (1992) Long-term and short-term electrophysiological effects of estrogen on the synaptic properties of hippocampal CA1 neurons. J Neurosci 12: 3217-3225 Woolley CS, McEwen BS (1994) Estradiol regulates hippocampal dendritic spine density via an NMDA receptor dependent mechanism. J Neurosci 19: 379-386 Woolley CS, Weiland NG, McEwen BS, Schwartzkroin PA (1997) Estradiol increases the sensitivity of hippocampal CAl pyramidal cells to NMDA receptor-mediated synaptic input: correlation with dendritic spine density. J Neurosci 17: 1848-1859 Yamashita S (1998) Localization and functions of steroid hormone receptors. Histol Histopathol 13: 255-70 Yang S-H, Shi J, Day AL, Simpkins JW (2000) Estradiol exerts neuroprotective effects when administered after ischemic insult. Stroke 31: 75-79 Yankner BA (1996) Mechanism of neuronal degeneration in Alzheimer’s disease. Neuron 16: 921-932

30

2 Estrogens and Schizophrenia Anita Riecher-Rössler

Introduction There is increasing evidence from both clinical and epidemiological as well as basic research that estrogens exert protective effects in schizophrenia. A brief overview of these protective effects will be provided in this chapter and potential therapeutic implications will be discussed. If these effects are confirmed, it could have important consequences for prophylaxis and therapy. For instance, consideration would need to be given to estrogen replacement therapy in periand postmenopausal women with schizophrenia, adjunct estrogen therapy in women with estrogen deficiency syndromes, cycle-modulated neuroleptic therapy in women with frequent perimenstrual relapses, and/or emphasis on prolactin-sparing atypical neuroleptics in women with hypoestrogenism. Further research is urgently needed since direct therapeutic benefits for women might result.

Estrogens and Schizophrenia: Historical Evidence The interest for the association between female sex hormones and mental diseases is not new. Particularly concerning psychosis, already at the beginning of the twentieth century psychiatrists considered a possible association of schizophrenia with estrogens: Early clinicians, such as Kraepelin (1909) or Kretschmer (1921), reported that many schizophrenic women show physical signs and other anatomical abnormalities indicating “insufficient functioning of the sexual glands” with “hypoestrogenism”. In the 1930s, studies analyzing estrogen levels in the blood and urine confirmed these observations. Researchers found decreased blood levels of estrogens in most of the schizophrenic inpatients they examined (for reviews see Diczfalusy and Lauritzen 1961; Riecher-Rössler and Häfner

31

Anita Riecher-Rössler

1993). As at this time neuroleptic therapy had not been introduced, the observed abnormalities cannot be interpreted as side effects of drug treatment. In later decades, studies showed that the gonadotrophins folliclestimulating hormone (FSH) and luteinizing hormone (LH) are low in women with schizophrenia as compared to controls. Many authors also reported irregular cycles in these women (for review see Riecher-Rössler and Häfner 1993). All these disturbances can possibly be attributed to a gonadal dysfunction with insufficient estrogen production from the ovaries. Furthermore, there have always been observations indicating a possible influence of estrogens on psychotic symptomatology. Krafft-Ebing (1896) was among the first to describe how some women became psychotic mainly before or during menstruation, i.e., when estrogen blood levels are relatively low. Kraepelin (1909) even created a separate diagnostic category, “menstrual psychosis”. Kretschmer (1921) reported on patients in whom the outbreak of schizophrenia was in temporal relation with “surgery of ovaries, pregnancy, delivery and puerperium.” Finally, Manfred Bleuler (1943) noted that late-onset schizophrenia with onset after age 40 years was much more frequent in women than in men, which he associated with the “loss of ovarian function” starting at around the same age. Further evidence comes from early intervention studies. As early as in the 1940s, Manfred Bleuer (1943) reported the first unsystematic trials using a combination of ovarian and pituitary hormones. Mall (1959, 1960), a German psychiatrist in charge of a large hospital, examined 167 women suffering from schizophrenia with respect to estrogen excretion in a 24-h urine sample, basal temperature, and vaginal cytology. Based on his findings, he divided the psychoses into two groups: hypofollicular and hyperfollicular. In the former group, he replaced estrogens and found that “hypofollicular psychosis can be healed relatively easily by this substitution therapy”. Unfortunately, Mall does not give many details about these interesting studies. Two hypotheses can be derived from these early observations (Riecher-Rössler and Häfner 1993): (1) The hypothesis of hypoestrogenism Some schizophrenic women suffer from chronic gonadal hypofunction and hypoestrogenism.

32

Estrogens and Schizophrenia

(2) The estrogen protection hypothesis Estrogens have a protective effect in schizophrenia. Recently, research has shown a renewed interest in this topic. In fact, there is a growing body of evidence in the meantime for both hypotheses. However, there is still need for more research. Especially as regards the causal explanation of the first finding – the gonadal hypofunction and hypoestrogenism in schizophrenic women –, data are still very limited.

The Estrogen Protection Hypothesis Research in the Past Few Decades Research in the last decades has confirmed many of the historical findings concerning a protective effect of estrogens in schizophrenia. Clinically, psychotic symptomatology has been shown to be exacerbated pre- or perimenstrually, i.e., in the low estrogen phase of the cycle (Endo et al. 1978; Glick and Stewart 1980; Gerada and Reveley 1988; Riecher-Rössler et al. 1994a, b). During pregnancy, when estradiol levels are about 200-fold higher than normal, chronic psychoses seem to improve (Chang and Renshaw 1986), but there is a 20-fold excess of psychosis after delivery (Kendell et al. 1987), when estrogen levels suddenly drop to normal. Psychoses associated with other forms of estrogen withdrawal have also been described (Mahé and Dumaine 2001). Furthermore, Seeman (1983) noted that schizophrenic women in the fertile age group of 20–40 years, i.e., the time of the highest ovarian estradiol production, require lower doses of neuroleptics than older women or men of the same age group – even when controlled for body weight. The same group has recently shown that early puberty is associated with a late onset of schizophrenia (Cohen et al. 1999). That means physiological estrogens might delay the outbreak of the disease. In the early 1980s, it was also observed that the effect of estrogens in laboratory animals is in some respects similar to that of neuroleptics. Estrogens can, for example, enhance neuroleptic-induced catalepsy and reduce amphetamine- and apomorphine-induced behavioral changes such as stereotypies (Gordon et al. 1980; Hruska and Silbergeld 1980; Di Paolo et al. 1981; Nicoletti et al. 1983). It has also been shown that estrogens can modulate the sensitivity and

33

Anita Riecher-Rössler

number of dopamine receptors (Koller et al. 1980; Gordon and Diamond 1981; Bédard et al. 1984). The identification of estrogen receptors in the limbic system further supported the assumption that estrogens not only play a role in the modulation of endocrine functions, but also have a “neuromodulating function” (for review see Riecher-Rössler and Häfner 1993). Today we know that estrogens produce many other effects. They do not only improve cerebral blood flow and glucose metabolism (Rasgon et al. 2001), but also promote neuronal sprouting and are regarded as neuroprotective. Their modes of action are not only classical genomic ones, but also involve nongenomic, rapid interactions, which explains the different latency of effects. They modulate monoaminergic neurotransmission and appear to have specific and significant effects not only on dopamine, but also on serotonin and GABA (Stahl 2001a, b; De Battista et al. 1999; Garcia-Segura et al. 2001; Kuhl 2001) to such a degree that they have even been called “nature’s psychoprotectant” (Fink et al. 1996).

Own Studies In an epidemiological study on a representative sample of 392 firsttime admitted patients with schizophrenia, the ABC Study, we found that schizophrenic women have a later peak of illness onset in comparison with schizophrenic men (Riecher et al. 1990; Häfner et al. 1991a, b; Häfner see this volume). They also exhibit an additional, smaller peak after age 45. We postulated that estrogens raise the vulnerability threshold for the outbreak of the disease. According to this hypothesis, women would be protected against schizophrenia between puberty and menopause to some extent by their relatively high gonadal estrogen production during this time. Then, around age 45, several years before menopause sets in at a mean age of 51.4 years, estrogen production begins to fall (Labhart 1978). Thus, women would lose the protection estrogens give, which could account for their second peak of illness onset after age 45. Based on these findings and on the aforementioned results from basic research which pointed to antidopaminergic properties of estrogens, Häfner together with Gattaz examined the neurohormonal effects of estradiol, the main component of estrogens, by means of animal experiments. They found evidence that chronic estradiol

34

Estrogens and Schizophrenia

treatment reduces the sensitivity of dopamine-D2 receptors (Häfner et al. 1991c; Gattaz et al. 1992). We also tried to examine the influence of estrogens on schizophrenic symptomatology in a clinical study (Riecher-Rössler et al. 1994a, b). We chose the female menstrual cycle as a “natural experiment” and tested the hypothesis that schizophrenic symptomatology varies with estradiol blood levels throughout the menstrual cycle. We examined 32 acutely admitted female schizophrenics, who gave a history of regular menstrual cycles. We saw a significant excess of admissions during the perimenstrual low-estrogen phase of the cycle (p < .005). During the hospital stay of the 32 women, a significant association emerged between estradiol levels, on the one hand, and psychopathology scores, on the other: Psychopathology seemed to improve when estradiol levels rose and vice versa. This was not only true for the total score of the Brief Psychiatric Rating Scale (BPRS, Overall and Gorham 1962) and for almost all the subscores of this scale, including anergia, thought disturbance, activation, and hostile suspiciousness, but also for general behavior on the ward as rated by the nurses (NOSIE, Honigfeld et al. 1976) and for general well-being and paranoid feelings as rated by the patients themselves (BfS and paranoid subscore of PDS, both by von Zerssen and Koeller 1976). The only exceptions were the anxiety/depression score of the BPRS and self-rated depression (depression score of PDS), which did not show this association. We interpreted this finding as further evidence for a protective effect of estradiol in schizophrenia. Our findings, however, rose various questions, especially concerning their specificity for schizophrenia: Is the observed estradioldependent fluctuation of the symptoms specific for schizophrenia or can it also be found in other mental diseases? We therefore tested the specificity of our results for schizophrenia by examining a control group of 29 acutely admitted females with psychiatric disorders other than schizophrenia and comparing them to the 32 schizophrenic women described above (Riecher-Rössler et al. 1998). In the control group, consisting mainly of women with depressive disorders, symptoms were not correlated with estradiol blood levels (Table 1). Based on the observations made on depressive symptoms in the schizophrenic women, these results were hardly surprising. The only exception in the control group was general behavior on the ward as assessed by the nursing staff. It was obviously positively

35

Table 1. Correlations between psychopathology scores and estradiol serum levels throughout hospital stay of schizophrenic and nonschizophrenic womena Schizophrenics (n = 32)

a b

c

mean –0.25

(SD) (0.41)

p < 0.01

nb 29

mean 0.03

(SD) (0.55)

p n.s.

31 –0.10 31 –0.15 31 –0.28 31 –0.27 31 –0.19 32 0.25 32 –0.20 32 –0.17 32 –0.10 – not evaluated –

(0.52) (0.42) (0.44) (0.41) (0.47) (0.49) (0.43) (0.42) (0.52)

n.s. < 0.10 < 0.01 < 0.01 < 0.10 < 0.01 < 0.05 < 0.05 n.s.

29 28 15 29 22 29 29 27 29 29

0.00 –0.04 –0.13 –0.06 –0.13 0.22 –0.02 0.06 0.01 0.03

(0.52) (0.53) (0.55) (0.49) (0.45) (0.43) (0.56) (0.51) (0.56) (0.53)

n.s. n.s. n.s. n.s. n.s. < 0.01 n.s. n.s. n.s. n.s.

nb 31

Shown as means of individual correlation coefficient, i.e., cross-correlations (Jenkins and Watts 1968). The slightly varying numbers are partly due to missing data, partly due to exclusions: patients who never showed symptoms of a certain score or never showed variability in a certain score were excluded from analysis concerning the respective score (14 controls concerning “thought disturbance”, 6 controls concerning “hostile-suspiciousness” and 2 controls concerning “PDS-paranoid subscore”). Contrary to the other scores, in the total NOSIE score a higher value means less psychopathology.

Source: Riecher-Rössler A, Häfner H, Dütsch-Strobel A, Stumbaum M (1998) Gonadal function and its influence on psychopathology – A comparison of schizophrenic and non-schizophrenic female inpatients. Arch Women Ment Health 1: 15-26

Anita Riecher-Rössler

36

Parameters BPRS-total score BPRS subscores Anxiety/depression Anergia Thought disturbance Activation Hostile–suspiciousness NOSIE-total scorec BfS-total score PDS-paranoid subscore PDS-depression subscore HAMD-total score

Non-schizophrenics (n = 29)

Estrogens and Schizophrenia

associated with estradiol levels in the control group as well. If only the control patients with major depressive disorders were taken into account, the results remained essentially unchanged. Thus, the clinical effect of estrogens seems to resemble that of neuroleptics: They appear to ameliorate psychotic symptoms and general behavior, but not depressive symptoms as much (RiecherRössler et al. 1998). We have also looked in more detail into our epidemiological ABC sample – with a special focus on late onset cases and the second peak of onsets after age 45 in women. This second peak reflects what has been described by many psychiatrists for a long time, namely, the fact that the incidence of late onset schizophrenia is about twice as high in women as in men (Bleuler 1943). We could almost exactly confirm this finding. First admission for schizophrenia after age 40 occurs in only 10% of all schizophrenic men, but in about 21% of all schizophrenic women. Incidence in women over 40 is 8.9 per 100,000, whereas it is only 4.2 per 100,000 in men (Riecher-Rössler et al. 1997). Furthermore, we had a very interesting new finding as regards the symptoms and disease course of these late-onset women. As for symptoms, we saw hardly any differences between late- and earlyonset patients when we compared these age groups without separation of the two sexes. Late-onset patients just showed slightly fewer nonspecific symptoms than early-onset patients. However, if we looked at the two genders separately, it was only the late onset men who had distinctly milder symptoms – not the late-onset women! Late-onset women obviously suffer from distinctly more severe symptoms than late-onset men – their symptoms are as severe as those of early onset women and men (Riecher-Rössler et al. 1997). The results concerning the course of the disease (based on Danish case register data) were very similar: Late-onset and also very late onset patients generally had a distinctly better institutional course than early-onset patients. For example, they spent significantly fewer days in the hospital than early-onset patients. Again, this was mainly due to the over 40-year-old men, who spent the shortest time in hospital, whereas women over 40 spent almost as much time in the hospital as the early-onset women (Riecher-Rössler et al. 1997). Thus, we have found that the over-40-year-old men have distinctly milder symptoms and a better institutional course than early-onset patients, whereas the over 40-year-old women not only

37

Table 2. Psychopathology scores with significant differences between early- and late-onset schizophrenia cases (ICD-9: 295) by gender (ABC Study). (SD)

p(t)

Womenb M

(SD)

p(t)

TOT < 40 years ≥ 40 years

42.5 24.6

(16.4) (14.8)

**

41.9 39.5

(16.1) (14.5)

NSN < 40 years ≥ 40 years

16.1 8.3

(7.3) (5.4)

**

15.1 12.9

DAH < 40 years ≥ 40 years

10.6 5.3

(7.3) (3.3)

**

Psychotic symptoms < 40 years ≥ 40 years

10.7 7.1

(5.5) (3.9)

First-rank symptoms < 40 years ≥ 40 years

2.2 0.7

(2.0) (1.3)

n.s. not significant; ° p < 0.1; ** p < 0.01 n < 40 years: 106; n ≥ 40 years: 7 b n < 40 years: 99; n ≥ 40 years: 22 c n < 40 years: 205; n ≥ 40 years: 29 a

Totalc M

(SD)

p(t)

n.s.

42.2 35.9

(16.2) (15.7)

°

(6.1) (6.1)

n.s.

15.6 11.8

(6.8) (6.2)

**

11.4 11.7

(7.7) (7.8)

n.s.

11.0 10.1

(7.5) (7.5)

n.s.

°

11.3 11.1

(4.8) (4.8)

n.s.

11.0 10.2

(5.2) (4.8)

n.s.

°

2.3 2.0

(1.9) (1.8)

n.s.

2.2 1.7

(2.0) (1.8)

n.s.

TOT total score of PSE (Wing et al. 1973; 1974) NSN non-specific neurotic syndromes, subscore of PSE DAH delusions and hallucinations, subscore of PSE

Source: Riecher-Rössler A, Löffler W, Munk-Jorgensen P (1997) What do we really know about late-onset schizophrenia? Eur Arch Psychiatry Clin Neurosci 247: 195-208

Anita Riecher-Rössler

38

Mena M

Estrogens and Schizophrenia

fall ill twice as often as elderly men, they obviously also suffer from a disease almost as severe as young patients. One explanation for this could again be the estrogen effect: if illness onset of women with a relatively high underlying vulnerability is delayed by estrogens, this high vulnerability is “unmasked” by the loss of this protective factor around menopause. These women then not only fall ill more frequently, but the disease is more severe as regards symptoms and course.

The Hypoestrogenism Hypothesis We and several other authors in the meantime also found a disturbed gonadal function and/or hypoestrogenism in schizophrenic women (Riecher-Rössler et al. 1994a, 1998; Kulkarni et al. 1996; Bergemann et al. 1996; 2002, see also this volume; Choi et al. 2001; Hoff et al. 2001; Huber 2001; Huber et al. 2001; Canuso et al. 2002; Smith et al. 2002; Zhang and Seeman 2002; for review see Riecher-Rössler 2003). Findings are menstrual irregularities and pathologically low estradiol and progesterone blood levels throughout the menstrual cycle as well as anovulation in the majority of women with schizophrenia. Reduced fertility rates have also been described in the past and recently confirmed (Hutchinson et al. 1999). In our own study on 32 acutely admitted schizophrenic women (see above), we had interesting results regarding this hypothesis. To start with, many schizophrenic women failed to qualify for our study because of considerable menstrual irregularities. However, even the 32 women who gave a history of regular menstrual cycles and were therefore included in our study showed many hints of a severely disturbed gonadal function (Riecher-Rössler et al. 1998). As compared to the 29 controls, they not only had greater variation in their cycle length observed on ward, but they also had significantly lower estradiol and progesterone blood levels throughout their menstrual cycle. Of these women, 56% presumably suffered from anovulation, which is a much higher proportion than in the control patients with depressive disorders, where anovulation had to be suspected in only 19% of those assessed (Riecher-Rössler et al. 1998). Furthermore, these abnormalities were obviously not (solely) due to neuroleptics. Thus, we could not, for example, find any associa-

39

Table 3. Menstrual and hormonal status of the schizophrenic group and the control group with depression Schizophrenic patients (n = 32a/n = 27b) Mean Range

Controls (n = 29) Mean

Range

(n = 32)a (n = 27)b (n = 32)a (n = 27)b

27.1 27.1 28.4 29.2

(23.0-31.5) (23.0-31.5) (11.0-66.0) (11.0-66.0)

27.6

(16.5-35.0)

28.4

(20.0-41.0)

Hormonal status Mean values throughout observation: Estradiol (n = 32)a [pmol/l] (n = 27)b Progesterone (n = 32)a [nmol/l) (n = 27)b

158.6 172.5 3.1 3.5

(40.0-824.0) (45.0-824.0) (0.1-37.2) (0.1-37.2)

282.3

(15-1258)

6.5

(0.1-63.3)

43.1 42.0 1.8 2.0 176.5 196.7 4.3 4.9

(4.8-99.9) (4.8-99.9) (0.8-4.8) (0.8-4.8) (45.0-502.0) (45.0-502.0) (0.1-37.2) (0.1-37.2)

9.9

(1.9-40.8)

1.6

(0.6-3.1)

Length of menstrual cycle (days) Before admission (As stated by patient) During hospital stay (As observed)

a b

(n = 32)a (n = 27)b (n = 32)a (n = 27)b (n = 32)a (n = 27)b (n = 32)a (n = 27)b

271.7

(63-1165)

2.8

(0.1-26.2)

n.s. n.s. n.s. n.s.

< 0.001 < 0.001 < 0.01 < 0.05 < 0.001 < 0.001 n.s. < 0.1 < 0.1 n.s. n.s. n.s.

All schizophrenic women investigated (n = 32) Only the schizophrenic women which had not taken oral contraceptives in the months preceeding admission (n = 27)

Source: Riecher-Rössler A, Häfner H, Dütsch-Strobel A, Stumbaum M (1998) Gonadal function and its influence on psychopathology – A comparison of schizophrenic and non-schizophrenic female inpatients. Arch Women Ment Health 1: 15–26

Anita Riecher-Rössler

40

On admission: Prolactin [ng/ml] Testosterone [nmol/l] Estradiol [pmol/l] Progesterone [nmol/l]

p(t)

Estrogens and Schizophrenia

tion between neuroleptic dosage (in chlorpromazine equivalents) and estradiol serum levels or ovulation (Riecher-Rössler et al. 1998). For theoretical reasons and also historically, the most interesting question was whether gonadal dysfunction with estrogen deficiency was just a state or rather a trait marker. We could not answer this from our clinical study sample as these women had been selected: Only women with a history of regular menstrual cycles had been included. Regarding this question we, therefore, investigated the women aged 18–45 of our representative ABC sample of first-time admitted schizophrenia patients. Forty-four women took part and were compared to 33 age-matched healthy women regarding potential indicators of preexisting chronic gonadal dysfunction with estrogen deficiency such as age at menarche, irregular cycles, midcycle bleeding, infertility, abortions, signs of relative hypoestrogenism or hyperandrogenism (Schepp 1997). The results were most interesting: Women with schizophrenia could be distinguished from the controls by several factors: Their menarche had been later and they had suffered more often from loss of hair, midcycle bleeding, mild bleeding, and hirsutism. There was also a tendency to more infertility in the patients.

Discussion Although closely connected, the two hypotheses shall be discussed separately. Firstly concerning the hypoestrogenism hypothesis we must discuss the causes for the menstrual irregularities and hormonal disturbances actually observed. Several explanations are feasible: The disturbances could – at least partly – be a secondary phenomenon due to general mental “stress,” which patients experience during an acute psychiatric episode. It is well known that stress can induce hyperprolactinemia, and hyperprolactinemia can suppress gonadal function (Scriba and von Werder 1976; Maguire 2002). However, if this was the main cause of the disturbances, they should be seen not only in schizophrenia but in other acute psychiatric conditions as well. Thus, the fact that the irregularities were mainly seen in schizophrenic women and only very rarely in the control group both in our study and in other studies (e.g., Huber 2001; Huber et al. 2001) speaks more against this explanation – at least against it being the main cause.

41

Anita Riecher-Rössler

Another explanation to be considered is the influence of neuroleptics. Although this explanation is often given, a critical review of the literature showed that it is not at all clear whether the disturbances observed are solely a side effect of neuroleptic treatment (Riecher-Rössler et al. 1998; Riecher-Rössler 2003). Thus, there are several studies showing no association between dosage of typical neuroleptics on the one hand and either estradiol serum levels or irregularity of menstrual cycle on the other hand (Carter et al. 1982; Riecher-Rössler et al. 1998; Canuso et al. 2002). Furthermore, hypooestrogenism and menstrual irregularities were also present in women on atypical prolactin-sparing neuroleptics (Bergemann et al. 2002; Zhang and Seeman 2002). Finally, gonadal hypofunction and “hypoestrogenism” in schizophrenic women had already been observed and reported a long time before the introduction of neuroleptics (Kretschmer 1921; Ripley and Papanicolaou 1942; Bleuler 1943). Thus, even if we suspect that neuroleptics are partly responsible for the hypoestrogenism observed, we must be careful not to overlook (additional) endocrine disturbances associated with the disease itself. Therefore, we have finally to consider the possibility that the menstrual and/or hormonal abnormalities observed in our schizophrenic patients are – at least partly – directly associated with the disease itself. In principle, such an association is conceivable in two ways: Either the pathological processes of schizophrenia can lead to the observed abnormalities or – the other way around – gonadal dysfunction and estrogen deficiency of other origin can enhance the vulnerability for schizophrenia or trigger the outbreak of the disease. Both possibilities are theoretically feasible either acutely (as a state) or chronically (as a trait). In our studies we could find hints of both acute and of chronic gonadal hypofunction with physical signs of hypoestrogenism in schizophrenia. Regarding chronic hypoestrogenism, we found long-standing abnormalities in our patients as described by Kretschmer and other psychiatrists at the beginning of the last century. Although we could only examine a small group of 44 first-episode patients, they showed significant differences to the healthy control women as regards various indicators of a relative estrogen deficiency, such as delayed menarche, mid-cycle bleeding, weak bleeding, loss of hair, and hirsutism. The finding of a later menarche in schizophrenic women as compared to healthy controls is well in line with the finding of

42

Estrogens and Schizophrenia

Cohen et al. (1999) concerning the association of an earlier onset of schizophrenia in women with later puberty. A later physiological rise in serum estrogen levels might be associated with a higher risk of developing psychosis and an earlier onset. These results could also be explained by the hypoestrogenism hypothesis: There might be a subgroup of schizophrenic women with a chronic and even preexisting estrogen-deficiency state. This chronic state might have enhanced their vulnerability for the psychosis. We could also show that during acute schizophrenia most women suffer from marked menstrual irregularities and decreased estradiol and progesterone levels and many even from anovulation. Just like in chronic hypoestrogenism, we cannot identify the direction, let alone the cause of the association between this acute estrogen deficiency state and the psychotic episodes. However, as there is strong evidence that estrogens influence the dopaminergic system, it is intriguing to speculate that some schizophrenic episodes of women predisposed to schizophrenia are induced by acute gonadal dysfunction with decreased estrogen production. The causes for the latter might be manifold, ranging from premenstrual to postnatal and (pre-)menopausal drop in estrogen levels and even to surgical loss of ovaries or pharmacological suppression of physiological estrogen production. The resulting decrease in estradiol serum levels would, in this case, mean the loss of a protective factor and could thereby – in women predisposed to schizophrenia – lead to the outbreak of schizophrenic symptoms. As regards the estrogen protection hypothesis, well in line with our findings, Hallonquist et al. (1993) observed lower psychopathology scores in schizophrenic women during the mid-luteal phase than in the early follicular phase of the menstrual cycle. They concluded that estrogens may act as “endogenous neuroleptics”. Together with Gattaz et al. (1994), in a further study we found a better therapeutic response in schizophrenic inpatients admitted during their perimenstruum, i.e., the low-estrogen phase of their cycle. These patients required lower doses of neuroleptics and a shorter treatment to reach the same degree of remission as the patients admitted during their intermenstruum, i.e., their high-estrogen phase. We speculated that the better therapeutic response of the low-estrogen admission group may be related to the increasing levels of circulating estradiol just after admission. In the meantime several investigators have reported promising in-

43

Anita Riecher-Rössler

itial results using estrogens as a therapeutic agent. Thus, Kulkarni et al. (1996, 2001) found that schizophrenic women receiving estradiol as an adjunct to neuroleptic treatment showed more rapid improvement in psychotic symptoms compared with the group receiving neuroleptics only. Similar effects were reported by Lindamer et al. (1997) in a case report on a postmenopausal women. Recently, Lindamer et al. (2001) reported on a community-dwelling sample of women with schizophrenia. Twenty-four women received hormone replacement therapy (HRT), and 28 women had never received such therapy. Interestingly, the users of HRT required a lower average dose of antipsychotic medication and suffered from less severe negative symptoms. Jeste et al. (2001) report on “encouraging” results using estrogen “augmentation” of antipsychotics in an ongoing study in postmenopausal women with schizophrenia. Ahokas et al. (2000) could show positive effects of estrogen substitution in two women with postpartum psychosis. The results of our studies further imply that the effects of estrogens are not restricted to schizophrenia and psychotic symptoms, but that they can also influence general behavior. This could reflect a generally stabilizing effect of estrogens in addition to their presumed antipsychotic properties. Such an effect might also play a role in otherwise healthy women, who suffer from estrogen deficiency around and after menopause. Estrogen substitution in these cases not only seems to ameliorate physical complaints, but also the women’s mental state. Furthermore, this unspecific effect might be relevant in the pathogenesis of premenstrual syndrome. Well in line with this are the results of another study which we conducted in a consecutive sample of 88 female patients admitted to an emergency service of a general hospital because of a suicide attempt (Riecher-Rössler 1994). We found a significant excess of suicide attempts in the perimenstrual low-estrogen phase of the cycle as compared to the high-estrogen phase (p < .005). Choi et al. (2001) found only affective and somatic symptoms to correlate with estradiol levels in women with chronic schizophrenia, not psychotic symptoms. However, as the authors themselves state, their 24 patients were not in an acute psychotic episode, but had been treated with antipsychotics for at least 6 months. The variation in psychotic symptoms, thus, was possibly not sufficiently high to find significant statistical correlations. Furthermore, there was no self-rating of paranoid symptoms in this study.

44

Estrogens and Schizophrenia

Also regarding the potential effect of estrogens on depression there have been findings that are at odds with ours. Thus, Klaiber et al. (1979), and Holsboer (1983) reported on positive effects. Particularly in postpartum depression (Sichel et al. 1995; Gregoire et al. 1996; Ahokas et al. 1998) or in postmenopausal depression (Zweifel and O’Brien 1997; Schmidt et al. 2000; de Noaves Soares et al. 2001), positive effects of estrogens could be shown. Postpartum and postmenopausal disorders, however, might be partly due to the sudden drop or even a sustained lack of estrogens. Furthermore, the effects may depend on the severity of the depression and the estrogen state of the women, i.e., estrogens might only be helpful if there is a deficiency.

Implications for Therapy and Prophylaxis Further research on the potentially protective effect of estrogens in psychoses is urgently needed, as if this effect could be confirmed, it would have interesting consequences for prophylaxis and therapy of these disorders. First intervention trials in schizophrenia indicate that estradiol should be used as an augmentation compound, an adjunct to neuroleptic medication. Replications of these results in larger controlled studies are needed though before recommendations for broad clinical application can be made. Hormonal replacement with estrogens might be even more promising for women with schizophrenia during and after perimenopause, as estrogens in other disorders such as depression have proven to be especially helpful when there a hormone deficiency is substituted. Estrogen replacement therapy definitely ameliorates perimenopausal complaints, which can act as stressors and theoretically provoke relapses. HRT has also been recommended for other reasons, e.g., for prophylaxis of osteoporosis and possibly also Alzheimer’s disease. Schizophrenia might represent an additional indication to be studied. As there is growing evidence that many even younger women are in an estrogen-deficiency state, estrogens and the gonadal axis should, in the future, be more seriously considered in the treatment of women with schizophrenia. Psychiatric history taking should always include questions regarding menstrual irregularities, amenorrhea, galactorrhoea, etc. Prolactin and estrogen serum levels should be tested, if necessary.

45

Anita Riecher-Rössler

Even in menstruating women gonadal dysfunction and hypoestrogenic states can often be found (Riecher-Rössler et al. 1994a, 1998; Smith et al. 2002). In addition, hyperprolactinaemia is obviously underdiagnosed (Maguire 2002). Some authors therefore suggest routine laboratory tests (e.g., Smith et al. 2002). Most neuroleptics can cause hyperprolactinaemia and can – especially if taken over years – theoretically induce “iatrogenic early menopause” via suppression of physiological estradiol production. The risks of that are not only short-term effects such as hot flushes and sexual dysfunction, but also long-term consequences such as osteoporosis and potentially also cardiovascular disease or cognitive deterioration (Maguire 2002; Oesterlund 2002). In schizophrenia patients these risks are further increased by additional risk factors such as smoking, poor diet, and reduced exercise (Smith et al. 2002). Furthermore, the menopause complaints might lead to compliance problems. In hyperprolactinemia with secondary estrogen deficiency, prolactin-sparing neuroleptics (e.g., clozapine, quetiapine, or olanzapine – Maguire 2002) should therefore be preferred. If a switch to these neuroleptics is not possible for clinical reasons or if hypoestrogenism persists in spite of switching, estrogens should be substituted. The question of contraception needs to be taken into account in these cases as, when switching to prolactin-sparing neuroleptics, the menstrual cycle often normalizes and fertility is regained, with a high risk of unplanned pregnancy (Neumann and Frasch 2001). In women suffering from frequent perimenstrual psychotic relapses, “cycle-modulated” neuroleptic therapy or – if contraception is required at the same time – continuous use of oral contraceptives without hormone-free intervals might represent strategies worth being researched (Braendle et al. 2001; Riecher-Rössler 2002). Research should also be conducted on the best mode of HRT for psychiatric patients. Progestogens are usually added to estrogens to prevent endometrial cancer, but can antagonize the positive effects of estrogens with respect to mental state (Braendle et al. 2001; Cyr et al. 2002). Furthermore, we have to consider other risks of hormone therapy, including breast cancer or cardiovascular disease for certain combinations (Barrett-Connor and Stuenkel 2001; Writing Group for Women’s Mental Health 2002). As an alternative to conventional HRT, compounds with more

46

Estrogens and Schizophrenia

specific and potent estrogenic activity in the brain as opposed to other tissues should be searched for (Halbreich 2002; Riecher-Rössler 2002). This would not only minimize the side effects of hormonal therapy, but may also allow new therapeutic strategies in men. Possible candidates are the selective estrogen receptor modulators (SERMS), whose agonist or antagonist properties depend on the target tissue. The effects of the so far existing SERMS on the brain, however, remain to be clarified. Raloxifene, e.g., seems to exert its main effects on the bone, although there are recent data suggesting that it also acts on different brain receptors (Cyr et al. 2002). The synthetic steroid tibolone also seems to cause less endometrial proliferation, but its effects on the central nervous system are still not clear, apart from the fact that it seems to have an androgenic effect and increase β-endorphin levels with improvement of mood and libido (Davis 2002). Further studies on the brain effects of SERMS and other estrogenic compounds (e.g., phytoestrogens, xenoestrogens, DHEA) are urgently needed. In summary, there are emerging hopes that estrogens as neuroand psychoprotective adjunctive therapy may complement the traditional drug therapies in schizophrenia in the future. However, it must be emphasized that most strategies are still being researched. Especially before using estrogens as an adjunct therapy in younger women without proven estrogen deficiency, the results of larger controlled studies are needed. Other strategies, however, should even now be part of standard clinical care (Grigoriadis and Seeman 2002). These include examination of the gonadal axis and to draw therapeutic consequences, if indicated. Regarding estrogen substitution and replacement therapy, it has to be stressed that the decision must always be made on the basis of an individual risk-benefit assessment (NAMS 2000; Writing Group for Women’s Mental Health 2002) and in close cooperation with a gynecologist. For future research, many questions arise, not only regarding new therapeutic strategies and compounds, but also regarding some so far not well explained disturbances of estrogens and the HPG axis in women with schizophrenia. Further research into this area could hopefully even contribute to better understanding the pathogenesis of this disease, at least in a subgroup of women.

47

Anita Riecher-Rössler

Acknowledgement Shorter, preliminary versions of this article were previously published in the “Archives of Women’s Mental Health” (2002; 5: 111–118) and in the “Nervenarzt” in 2003.

References Ahokas AJ, Turtiainen S, Aito M (1998) Sublingual oestrogen treatment of postnatal depression. Lancet 351: 109 Ahokas A, Aito M, Turtiainen S (2000) Association between oestradiol and puerperal psychosis. Acta Psychiatr Scand 101: 167-170 Barrett-Connor E, Stuenkel CA (2001) Hormone replacement therapy (HRT): risks and benefits. Int J Epidemiol 30: 423-426 Bédard P, Boucher R, Daigle M, Di Paolo T (1984) Similar effect of estradiol and haloperidol on experimental tardive dyskinesia in monkeys. Psychoendocrinology 9: 375-379 Bergemann N, Mundt Ch, Resch F, Parzer P, Nagl I, Eller A, Runnebaum B (1996) On the hypoestrogen hypothesis in female schizophrenia: Preliminary results. Schizophr Res 18: 158 Bergemann N, Parzer P, Nagl I, Salbach B, Runnebaum B, Mundt Ch, Resch F (2002) Acute psychiatric admission and menstrual cycle phase in women with schizophrenia. Arch Women Ment Health 5: 119-126 Bleuler M (1943) Die spätschizophrenen Krankheitsbilder. Neurol XV 9: 259-290 Braendle W, Breckwoldt M, Kuhl H, Riecher-Rössler A, Rohde A, Rüther E (2001) Sexualhormone und Psyche – Ergebnisse des 2. Interdisziplinären Frankfurter Gesprächs zur Kontrazeption. Frauenarzt 42: 154-160 Canuso CM, Goldstein JM, Wojcik J, Dawson R, Brandman D, Klibanski A, Schildkraut JJ, Green AI (2002) Antipsychotic medication, prolactin elevation, and ovarian function in women with schizophrenia and schizoaffective disorder. Psychiatry Res 111: 11-20 Carter DA, McGarrick GM, Norton KRW, Paykel ES, Prysor-Jones RA, Whitehead SA (1982) The effect of chronic neuroleptic treatment on gonadotrophin release. Psychoneuroendocrinology 7: 201-207 Chang SS, Renshaw DC (1986) Psychosis and pregnancy. Compr Ther 12: 36-41 Choi SH, Kang SB, Joe SH (2001) Changes in premenstrual symptoms in women with schizophrenia: a prospective study. Psychosom Med 6: 822-829 Cohen RZ, Seeman MV, Gotowiec A, Kopala L (1999) Earlier puberty as a predictor of later onset of schizophrenia in women. Am J Psychiatry 156: 1059-1064 Cyr M, Calon F, Morissette M, Di Paolo T (2002) Estrogenic modulation of brain activity: implications for schizophrenia and Parkinson’s disease. J Psychiatry Neurosci 27: 12-27 Davis SR (2002) The effects of tibolone on mood and libido. Menopause 9: 162-170 De Battista C, Lawrence Smith D, Schatzberg AF (1999) Modulation of monoamine neurotransmitters by estrogen: clinical implications. Rev Psychiatry 18: 137-166

48

Estrogens and Schizophrenia de Noaves Soares C, Almeida OP, Joffe H, Cohen LS (2001) Efficacy of estradiol for the treatment of depressive disorders in perimenopausal women. Arch Gen Psychiatry 58: 529-534 Diczfalusy E, Lauritzen C (1961) Psychiatrische und neurologische Erkrankungen. In: Diczfalusy E, Lauritzen C (eds) Östrogene beim Menschen. Springer, Berlin Heidelberg, pp 461-462 DiPaolo T, Payet P, Labrie F (1981) Effect of chronic estradiol and haloperidol treatment on striatal dopamine receptors. Eur J Pharmacol 73: 105-106 Endo M, Daiguji M, Asano Y, Yamashita I, Takahashi S (1978) Periodic psychosis recurring in association with menstrual cycle. J Clin Psychiatry 39: 456-466 Fink G, Sumner BEH (1996) Oestrogen and mental state. Nature 383: 306 Fink G, Sumner BEH, Rosie R, Grace O, Quinn JP (1996) Estrogen control of central neurotransmission: effect on mood, mental state, and memory. Cell Mol Neurobiol 16: 325-344 Garcia-Segura LM, Azcoitia I, Don Carlos LL (2001) Neuroprotection by estradiol. Progr Neurobiol 63: 29-60 Gattaz WF, Behrens S, De Vry J, Häfner H (1992) Östradiol hemmt Dopaminvermittelte Verhaltensweisen bei Ratten: Ein Tiermodell zur Untersuchung der geschlechtsspezifischen Unterschiede bei der Schizophrenie. Fortschr Neurol Psychiatrie 1: 1-44 Gattaz WF, Vogel P, Riecher-Rössler A, Soddu G (1994) Influence of the menstrual cycle phase on the therapeutic response in schizophrenia. Biol Psychiatry 36: 137-139 Gerada C, Reveley A (1988) Schizophreniform psychosis associated with the menstrual cycle. Br J Psychiatry 152: 700-702 Glick J, Stewart D (1980) A new drug treatment for premenstrual exacerbation of schizophrenia. Compr Psychiatry 21: 281-287 Gordon JH, Borison RL, Diamond BL (1980) Modulation of dopamine receptors sensitivity by estrogen. Biol Psychiatry 15: 389-396 Gordon JH, Diamond BI (1981) Antagonism of dopamine supersensitivity by estrogen: neurochemical studies in an animal model of tardive dyskinesia. Biol Psychiatry 16: 365-371 Gregoire AJP, Kumar R, Everitt B. Henderson AF, Studd JWW (1996) Transdermal oestrogen treatment of severe postnatal depression. Lancet 347: 930-933 Grigoriadis S, Seeman MV (2002) The role of estrogen in schizophrenia: implications for schizophrenia practice guidelines for women. Can J Psychiatry 47: 437-442 Häfner H, Riecher-Rössler A, Hambrecht M, Maurer K, Meissner S, Schmidtke A, Fätkenheuer B, Löffler W, an der Heiden W (1991a) Geschlechtsunterschiede bei schizophrenen Erkrankungen. Fortschr Neurol Psychiatrie 59: 343-360 Häfner H, Behrens S, De Vry J, Gattaz WF, Löffler W, Maurer K, Riecher-Rössler A (1991b) Warum erkranken Frauen später an Schizophrenie? Erhöhung der Vulnerabilitätsschwelle durch Östrogen. Nervenheilkunde 10: 154-163 Häfner H, Behrens S, De Vry J, Gattaz WF (1991c) An animal model for the effects of estradiol on dopamine-mediated behavior: implications for sex differences in schizophrenia. Psychiatry Res 38: 125-143

49

Anita Riecher-Rössler Halbreich U (2002) The spectrum of estrogens, estrogen agonists and SERMS – abstract. Int J Neuropsychopharmacol 5 (Suppl 1): S12 Hallonquist JD, Seeman MV, Lang M, Rector NA (1993) Variation in symptom severity over the menstrual cycle of schizophrenics. Biol Psychiatry 33: 207-209 Hoff AL, Kremen WS, Wieneke MH, Lauriello J, Blankfeld HM, Faustman WO, Csernansky JG, Nordahl TE (2001) Association of estrogen levels with neuropsychological performance in women with schizophrenia. Am J Psychiatry 158: 1134-1139 Holsboer F (1983) Hormones. In: Hippius H, Winokur G (eds) Psychopharmacology 1. Excerpta Medica, Amsterdam Oxford Princeton, 144-161 Honigfeld G, Gillis RD, Klett CJ (1976) NOSIE: Nurses’ observation scale for inpatient evaluation. In: Guy W (ed) ECDEU Assessment Manual for Psychopharmocology. Rockville MD, NIMH Hruska RE, Silbergeld EK (1980) Estrogen treatment enhances dopamine receptor sensitivity in the rat striatum. Eur J Pharmacol 61: 397-400 Huber TJ (2001) Hormonspiegel bei Frauen mit psychotischen Erkrankungen. In: Riecher-Rössler A, Rohde A (eds) Psychische Erkrankungen bei Frauen – für eine geschlechtersensible Psychiatrie und Psychotherapie. Karger, Basel Freiburg Paris London New York New Delhi Bangkok Singapore Tokyo Sidney, 165-169 Huber TJ, Rollnik J, Wilhelms J, von zur Mühlen A, Emrich HM, Schneider U (2001) Estradiol levels in psychotic disorders. Psychoneuroendocrinology 26: 27-35 Hutchinson G, Bhugra D, Mallett R, Burnett R, Corridan B, Leff J (1999) Fertility and marital rates in first-onset schizophrenia. Soc Psychiatry Psychiatr Epidemiol 34: 617-621 Jeste DV, Lindamer LA, Lacro JP (2001) Gender differences in late-life schizophrenia and its treatment – abstract. Syllabus and Proc Summary of the American Psychiatric Association’s 2001 Annual Meeting, 310 Kendell RE, Chalmers JC, Platz C (1987) Epidemiology of puerperal psychoses. Br J Psychiatry 150: 662-673 Klaiber EL, Broverman DM, Vogel U, Kobayashi Y (1979) Estrogen therapy for severe persistent depressions in women. Arch Gen Psychiatry 36: 550-554 Koller WC, Weiner WJ, Klawans HL, Nausieda PA (1980) Influence of female sex hormones on neuroleptic-induced behavioral supersensitivity. Neuropharmacol 19: 387-391 Kraepelin E (1909) Psychiatrie, vol. 1–4, 8th ed. Barth, Leipzig Krafft-Ebing G (1896) Untersuchungen über Irresein zur Zeit der Menstruation: Ein klinischer Beitrag zur Lehre vom periodischen Irresein. Arch Psychiatry 8: 65-107 Kretschmer E (1921) Körperbau und Charakter. Untersuchungen zum Konstitutionsproblem und zur Lehre von den Temperamenten, 25th ed. Springer, Berlin Kuhl H (2002) Einfluss von Östrogenen und Gestagenen auf das Zentralnervensystem. In: Kuhl H (ed) Sexualhormone und Psyche. Thieme, Stuttgart, 9-17 Kulkarni J, de Castella A, Smith D, Taffe J, Keks N, Copolov D (1996) A clinical trial of the effects of estrogen in acutely psychotic women. Schizophr Res 20: 247-252

50

Estrogens and Schizophrenia Kulkarni J, Riedel A, de Castella AR et al. (2001) Estrogen – a potential treatment for schizophrenia. Schizophr Res 48: 137-144 Labhart A (1978) Klinik der inneren Sekretion. Springer, Berlin Heidelberg New York Lindamer LA, Lohr JB, Harris MJ et al. (1997) Gender, estrogen, and schizophrenia. Psychopharmacol Bull 33: 221-228 Lindamer LA, Buse DC, Lohr JB, Jeste DV (2001) Hormone replacement therapy in postmenopausal women with schizophrenia: positive effect on negative symptoms? Biol Psychiatry 49: 47-51 Maguire GA (2002) Prolactin elevation with antipsychotic medications: mechanisms of action and clinical consequences. J Clin Psychiatry 63 (Suppl 4): 56-62 Mahé V, Dumaine A (2001) Oestrogen withdrawal associated psychoses. Acta Psychiatr Scand 104: 323-331 Mall G (1959) Neuere Ergebnisse der Psycho-Endokrinologie. Ärztliche Praxis XI: 1357-1360 Mall G (1960) Diagnostik und Therapie ovarieller Psychosen. Zentralblatt Gesamte Neurol Psychiatrie 155: 250 McEwen BS, Alves SE, Bulloch K, Weiland NG (1998) Clinically relevant basic science studies of gender differences and sex hormone effects. Psychopharmacol Bull 34: 251-259 NAMS (2000) A decision tree for the use of estrogen replacement therapy or hormone replacement therapy in postmenopausal women: consensus opinion of the North American Menopause Society. Menopause 7: 76-86 Neumann NU, Frasch K (2001) Olanzapin und Schwangerschaft – zwei Kasuistiken. Nervenarzt 72: 876-878 Nicoletti F, Ferrara N, Patti F, Viglianesi M, Rampello L, Bianchi A, Reggio A, Scapagnini U (1983) Influence of sex steroids and prolactin on haloperidol-induced catalepsy. Brain Res 279: 352-358 Oesterlund MK (2002) The role of estrogens in neuropsychiatric disorders. Curr Opin Psychiatry 15: 307-312 Overall JE, Gorham DR (1962) The brief psychiatric rating scale. Psychol Report 10: 799-812 Rasgon NL, Small GW, Siddarth P, Miller K, Ercoli LM, Bookheimer SY, Lavretsky H, Huang SC, Barrio JR, Phelps ME (2001) Estrogen use and brain metabolic change in older adults. A preliminary report. Psychiatry Research: Neuroimaging Section 107: 11-18 Riecher A, Maurer K, Löffler W, Fätkenheuer B, an der Heiden W, Munk-Jørgensen P, Strömgren E, Häfner H (1990) Gender differences in age at onset and course of schizophrenic disorders. In: Häfner H, Gattaz WF (eds) Search for the causes of schizophrenia, vol. 2, 15-33 Riecher-Rössler A, Häfner H (1993) Schizophrenia and oestrogens – is there an association? Euro Arch Psychiatry Clin Neurosci 242: 323-328 Riecher-Rössler A (1994) Die Spätschizophrenie – eine valide Entität? Eine empirische Studie zu Risikofaktoren, Krankheitsbild und Verlauf. Habilitationsschrift, Universität Heidelberg-Mannheim Riecher-Rössler A, Häfner H, Stumbaum M, Maurer K, Schmidt R (1994a) Can estradiol modulate schizophrenic symptomatology? Schizophr Bull 20: 203-214

51

Estrogens and Schizophrenia Riecher-Rössler A, Häfner H, Dütsch-Strobel A, Oster M, Stumbaum M, van Gülick-Bailer M, Löffler W (1994b) Further evidence for a specific role of estradiol in schizophrenia? Biol Psychiatry 36: 492-494 Riecher-Rössler A, Löffler W, Munk-Jørgensen P (1997) What do we really know about late-onset schizophrenia? Eur Arch Psychiatry Clin Neurosci 247: 195-208 Riecher-Rössler A, Häfner H, Dütsch-Strobel A, Stumbaum M (1998) Gonadal function and its influence on psychopathology. A comparison of schizophrenic and non-schizophrenic female inpatients. Arch Women Ment Health 1: 15-26 Riecher-Rössler A (2002) Estrogen effects in schizophrenia and their potential therapeutic implications – review. Arch Women Ment Health 5: 111-118 Riecher-Rössler A (2003) Estrogens and schizophrenia. Curr Opin Psychiatry 16: 187-192 Ripley HS, Papanicolaou GN (1942) The menstrual cycle with vaginal smear studies in schizophrenia, depression and elation. Am J Psychiatry 98: 567-573 Schepp A (1997) Pilotstudie zur Frage eines überdauernden relativen Hypoöstrogenismus bei schizophrenen Frauen. Inauguraldissertation zur Erlangung des medizinischen Doktorgrades, Universität Heidelberg-Mannheim Schmidt PJ, Nieman L, Danaceau MA, Tobin MB, Roca CA, Murphy JH, Rubinow DR (2000) Estrogen replacement in perimenopause-related depression: a preliminary report. Am J Obstet Gynecology 183: 414-420 Scriba PC, von Werder K (1976) Hypothalamus und Hypophyse. In: Siegenthaler W (ed) Klinische Pathophysiologie. Thieme, Stuttgart, pp 278-306 Seeman MV (1983) Interaction of sex, age and neuroleptic dose. Compr Psychiatry 24: 125-128 Sichel DA, Cohen LS, Robertson LM et al. (1995) Prophylactic estrogen in recurrent postpartum affective disorders. Biol Psychiatry 38: 814-818 Smith S, Wheeler MJ, Murray R, O’Keane V (2002) The effects of antipsychoticinduced hyperprolactinaemia on the hypothalamic-pituitary-gonadal axis. J Clin Psychopharmacol 22: 109-114 Stahl SM (1998) Basic psychopharmacology of antidepressants, part 2: estrogen as an adjunct to antidepressant treatment. J Clin Psychiatry 59 (Suppl 4): 15-24 Stahl SM (2001a) Effects of estrogen on the central nervous system. J Clin Psychiatry 62: 317-318 Stahl SM (2001b) Why drugs and hormones may interact in psychiatric disorders. J Clin Psychiatry 62: 225-226 von Zerssen D, Koeller DM (1976) Klinische Selbstbeurteilungs-Skalen (KSb-Si) aus dem Münchener Psychiatrischen Informations-System (PSYCHIS München). Manual. Beltz, Weinheim Writing Group for the Women’s Health Initiative Investigators (2002) Risks and benefits of estrogen plus progestin in healthy postmenopausal women. JAMA 288: 321-333 Zhang-Wong J, Seeman MV (2002) Antipsychotic drugs, menstrual regularity, and osteoporosis risk. Arch Women Ment Health 5: 93-98 Zweifel JE, O’Brien WH (1997) A meta-analysis of the effect of hormone replacement therapy upon depressed mood. Psychoneuroendocrinology 22: 189-212

52

3 Gender Differences in Schizophrenia Heinz Häfner

Gender differences in schizophrenia have long been reported and discussed. Nearly 100 years ago Kraepelin (1909–1915) pointed to the higher age in women by several years at first admission for dementia praecox compared to men. Since then the finding has been replicated in more than 50 studies (for a review see Angermeyer and Kühn 1988). However, there are also other important domains of difference between the sexes, e.g., (1) Diagnoses, subtypes, and symptoms (2) Lifetime risk for psychosis and distribution of onset throughout life (3) Proxy (antecedent) and distant (genetic, pre- and perinatal) risk factors (4) Determinants and consequences of the gender difference in age at onset (5) Gender-specific illness behavior versus illness-specific deficits and symptoms (6) Course and outcome (7) Brain development and brain functioning (8) Treatment and care. We will deal with these eight domains, as far as informative data are available.

Material and Method Our analyses are based on the relevant literature and a populationbased sample of 232 first-illness episodes of a broad diagnosis of schizophrenia (ICD-9: 295, 297, 298.3 and .4), the ABC sample (= 84% of first admissions). The patients were aged 12–59 years and came from a semi-urban, semi-rural German population of 1.5 million. A detailed description of the sample has been given elsewhere (Häfner

53

Heinz Häfner

Fig. 1. ABC schizophrenia study, medium-term course (Source: Häfner and an der Heiden 1999)

et al. 1993a). The patients were assessed using the PSE (Wing et al. 1974), the SANS (Andreasen 1983), the PIRS (Biehl et al. 1989), the DAS (World Health Organization 1988; Jung et al. 1989), and other instruments immediately upon hospitalization. The onset and early course of the disorder were assessed retrospectively using the IRAOS interview (Häfner et al. 1992, 1999a, 2003). A subsample of 57 patients was compared with 57 controls matched for age, sex, and place of residence. The further illness course from first admission on was assessed prospectively in a subsample of 115 first-episode cases at five cross sections over 5 years using the same instruments and, additionally, the FU-HSD (WHO 1980) (Fig. 1).

Results Gender Differences in the Diagnosis and Symptoms of First-Episode Schizophrenia (Domain 1) The literature on sex differences in the symptoms and subtypes of schizophrenia mostly reports a greater frequency of positive and affective symptoms in women and a greater frequency of negative

54

Gender Differences in Schizophrenia

symptoms and insidious types of onset in men (Castle et al. 1993; Castle 1999). To test these hypotheses we compared diagnoses, subtypes, symptom clusters, and symptoms at exactly the same stages of illness: (1) in the first psychotic episode, (2) cumulatively in the early illness course from onset to first admission, and (3) at illness onset. As shown in Table 1, no significant differences were found in any of the clinical or the operationalized diagnoses, scores, or syndromes in the psychotic episode. A representative sample and the wide age range of 12–59 years may account for this result. The results of comparative neuropsychological studies are also inconsistent (Goldstein and Lewine 2000; Fitzgerald and Seeman 2000). Goldberg et al. (1995), for example, studied four independent cohorts of men and women with schizophrenia using a large test battery, but did not find any substantial neuropsychological gender differences. Table 1. Comparison of clinical and operationalised diagnoses and CATEGO subclasses, scores and index of definition at first admission – ABC study sample of 232 first-episode cases (= 84% of 276 first admissions) Diagnosis (%)* Schizophrenia broad definition (ICD-9: 295, 297, 298.3/4) Schizophrenia ICD 295

Females n = 124

Males n=108

100%

100%

87.1%

88.0%

n.s

79.0% 73.4% 13.7%

73.1% 67.6% 13.0%

n.s n.s n.s

p

Operationalized diagnosis CATEGO ICD 295 CATEGO class S* CATEGO: affective psychosis Scores (mean values)** PSE: Index of definition CATEGO: total score CATEGO subscores: DAH (delusions, hallucinations) BSO (behaviour, speech) SNR (specific neurotic syndrome) NSN (nonspecific neurotic syndrome) ** Chi2-tests ** t-tests

55

7.47 40.67

7.49 41.44

n.s n.s

10.83 8.04 7.11 14.69

10.01 7.85 7.68 15.91

n.s n.s n.s n.s

Heinz Häfner Table 2. Symptom clusters in the first psychotic episode Cluster

Nonspecific, negative, depressive

Delusional

Psychotic thought disorder

Auditory hallucinations, substance abuse

Disorganization/ psychotic thought disorder

Low values on all dimensions

49.2 50.8

45.2 54.8

40.6 59.4

48.4 51.6

50.0 50.0

42.3 57.7

Age at first admission (years) F = 0.293, df=5; p = 0.91 31.1

29.8

29.9

29.3

30.3

31.4

Sex: males (%) females (%) Chi2 = 1.1, df: 5; p = 0.95

Source: Häfner 2000

Table 3. The ten most frequent earliest signs of schizophrenia (independent of the course) reported by the patients1

Restlessness Depression Anxiety Trouble with thinking and concentration Worrying Lack of self-confidence Lack of energy, slowness Poor work performance Social withdrawal, distrust Social withdrawal, communication

Total (n = 232) %

Men (n = 108) %

Women (n = 124) %

19 19 18

15 15 17

22 22 19

16 15 13 12 11

19 9 10 8 12

14 20 15 15 10

10

8

12

10

8

12

Based on closed questions, multiple counting possible. All items tested for sex differences: * p ≤ 0.05 (Source: Häfner et al. 1995, modified) 1

56

p

*

Gender Differences in Schizophrenia

The six symptom clusters we derived from the psychotic prephase – representing empirical subtypes – showed no significant gender differences when controlled for age (Table 2), nor did the 10 most frequent initial signs of illness onset, except “worrying” (Table 3). However, worrying is also observed more frequently in women than men in populations not suffering from schizophrenia. Jablensky (1995) summed up results from the literature on sex differences in the expressions of schizophrenia as follows: “There is no unequivocal evidence of consistent sex differences in the symptom profiles of schizophrenia and particularly in the frequency of positive and negative symptoms.” To define clinical subtypes we used symptoms, type of illness onset, and course from the first sign to the climax of the first episode. To test the hypothesis postulating that an insidious type of onset and early course and a hebephrenic subtype are more frequent in men, an acute onset with predominantly positive symptoms is more frequent in women, we compared three types of onset – acute, subacute, and insidious – and three categories of initial symptoms – positive, negative and unspecific. As shown in Table 4, none of these criteria differed significantly between the sexes. This indicates that the proxy characteristics of schizophrenia, symptoms, psychopathological subtypes, types of onset, and early illness course show no major Table 4. Type of onset of schizophrenia – ABC first-episode sample n = 232

Type of onset* Acute (≤ 1 month) Subacute (> 1 month ≤ 1 year) Insidious or chronic (> 1 year) Type of first symptoms* Negative or nonspecific Positive Both

Total n = 232

Men n = 108

Women n = 124

18%

19%

17%

15% 68%

11% 70%

18% 65%

73% 7% 20%

70% 7% 22%

76% 6% 19%

* The variables listed, except “worrying”, showed no significant sex differences. (Source: Häfner et al. 1995, modified)

57

Heinz Häfner

differences between men and women, when carefully compared on the basis of mean values in a wide age range.

Gender Difference in Morbidity Risk (Domain 2) The male-to-female ratios of the annual incidence and lifetime risk of schizophrenia vary considerably in the literature. We have discussed the methodological pitfalls that presumably explain the male predominance in the majority of the studies on the topic (Hambrecht et al. 1994; Häfner and an der Heiden 1997; see also Lewine et al. 1984; Goldstein and Lewine 2000; Castle et al. 1993). As reasons for this, an overrepresentation of young males and an underrepresentation of primarily female late-onset cases in many of the samples studied have been identified (e.g. Bland 1977). Another problem is the representativeness of the populations of origin and the samples studied for the total populations. These preconditions for studying sex differences in schizophrenia are difficult to fulfill. Hegarty et al. (1994), for example, did not find a single study using a representative sample among 320 follow-up studies of schizophrenia from a period of nearly a 100 years (1895–1992). The rare methodologically sophisticated studies show a trend towards convergence in the male-to-female lifetime prevalence rates for schizophrenia of a broad, but precise diagnostic definition (e.g., Jablensky et al. 1992; Häfner and an der Heiden 1997). However, the same does not apply to rates based on diagnoses including a 6-month social course prior to first contact as a criterion, as is the case, for example, with the DSM-III-R and -IV diagnoses (APA 1987, 1994). This will be explained later (Castle 2000). In addition, it is not clear whether the rates would converge if schizophrenia-like delusional disorders of old age (late paraphrenia, etc.), which show markedly higher incidence rates for women, were included. We calculated cumulative incidence rates – a good indicator of the lifetime risk – as based on 5-year age bands of the population studied, until age range 54–59 years at first admission. As Fig. 2 shows, men consumed their lifetime risk until age-range 30–35 years more rapidly than women did. From that age on, however, women caught up with men, finally reaching almost the same lifetime rate at about 13/100,000.

58

Gender Differences in Schizophrenia

Fig. 2. Cumulative incidence rates for schizophrenia, broad definition (ICD 295, 297, 298.3 and 298.4; source: Häfner et al. 1991)

Consequently, lower age cut-offs are bound to lead to a male predominance in the risk ratios. This result provided further support for our hypothesis that the disorder as such is essentially the same in men and women.

Childhood and Youth Antecedents of Schizophrenia (Domain 3) Normal early childhood development (Richman et al. 1982; Earls 1987) differs very little between the sexes. In late childhood, boys exhibit more expansive behaviors and a slightly higher frequency of attention deficits and girls more anxiety. From puberty on, the mental health risks of males and females follow different lines, males showing a greater frequency of hyperactivity, attention deficit dis-

59

Heinz Häfner

Males – age 11

Females – age 7

Females – age 11

An

An

xie ty An tow Ho xiety ards sti to ch lit w il Ho y tow ards dren Inc stili ard ad on ty t s c ults se ow hi qu ar ldr en ds en tia ad lb u e lt Un Re hav s for stl iou thc ess r om ne in ss W gne ith ss d D ra W epr wal riti es n s Mi sc Mis g of ion . n c. ad erv sy ul ou mp ts s s to ym ms pt om s

Males – age 7

xie ty An tow Ho xiety ards sti to ch lit w il Ho y tow ards dren Inc stili ard ad on ty t s c ults se ow hi qu ar ldr en ds en tia ad lb u e lt Un Re hav s for stl iou thc ess r om ne in ss W gne ith ss d D ra W epr wal riti es n s Mi sc Mis g of ion . n c. ad erv sy ul ou mp ts s s to ym ms pt om s

Mean percentage score

order, dissocial behavior, aggressiveness, and antisocial personality disorder and females a greater frequency of neurotic and affective disorders. These different age- and sex-specific behavioral trends must be taken into account in schizophrenia, too. As shown in studies based on teachers‘ and parents‘ reports (Watt et al. 1984), on offspring of schizophrenic mothers (Erlenmeyer-Kimling et al. 1993; Cannon et al. 1993; Cannon and Mednick 1993; Parnas et al. 1993), on two British (Jones et al. 1995) and one North-Finnish (Isohanni et al. 1998a, b) birth cohort of the respective populations and in a retrospective sibling study, adult-onset schizophrenia is preceded by mild neuromotor, cognitive, and behavioral anomalies. The minor early-childhood deficits in neuromotor and speech development seem to occur at the same frequency in boys and girls. From school age on, as Fig. 3 shows, behavioral anomalies as antecedents of schizophrenia manifest themselves several years later in girls than in boys. These anomalies are particularly severe in children of schizophrenic mothers, boys clearly scoring higher than girls on cognitive impairment (Erlenmeyer-Kimling et al. 1984; Castle 2000). Walker et al. (1995) compared childhood videos of siblings discordant for schizophrenia. They showed that premorbid behavioral signs become manifest somewhat later in girls than in boys (Fig. 4). However, similarly to the behavioral patterns of normal children,

Fig. 3. Antecedents of schizophrenia: social adjustment (BSAG) of later schizophrenic boys and girls at age 7 and 11, British Child Development Study (Source: Crow et al. 1995)

60

Gender Differences in Schizophrenia Mean score

Mean score

Fig. 4. Antecedents of schizophrenia: comparison of preschizophrenic and control siblings, mean externalized/internalized behavior problem scores by age period and sex (Source: Walker et al. 1995, modified)

boys exhibit primarily externalizing behaviors (e.g. hyperactivity, physical and verbal aggression, failure of behavioral inhibition), whereas girls manifest mainly “internalizing” (introvertive) behaviors, e.g., shyness, social withdrawal, depressive mood, and social anxiety. Most studies report slightly superior premorbid social and cognitive functioning and better school and work achievement for females than for males (Mueser et al. 1990). This was also shown by the Israeli conscript study with follow-ups of 4–10 years before first admission (Weiser et al. 2000). However, it must be borne in mind that in several studies the “premorbid” period prior to first hospital admission or first psychotic episode is contaminated with age- and gender-specific behavior and an often lengthy prodromal phase of the disorder. This may affect the gender distributions of diagnoses that include the “premorbid” social course as a criterion (Castle 2000).

The Gender Difference in Age at Onset and Its Consequences (Domain 4) The gender difference in age at first admission is a hallmark of the disorder, as Lewine (1980), Seeman (1983), Angermeyer and Kühn

61

Heinz Häfner Age in years

36

Moskau

Females

Males

36,2

35 34 33 32 31

Aarhus

32,8

Moskau

Prag

32,6

31,7

Nottingham

Aarhus

30,4

29,6

Total

30

30,1

Dublin

29

Dublin

29,4

30,0

28

28,5

Chandigarh

29,0

27

27,1

26

26,2

25

25,5

25,1

25,4

25,1

24 23

Nottingham Chandigarh

Ibadan

27,5

Total

Nagasaki

26,7

Honolulu

Agra

Prag 26,2

Agra

Honolulu

Nagasaki

Cali

24,9

24,7

Ibadan

24,2

22

Cali

21,7

21

Fig. 5. WHO Study: mean age at first admission (Source: Hambrecht et al. 1992)

(1988), and others have pointed out. The pooled data of the World Health Organization ten-country Determinants of Outcome Study (DOSMED) (Hambrecht et al. 1992; Hambrecht et al. 1994) revealed a 3.4 years higher mean age at onset for women than men (Fig. 5). In this context most of the conflicting results have been obtained on selected samples, mostly based on low age cut-offs (Hambrecht et al. 1992). In the ABC first-episode sample, illness onset and the consecutive milestones of the early course were dated by means of the IRAOS interview (Häfner et al. 1992; 1999a). Mean age at the emergence of the first sign of the disorder, at the first negative and first positive symptom, and at the climax of the first episode – defined by the maximum of positive symptoms – ranged from 24 to 30 years (Fig. 6). All these milestones differed significantly by 3–4 years between men and women, similarly to the result of the multinational DOSMED study. As a result, it seems well established that the disorder becomes manifest clearly later in women than in men. Looking at the distribution of onsets in 5-year age bands over the entire age range, we found an early and steep increase with a

62

Gender Differences in Schizophrenia first positive symptom

20

Males

first episode (max. of pos. symptoms)

first negative symptom

first sign

first admission

25 22.5

30

24.1

35 age in years

26.7 27.8 28.2

(N=108)

Total

24.0

25.5

29.0

30.1

30.3

(N=232)

Females

25.4

(N=124) * = p ≤ 0.05 ** = p ≤ 0.01

26.7

* 20

30.9

*

32.1

*

25

30

32.2

** ** 35

age in years

Fig. 6. Mean age values at five definitions of onset until first admission, firstepisode sample of schizophrenia, broad definition (n = 232) (Source: Häfner 1996)

* p < 0.05

Fig. 7. Distribution of age at onset of schizophrenia (first ever sign of mental disorder) by sex, ICD 9: 295, 297, 298.3 and .4), ABC Schizophrenia Study (Source: Häfner et al. 1993b)

63

Heinz Häfner

maximum between 15 and 25 years for men (Fig. 7). After that short period of maximum risk, the onset rates for men fell monotonously to a very low level. In women the rate of onsets rose slightly less steeply and reached a lower and broader peak in the age range 15–30 years. After a decline, similar to that in the male rates, female onsets reached a second, somewhat smaller peak in age group 45–50 years around premenopause with a significant difference to the male rates. The same pattern also emerged in the study of Castle et al. (1993) , based on the Camberwell case register, and in our analysis of first admission rates in a 1-year period from the Danish case register (Löffler et al. 1994).

The Estrogen Hypothesis and Testing It at Different Levels (Domain 4) The distribution of schizophrenia onsets across the female life-cycle inspired us to hypothesize that it might have something to do with the distribution of estrogen secretion. A protective effect of estrogen had previously been suggested by Mendelson et al. (1977), Seeman (1983), Loranger (1984), Lewine (1988), and Seeman and Lang (1990). A neuroleptic-like effect of short-term estrogen administration in animals had been shown by DiPaolo and Falardieu (1985), Fields and Gordon (1982), and Hruska (1986). We tested the estrogen hypothesis by a 4-week estrogen treatment of ovariectomized rats and found a significant attenuation of apomorphine-stimulated dopaminergic behavior compared with two control groups (the one with placebo, the other sham-operated). The effects were highest in young animals. We were able to demonstrate that estrogen reduces the sensitivity of D2 receptors (Häfner et al. 1991; Gattaz et al. 1992). A decade of experimental estrogen studies has shown that the sex hormone has potent neuromodulatory and neuroprotective effects. Sumner and Fink (1995), Fink et al. (1998), Shughrue et al. (1997), Sumner et al. (1999), McEwen et al. (1981), and Woolley and McEwen (1994) showed that estrogen acts with similar functional effects not only on D2 receptors, but also on 5-HT2, glutamate (NMDA), and GABA receptors at both the protein and the genomic level. In animal experiments, for example, estrogen stimulates the serotonin transporter gene. The functional effect seems to be analogous to the protective, antipsychotic effects of a reduced D2

64

Gender Differences in Schizophrenia

receptor sensitivity (Fink 1995). The effects of estrogen secretion on menstrual cycle-related and premenstrual cognitive functioning as well as in other mental disorders is not the topic of the present contribution. The applicability of the results of our animal experiments to human schizophrenia was shown by Riecher-Rössler et al. (1994a, b). Comparing 32 women with schizophrenic and 29 women with depressive episodes, both with normal menstrual cycles, we found significant negative correlations of increasing estrogen plasma levels with schizophrenia symptom scores in both groups of women, but no correlation with depressive symptom scores in either group (see Riecher-Rössler, this volume). From this we concluded that estrogen also has a weak neuroleptic-like effect on schizophrenic symptoms. An analogous variation in symptom severity over the menstrual cycle was also reported by Hallonquist et al. (1993), and similar clinical observations had previously been published by Dalton (1959) and Endo et al. (1978). Seeman and Cohen (1999), early proponents of the estrogen hypothesis, tested the hypothesis that an earlier onset of functional estrogen secretion with puberty might be associated with a later onset of schizophrenia in women. In line with the hypothesis, they found a significantly negative correlation between age at puberty and age at schizophrenia onset in women, but no correlation between age of puberty and age of schizophrenia onset in men (see Cohen and Seeman, this volume).

Age Difference in Severity of Illness Between Men and Women (Interaction of Age and Gender) (Domain 4) Assuming that a greater severity of illness is associated with an early outbreak of the illness, men, lacking the protective effect of estrogen, would be expected to develop the most severe forms of the disorder fairly early and, with increasing age, increasingly milder forms (Fig. 8). In women, as long as estrogen remains effective, the disease should be slightly milder, and a certain proportion of schizophrenias should not become manifest until menopause. From premenopause on, with decreasing estrogen secretion, women should not only show higher incidence rates, as depicted in Table 5, but also present more severe forms of the disorder.

65

Heinz Häfner

Fig. 8. A theoretical model of age-dependent effects of different vulnerability theresholds in men and women (Source: Häfner et al. 1998b)

Testing these hypotheses we found significantly lower PSE total scores for late-onset patients, taking men and women together. A comparison of symptom scores in early- and late-onset schizophrenia (age at onset 20 years or younger versus 40 years or older) by gender yielded different age trends for men and women, which was in line with the hypothesis (Table 6). Four out of eight symptom scores were significantly lower for late-onset men than for their early-onset counterparts. In contrast, in late-onset women not a single symptom score was significantly lower and one, the SANS global score denoting negative symptoms, was significantly higher than in earlyonset cases. This means that the milder symptoms of late-onset schizophrenia are accounted for by men alone. Compared with earlyonset episodes, late-onset psychotic episodes in women are not less severe, but sometimes even more severe. These gender-different age

66

Gender Differences in Schizophrenia Table 5. Onset of schizophrenia by age and sex – ABC first-episode sample (n = 232) Age at first psychotic symptom

n

Men

Women

12–20 years 21–35 36–59

49 136 47

57% 48% 32%

43% 52% 68%

* Odds ratio = 2.16

m/f ratio 1.33 0.92 0.47*

(sex ratio in the age group against the sex ratio in the remaining age groups); p < 0.05

m/f = male/female (Source: Häfner and Nowotny 1995)

trends in the severity of psychotic episodes support our hypothesis of a clearly age-dependent protective effect of estrogen secretion. Men develop relatively severe first episodes at young age, whereas young women present slightly milder cases. In clear agreement with our results at the symptom level, Lewine et al. (1997), who studied the interaction of sex and age of onset of schizophrenia, found a worse cognitive outcome for early-onset males (< 25 years) than for late-onset males and a poorer outcome for late-onset females than for early-onset females. This finding had been Table 6. Symptoms at early and late onset of schizophrenia in comparison (age at first psychotic symptom < 21 years vs. ≥ 40 years) Men Symptoms

Early onset n = 28

Vs. Wilcoxon

DAS BSO SNR NSN Total score SANS PIRS DAS-M

12.1 8.6 10.7 18.9 50.3 9.3 10.7 3.0

.02* .29 .11 .03* .02* .29 .29 .06t

Women Late onset n=9 ↓5.7 7.3 7.3 ↓11.4 ↓31.8 6.6 8.4 ↓1.8

t p ≤ 0.1; * p ≤ 0.05; ↓, ↑ = direction of difference (Source: Häfner et al. 1998a)

67

Early onset n = 21

Vs. Wilcoxon

Late onset n = 24

10.0 8.9 8.2 13.0 40.0 6.7 9.8 1.9

.95 .44 .42 .58 .80 .08t .73 .61

10.5 7.9 7.1 13.8 39.2 ↑9.5 10.5 1.8

Heinz Häfner

missed in previous studies because of lack of attention to the age variable. Also in line with our hypothesis are the results of several longterm studies showing that schizophrenia of an early onset and a lengthy course has a poorer symptom-related outcome from menopause on in women than in men of the same age or in younger women (Opjordsmoen 1991).

Strength of Predisposition and the Protective Effect of Estrogen (Domain 4)

Age in years

Leboyer et al. (1992), DeLisi et al. (1994), and Albus and Maier (1995) showed that there is no major gender difference in age at onset in familial schizophrenia. In contrast, the definitely nonfamilial cases of Albus and Maier’s sample had a gender difference as pronounced as 5.7 years. In our replication study of the ABC first-episode sample, the gender difference in age at psychosis onset did indeed fall from 4.2 in the total sample to a nonsignificant 1.6 years in familial cases (i.e., study subject has at least one first-degree relative with schizophrenia) and, thus, below the level of signifigance (Könnecke et al.

Fig. 9. Mean age at first psychotic symptom by gender and familial load (first episode sample of schizophrenia n = 232) (Source: Könnecke et al. 2000)

68

Age in years

Gender Differences in Schizophrenia

Fig. 10. Mean age at first psychotic symptom by gender and presence/absence of pre- and perinatal complications (PPC), follow-up sample n = 87 (Source: Könnecke et al. 2000)

2000) (Fig. 9). In sporadic cases (i.e., study subject has no relative with schizophrenia) the gender difference was a highly significant 4.9 years. As predicted by the estrogen hypothesis, the age at onset difference between familial and nonfamilial cases was almost entirely limited to women, whereas men showed no significant difference in age at onset between familial and nonfamilial cases. These results supported Albus and Maier’s hypothesis: the stronger the patients‘ genetic liability, the weaker the effect of delay that estrogen has on schizophrenia onset (Albus and Maier 1995). We then tested whether the other risk factor of etiological relevance for schizophrenia, pre- and perinatal complications, also weakens the protective effect of estrogen. And it indeed does, though to a somewhat lesser extent than familial load (Fig. 10). Hence, estrogen is capable of warding off schizophrenia onset the stronger the weaker individual’s vulnerability or strength of predisposition is.

Support for the Estrogen Hypothesis from Intervention Studies (Domain 8) Reliable evidence for causal effects of protective factors on the timepoint of illness onset, severity of illness, or the course of schizo-

69

Heinz Häfner

phrenia can be obtained from intervention studies varying the protective factors. Kulkarni et al. (1999) tested the estrogen hypothesis by estrogen substitution in the treatment of psychosis. Conducting a double-blind study with two different dosages of estradiol as an adjunct to haloperidol medication they found significant, but dose-related improvement in psychotic symptoms. In two pilot studies, one with male patients suffering from schizophrenia, Kulkarni et al. (1996a; 2002) also found significant, but only shortterm antipsychotic effects of adjunctive estradiol treatment. Improvement in psychotic symptoms has also been reported to occur with the addition of a combined estrogen-progesterone oral contraceptive (Felthous et al. 1980).

Abnormalities in Brain Development and Morphology (Domain 7) A series of neuroanatomical and neuroimaging studies on gender differences in structural brain abnormalities have shown more pathological morphology in male than female patients (Andreasen et al. 1990; Bogerts et al. 1990; Lewine and Seeman 1995; Goldstein 1996). However, there are also several studies that have found no exaggeration of the normal gender dimorphism of the brain in schizophrenia (Flaum et al. 1995). The reasons for this inconsistency may have something to do with the methodologies of the studies (Goldstein 1993; 1995a, b; Goldstein and Lewine 2000; Lauriello et al. 1997). However, comparative analysis based on functional imaging (e.g., Gur and Gur 1990; Gur et al. 1995) and neurophysiological paradigms (Reite et al. 1989; 1993) have equally failed to provide a definitive answer to sex differences in schizophrenia. For this reason in particular, the case on the effects of steroid hormones on brain development and brain functioning and their implications for the risk of schizophrenia and manifestation of symptoms cannot be closed yet.

Gender Differences in Illness Behavior (Domain 5) To analyze sex differences in the first-illness episode in greater detail, we compared all 303 single items from the instruments we used for

70

Gender Differences in Schizophrenia

measuring symptoms, functional impairment, and social disability (PSE, SANS, PIRS, DAS and IRAOS) in the first episode. Controlling for multiple testing we found no significant gender differences in the positive and negative core symptoms, except for some instances of delusions, such as sexual delusions and delusions of pregnancy, which were more frequent in women. The most pronounced gender difference emerged with socially adverse behavioral items, such as self-neglect, reduced interest in a job, social withdrawal, and deficits of communication, which were all significantly more frequent in men in the first psychotic episode (Table 7). Only one – socially favorable – behavioral item was observed significantly more frequently in women: overadaptiveness/ conformity. The cumulative prevalence of drug and alcohol abuse, also shown by several other studies, was found significantly more frequently in men. Restlessness was the only other item that showed a significantly higher frequency in women. These findings, however, are presumably not specific to schizophrenia, considering the behavioral gender differences in normal development mentioned above. In view of the consistent reports of a higher frequency of conduct disorders, disruptive, antisocial, and

Table 7. Behavioral items with significant sex differences (from a total of 303 PSE, PIRS, SANS, DAS and IRAOS items)* – ABC first-episode sample, n = 232 More frequent in women:

More frequent in men:

a) Cumulative until first admission ■ Restlessness

■ Drug abuse ■ Alcohol abuse

b) Cross-sectional: at first admission ■ Overadaptiveness/ conformity

■ ■ ■ ■ ■ ■ ■ ■ ■

* Validated by split-half method for Â-correction (Source: Häfner 1998)

71

Self-neglect, Reduced interest in a job Social inattentiveness Deficits of free time activities Deficits of communication Social disability (overall estimate) Loss of interests Deficits of personal hygiene

Heinz Häfner

violent behavior, and substance abuse among young men in comparison with their female counterparts from population studies (Choquet and Ledoux 1994; Döpfner et al. 1997), we are probably dealing with a reflection of normal gender- and age-specific behavior here. The normal psychology of behavioral gender differences, reviewed, for example, by Maccoby and Jacklin (1974), has shown that boys and young men exhibit a higher frequency of aggressive behavior, in particular antisocial aggression, than girls and young women do, who display more prosocial aggression or aggressive inhibition and a greater acceptance of authority. This socially adverse illness behavior of men with schizophrenia is strongly age-dependent, however, showing the highest frequency before age 30. With increasing age, men’s illness behavior in schizophrenia becomes socially more favorable, indicating improved adjustment and, as a result, the male disadvantage in the social course of schizophrenia probably also decreases (Fig. 11).

Mean score 4.4 ANOVA: p = not significant 4.2

4.1

4.0 3.8 3.6 3.6 3.4 3.4 3.2 3.0 Age at first admission:

12–20 years

21–35 years

36–59 years

* Self-neglect; Deficits of free time activities; Deficits of communication; Reduced interest in a job; social disability (overall estimate); loss of interests; Deficits in personal hygiene, social inattentiveness

Fig. 11. Socially negative behavior* of men (significantly different to women’s) by age at first admission. ABC first-episode sample N = 232 (Source: Häfner 2000)

72

Gender Differences in Schizophrenia

Sex Differences in the Course of Schizophrenia (Domain 4) The milestones of incipient illness with an age difference of 3–4 years appear in a parallel sequence in men and women. The curves illustrating the accumulation of the three symptom categories as based on mean symptom scores until the climax of the first episode do not show essential differences between men and women either, underscoring the impression of a uniform illness process (Fig. 12). However, the sex differences in illness behavior presumably influence the social course of the disorder: Associated with women’s higher tendency to prosocial behavior is greater cooperativeness and a better therapy compliance, whereas men’s socially adverse behavior leads not only to a reduced therapy compliance, but also to disadvantages in the domains of work and partnership. A wealth of studies have reported a poorer short- and mediumterm course of schizophrenia for male than for female patients (e.g., Angermeyer et al. 1990). When the dimensions of illness course and outcome have been looked at separately in methodologically sound studies, the difference has turned out to be accounted for by the social and not the symptom-related course (Biehl et al. 1986; Salokangas et al. 1987; Häfner et al. 1999b). In our study the symptom-related course over 5 years after first admission indeed showed no significant difference between men and women either in the CATEGO global scores (Fig. 13) nor in the four subscores. The mean symptom scores from remission of the first episode on indicated, rather than a trend to deterioration, no significant change over time. By contrast, the social course of the disorder, indicated by social disability and measured by the DAS (WHO 1988), was significantly poorer for men throughout the 5-year period studied (Fig. 14). This result lends support to our hypothesis that men’s socially unfavorable illness behavior might contribute to their poorer social course and outcome. Real-life disadvantages are caused by deficits in social functioning. This leads to the question of when and to what extent social disability emerges in the course of the disorder. We traced dysfunctional social roles and dysfunctional overall behavior by means of the Disability Assessment Schedule (WHO 1988) retrospectively from first admission back. As Fig. 15 shows, the disabilities measured by the DAS items – dysfunctional social roles and behaviors – became manifest on average as early as 2–4 years before first admission. Hence, it was long before the first psychiatric contact that the pa-

73

Heinz Häfner

Fig. 12. Cumulative numbers of positive, negative and unspecific symptoms of onset until first hospital admission for schizophrenia (males = 108, females = 124) (Source: Häfner et al. 1995)

tients with schizophrenia first exhibited deficits in social functioning. For a reliable assessment of the social course and consequences of

74

Gender Differences in Schizophrenia n.s.

Mean CATEGO total score

women men

n.s.

n.s. n.s.

n.s.

men

n.s.

women

Fig. 13. Five-year course (from first admission – 6 cross sections) for men and women by the CATEGO total score (n = 115) (Source: Häfner 1998a)

a disorder, it is necessary to proceed from a baseline, the level of social development at the onset of the disorder – which is only rarely done in long-term studies of schizophrenia. We chose six key social roles characteristic of the main period of risk for schizophrenia and compared them at age of illness onset in three age groups: 20 or younger, 21–35 years, and 36 and older. Figure 16 shows a trivial finding: significant age differences in the six roles studied, i.e., school education, occupational training, employment, own income, own accommodation, and marriage or stable partnership. These roles indicate the level of social development at illness onset. It is higher, the higher the patients‘ age. As men fall ill 3–4 years earlier than women, the disorder intrudes into their social biographies earlier. A comparison of social-role performance at illness onset between male and female patients showed significant advantages for women in the domains of employment, own income, and marriage or stable partnership, in particular (Table 8). From these results, we inferred that due to the disorder’s intrusion 3–4 years earlier in men’s social biographies, their lower baseline of social development at illness onset and their socially

75

Heinz Häfner

DAS total score

men

women

0 : * : ***: n.s. :

p < 0.1 p < 00.5 p < 0.001 not significant

Fig. 14. Five-year course of social disability for men and women from first admission (six cross sections) by the DAS total score (n = 115) (Source: Häfner 1998)

adverse illness behavior might explain their more unfavorable social course compared with that of women. To see how schizophrenia affects social development after illness onset in men and women, we looked at the social role of marriage or stable partnership, comparing 57 first-episode cases of schizophrenia from Mannheim with 57 controls matched for age, sex, and place of residence. At illness onset there was no significant difference between patients and healthy controls, male or female. However, because of the age difference of 4 years and men’s 2.5 years higher age of marrying in the general population at that time, male and female patients showed a significant difference (Fig. 17): during the period of 6 years the percentage of healthy men married or in a stable partnership gradually approached that of healthy women, whereas the figures for men and women with schizophrenia fell continuously after illness onset. Female patients, however, managed to retain their

76

Gender Differences in Schizophrenia DAS-Items

Dysfunctional overall behaviour

Dysfunctional in social roles

60

50

40 30 20 months before index admission

10

1.1

Self-care

1.2

Underactivity

1.3

Slowness

1.4

Social withdrawal

1.5

Social contacts

1.6

Emergencies

2.1

Participation

2.2

Marital / affective

2.3

Marital / sexual

2.4

Parental

2.5

Sexual

2.6

Work performance

2.7

Interest in job

2.8

Information

0

Fig. 15. Onset of social disabilities (months before index admission) (Source: Häfner et al. 1996, modified)

significant advantage over their male counterparts at first admission: 33% of the women with schizophrenia, but only 17% of the male patients were living with a spouse or partner, compared with 78% of the female and 60% of the male controls.

Predictor Model (Domain 4) The next step was to test the predictive power of the two main variables of social course and outcome – level of social development at psychosis onset and socially adverse illness behavior at first admission. Figure 18 illustrates two models: on the right a stepwise logistic regression including symptoms at first admission measured by the PSE, type of onset, age at first psychotic symptom, and gender, and on the left a pathanalytic model for analyzing partial correlations of age at onset and gender with the two social variables. Significant predictors of 5-year social outcome, operationalized by the ability to earn one’s living, were the number of nonfulfilled social roles at

77

Heinz Häfner

Fig. 16. Stage of social development by performance of 6 key roles at the onset of schizophrenia (first sign) by 3 age groups, ABC first-episode sample n = 232 (Source: Häfner et al. 1998a)

Table 8. Social-role performance at the emergence of the first sign of mental disorder – ABC first-episode sample

Age (in years)

Men n = 108 22.5 %

School education Occupational training Employment Own income Own accommodation Marriage or stable Partnership * p ≤ 0.05;

** p ≤ 0.01;

Women n = 124 25.4 %

Total n = 232 24.0 %

70 41 37 44 39

n.s * n.s. *

69 38 52 55 54

70 39 45 50 47

28

**

52

41

n.s. = not significant

(Source: Häfner 1996, modified)

78

Gender Differences in Schizophrenia

1st sign

** = p ≤ 0.01

* = p ≤ 0.05

1st psychotic symptom

First admission

t = p ≤ 0.1

Fig. 17. Social role performance in men and women with schizophrenia from illness onset to first admission: marriage or stable partnership-ABC first-episode sample N = 232 (Source: Häfner et al. 1999b)

psychosis onset and the number of items of socially adverse illness behavior at first admission. Symptoms and type of onset had no effect, age and gender merely that mediated by the first two variables. This is clearly shown in the pathanalytic model on the left, which revealed highly significant correlations of age at onset and gender with the two mediating variables, social development at illness onset and illness behavior. This means that the sex difference in the social course of schizophrenia, instead of reflecting a gender-different illness, is basically a result of the protective effect of estrogen in women and the sex-specific illness behavior of men.

79

Heinz Häfner

Prodromalphase

1st sign of mental disorder

Number of social roles not fulfilled at first signs

1st psychotic symptom

5 years after first admission

First admission

Financial independence

Number of social roles not fulfilled at onset

0.76 *

.682 ***

-.642 **

Age at first signs

.224 **

Gender

0.40 * Socially adverse illness behaviour

Logistic regression model - odds ratio -

-.321 ** Schizophrenia ABStudy C

path analysis - standard beta coefficients -

Source: Häfner et al. 2000 (ECNP)

Fig. 18. Prediction of 5-year social outcome (financial independence), first-episode sample n = 115

Sex Differences in the Long-term Course of Schizophrenia (Domain 6) There are only few methodologically high-standard follow-up studies of representative first-episode samples extending over 10 years or more. The studies by an der Heiden et al. (1995; 1996), Opjordsmoen (1991), and Goldstein (1988) showed that the gender differences in the early course of the disorder become diluted over long periods of follow-up. Harrison et al. (1996) reported a gender effect in a small first-admission sample sustained over a 13-year period after adjustment for sociodemographic variables and type of early course. The Mannheim cohort (an der Heiden et al. 1995; 1996) of the WHO Disability Study was assessed at ten cross sections over 15.6 years after first admission (Fig. 19). The repeated single measurements, based on the PSE total score, demonstrated a high degree of stability in the mean symptom scores over the long-term course. Women, showing significantly lower symptom scores only in the first 1.5 years after first admission – primarily because of their shorter first episodes – attained the level of male scores in the long-term. From 2 to 15.6 years after first admission, men and women showed almost equal symptom levels (Fig. 20).

80

Gender Differences in Schizophrenia

Fig. 19. Mannheim long-term schizophrenia project: cross-sectional assessments (Source: an der Heiden et al. 1996)

Fig. 20. PSE total score over 15.6 years after first admission by sex – first-admission sample of the WHO Disability study Mannheim cohort N = 70 * = p ≤ 0.05, *** = p ≤ 0.001 (Source: an der Heiden et al. 1995)

81

Heinz Häfner

Percentage of patients

The gender differences in social disability, still present at the 5year assessment after first admission, had disappeared at the later assessments, possibly as a result of the age-related decrease in socially negative male behavior. During the last 9 months before the 14.9-year assessment (no. 9) the percentages of good (no symptoms or disability present) and poor outcomes (suffering from at least one positive or negative core symptom or disability in the last 9 months) – 40% versus 60% – did not show any sex difference, as illustrated in Fig. 21. However, two thirds of the symptom-free women and only one tenth of the symptom-free men continued to be on antipsychotic drugs. We assume that this gender difference also resulted from sexspecific illness behavior and women’s generally better cooperativeness and compliance with treatment measures as compared with men’s. In spite of the similar symptom-related outcomes, there were considerable differences in social outcome between men and women: 71% of the male patients, but only 23% of the female patients had never married. Consequently, only 28% of the men, but 53% of the

Fig. 21. Two outcome groups 14.9 years after first admission by sex. Based on data from an der Heiden et al. 1996 (Source: Häfner 2000)

82

Gender Differences in Schizophrenia Table 9. Living situation of schizophrenic men and women 15.5 years after first admission (WHO “Disability Study” Mannheim cohort) N = 70 at inclusion in the study

Mean age Outcome: Symptoms or disability Present Living situation: Never married Married+ Lives with a spouse/ partner+ Lives in a home+ Own children+ Employment status: Has a regular job

Women n = 22

Men n = 34

44 years

41 years

n.s.

59%

62%

n.s.

23% 42%

71% 19%

** t

53% 5% 45%

28% 28% 26%

* t *

26%

31%

n.s.

n.s.: not significant t: p ≤ 0.1 *: p ≤ 0.05 **: p ≤ 0.01 (source: Based on data from an der Heiden et al. 1996)

women were living with a spouse and 28% of the men, but only 5% of the women were living in a supervised apartment or home. Naturally, more than twice as many women as men had children. Interestingly, however, there was no significant sex difference in employment status, which coincided with equal measures of social disability for men and women (Table 9). Women’s more favorable social outcome as compared with men’s, despite a similar symptom-related outcome, is obviously accounted for by women’s more favorable social conditions at illness onset and socially less adverse behavior in the course of the illness. However, a high proportion of the female patients who had married before illness onset or in the subsequent 15-year course of illness were divorced and some had re-married. It seems that because of their more prosocial behavior, presumably in conjunction with the traditional gender roles in society, women with schizophrenia, more often than their male counterparts, manage to find a new partner after a failed marriage or to some extent also to maintain an existing partnership.

83

Heinz Häfner

Premorbid Hypoestrogenism (Domain 4) In a historical review Riecher-Rössler and Häfner (1993) showed that the early writers in psychiatry, for example, E. Kraepelin (1909–1915) and E. Kretschmer (1921), had already noticed that persons with schizophrenia show deficits in the maturation of primary and secondary sex characteristics more frequently than their healthy peers. As a consequence, it was also early speculated that the risk for schizohprenia might be associated with gonodal hypofunctioning. Recently, several authors have fairly consistently reported subnormal serum levels of estrogen in women with first-onset schizophrenia not yet treated with psychotropic drugs (Riecher-Rössler, this volume). The temptation is now great to regard gonodal hypofunction as a risk factor amenable to preventive action. However, we are still a long way off from this goal, especially since the specificity and positive predictive power of hypoestrogenism for schizophrenia onset have not yet been demonstrated and are presumably rather low. In any case they do not justify any interventions involving risks. An unresolved question is whether the finding might be attributable to secondary, environmental causes, at least to some extent: The early deficits in communicability and mating behavior of preschizophrenic individuals presumably result in a lower stimulation of the gonodal function due to a sexually less stimulating environment. Last but not least, the equal lifetime risk of men and women for schizophrenia hardly speaks for a persistent risk factor specific to the female gender. Nothing is known yet on hypogonadism as a risk factor for schizophrenia in men.

Coping with Illness (Domain 5) Weber (1996) from our group looked into cognitive coping with the illness and subjective life satisfaction in this fairly homogeneous cohort by using a life goal and satisfaction questionnaire (FNL: Fragebogen zu Lebenszielen und Lebenszufriedenheit; Kraak and NordRüdiger 1989), based on Lehman’s (1983a, b) interaction model. She assessed subjective importance of life goals, goal achievement, and life domain-specific satisfaction. For the patients studied there was no correlation between symptom measures and social disability, on the one hand, and overall life satisfaction, on the other. However,

84

Gender Differences in Schizophrenia schizophrenics

controls

Fig. 22. Life-goal importance, achievement and satisfaction of men (n = 30) and women (n = 18) with schizophrenia 15.5 years after first admission compared with age- and sex-matched controls (Based on data from Weber 1996; Source: Häfner 2000)

patients and controls showed significant differences: 82% of the control men and 84% of the control women reported a high degree of overall life satisfaction, whereas only 43% of the male and 58% of the female patients did so (Fig. 22). Many life goals were considered only slightly less important by patients with schizophrenia than by healthy controls; for example, to be loved or to maintain stable relationships and self-esteem were almost equally important to both patients and healthy individuals. Only male patients considered sexual relationships and employment as important as did healthy controls. Women had significantly reduced their expectations in these life domains during the lengthy course of the disease, obviously as a means of coping with their diminished capacities. As a consequence, women with schizophrenia were generally more satisfied than their male counterparts, whose goal achievement and satisfaction with current status differed more markedly from their high expectations and from those of healthy controls. Nonetheless, women managed to achieve, to a greater

85

Heinz Häfner

extent than male patients, some of their highly valued aspirations in the domain of interpersonal relationships. Again, their illness behavior contributed to the more favorable social situation of women with schizophrenia and to their better coping with illness-related deficits and, as a result, to their slightly higher life satisfaction.

Summary and Conclusions The nuclear process of schizophrenia does not show hardly any significant differences between men and women in symptoms at illness onset, except for age effects, in the early course, in the first psychotic episode, and in the medium-term course. Significantly different between men and women is age at illness onset and at all the following milestones of the early illness course. This difference seems to be attributable to a protective effect of estrogen, which was also found to account for slight improvement in symptom scores and for different patterns of distribution of illness onsets over the male and female life cycles. When estrogen secretion starts waning in premenopause, women lose the protection. Subsequently, they show a second peak of onsets, more severe forms of first-episode illness, and a more severe course of early- and late-onset illness than men, who present fewer onsets and less severe types of schizophrenia in the second half of life. From the very beginning the social course of schizophrenia is more favorable in premenopausal women than men. The main reason is a difference in the baselines: the onset of schizophrenia hits men, on average younger than women, at a lower level of social and personality development. Hence, men’s further social development is impaired at a lower level than women’s, who have managed to attain a higher level of social and possibly also of cognitive and personality development before illness onset. In addition, detrimental to the social course of schizophrenia in men is their socially adverse illness behavior, most pronounced in adolescence and early adulthood. Men show a higher degree of social dysfunctioning and consequently suffer more severe social consequences than women, who tend to come to terms with the illness better than men and show a higher tendency to social conformity and therapy compliance. Conducive to a better coping with the illness in women is also the fact that women adjust their life goals and expectations to their

86

Gender Differences in Schizophrenia

changed capacities in the domains affected by the disorder. Hence, even in prolonged illness, women’s life satisfaction is better than men’s, who continue to hold on to goals in life that are no longer realistic to a greater extent than women do. Men’s more pronounced dissatisfaction possibly also contributes to their higher rate of suicide in the course of schizophrenia. To sum up briefly, our analyses showed that the nuclear disease process of schizophrenia does not essentially differ between the sexes. However, due to age-dependent hormonal and behavioral gender differences, the social and psychological course of the disorder shows considerable differences: men fare particularly poorly at young age and significantly better later in life and women considerably better until menopause, but worse afterwards. The dose-related antipsychotic effect of adjunctive estrogen treatment shown in the study Kulkarni et al. (1996a, b; 2002) offers the prospect of preventive and therapeutic medications that could be developed on the basis of the estrogen effects. It might even be possible to turn the selective estrogen receptor modulators (SERMs) into potent antipsychotic agents with low direct (e.g., feminizing in men) and long-term effects on breast and uterine tissue (risk of carcinoma). In any case the estrogen hypothesis seems to have therapeutic potential.

Acknowledgment A paper with the same title and some overlap with the present contribution has been published in the journal “Psychoneuroendocrinology” (2003; 28: 17–54).

References Albus M, Maier W (1995) Lack of gender differences in age at onset in familial schizophrenia. Schizophr Res 18: 51-57 an der Heiden W, Krumm B, Müller S, Weber I, Biehl H, Schäfer M (1995) Mannheimer Langzeitstudie der Schizophrenie. Nervenarzt 66: 820-827 an der Heiden W, Krumm B, Müller S, Weber I, Biehl H, Schäfer M (1996) Eine prospektive Studie zum Langzeitverlauf schizophrener Psychosen: Ergebnisse der 14-Jahres-Katamnese. ZMP 5: 66-75 Andreasen NC (1983) The scale for the assessment of negative symptoms (SANS). University of Iowa, Iowa City Andreasen NC, Ehrhardt JC, Swayze VW, Alliger RJ, Yuh WTC, Cohen G, Ziebell S (1990) Magnetic resonance imaging of the brain in schizophrenia: the patho-

87

Heinz Häfner physiologic significance of structural abnormalities. Arch Gen Psychiatry 47: 35-44 Angermeyer MC, Kühn L (1988) Gender differences in age at onset of schizophrenia. Eur Arch Psychiatry Neurol Sci 237: 351-364 Angermeyer MC, Kühn L, Goldstein JM (1990) Gender and the course of schizophrenia: Differences in treated outcomes. Schizophr Bull 16: 293-307 APA – American Psychiatric Association (1987) Diagnostic and statistical manual of mental disorders. 3rd, revised ed. American Psychiatric Association, Washington DC APA – American Psychiatric Association (1994) Diagnostic and statistical manual of mental disorders, 4th, revised ed. American Psychiatric Association, Washington, DC Biehl H, Maurer K, Jablensky A, Cooper JE, Tomov T (1989) The WHO Psychological Impairments Rating Schedule (WHO/PIRS). I. Introducing a new instrument for rating observed behaviour and the rationale of the psychological impairment concept. Br J Psychiatry 155: 68-70 Biehl H, Maurer K, Schubart C, Krumm B, Jung E (1986) Prediction of outcome and utilization of medical services in a prospective study of first onset schizophrenics – results of a prospective 5-year follow-up study. Eur Arch Psychiatry Neurol Sci 236: 139-147 Bland RC (1977) Demographic aspects of functional psychoses in Canada. Acta Psychiatr Scand 55: 369-380 Bogerts B, Ashtari M, Degreef G, Alvis JMJ, Bilder RM, Lieberman JA (1990) Reduced temporal limbic structure volumes on magnetic resonance images in first episode schizophrenia. Psychiatry Res Neuroimag 35: 1-13 Cannon TC, Mednick SA (1993) The schizophrenia high-risk project in Copenhagen: three decades of progress. Acta Psychiatr Scand 370 (Suppl): 33-47 Cannon TD, Mednick SA, Parnas J, Schulsinger F, Praestholm J, Vestergaard A (1993) Developmental brain abnormalities in the offspring of schizophrenic mothers: I. Contributions of genetic and perinatal factors. Arch Gen Psychiatry 50: 551-564 Castle DJ (1999) Gender and age at onset in schizophrenia. In: Howard R, Rabins P, Castle D (eds) Late onset schizophrenia. Wrightson Biomedical, Hampshire, 147-164 Castle DJ (2000) Women and schizophrenia: an epidemiological perspective. In: Castle DJ, McGrath J, Kulkarni J (eds) Women and schizophrenia. Cambridge University Press, Cambridge, 19-34 Castle DJ, Wessely S, Murray RM (1993) Sex and schizophrenia: effects of diagnostic stringency, and association with premorbid variables. Br J Psychiatry 162: 658-664 Choquet M, Ledoux S (1994) Epidémiologie et adolescence. In: Confrontations psychiatriques, vol 27 (no 35) Rhone-Poulenc Rorer Specia, Paris, 287-309 Crow TJ, Done DJ, Sacker A (1995) Birth cohort study of the antecedents of psychosis: ontogeney as witness to phylogenetic origins. In: Häfner H, Gattaz WF (eds) Search for the causes of schizophrenia, vol III. Springer, Berlin Heidelberg, 3-20

88

Gender Differences in Schizophrenia Dalton K (1959) Menstruation and acute psychiatric illness. Br Med J 1: 148-149 DeLisi LE, Bass N, Boccio A, Shilds G, Morganti C, Vita A (1994) Age of onset in familial schizophrenia. Arch Gen Psychiatry 51: 334-335 DiPaolo T, Falardeau P (1985) Modulation of brain and pituitary dopamine receptors by estrogens and prolactin. Progr Neuro-Psychopharmacol Biol Psychiatry 9: 473-480 Döpfner M, Pluck J, Berner W, Fegert JM, Huss M, Lenz K, Schmeck K, Lehmkuhl U, Poustka F, Lehmkuhl G (1997) Psychische Auffälligkeiten von Kindern und Jugendlichen in Deutschland – Ergebnisse einer repräsentativen Studie: Methodik, Alters-, Geschlechts- und Beurteilereffekte (Mental disturbances in children and adolescents in Germany. Results of a representative study: age, gender and rater effects). Z Kinder Jugendpsychiatr Psychother 25: 218-33 Earls F (1987) Sex differences in psychiatric disorders: origins and developmental influences. Psychiatr Dev 1: 1-23 Endo M, Daiguji M, Asano Y, Yanashita I, Lakahashi S (1978) Periodic psychosis occurring in association with the menstrual cycle. J Clin Psychiatry 39: 456-461 Erlenmeyer-Kimling L, Kestenbaum C, Bird H, Hilldoff U (1984) Assessment of the New York high-risk project subjects in sample A who are now clinical deviants. In: Watt N, Anthony N, Wynne EJ, Rolf LC (eds) Children at risk for schizophrenia: A longitudinal perspective. Cambridge University Press, Cambridge Erlenmeyer-Kimling L, Cornblatt BA, Rock D, Roberts S, Bell M, West A (1993) The New York high-risk project: anhedonia, attentional deviance, and psychopathology. Schizophr Bull 19: 141-153 Felthous AR, Robinson DB, Conroy RW (1980) Prevention of recurrent menstrual psychosis by an oral contraceptive. Am J Psychiatry 137: 245-246 Fields JZ, Gordon JH (1982) Estrogen inhibits the dopaminergic supersensitivity induced by neuroleptics. Life Sciences 30: 229-234 Fink G (1995) The psychoprotective action of estrogen is mediated by central serotonergic as well as dopaminergic mechanisms. In: Takada A, Curzon G (eds) Serotonin in the central nervous system and periphery. Elsevier Science, The Netherlands, 175-187 Fink G, Sumner B, McQueen JK, Wilson H, Rose R (1998) Sex steroid control of mood, mental state and memory. Clin Exp Pharmacol Physiol 25: 764-765 Fitzgerald P, Seeman MV (2000) Women and schizophrenia: treatment implications. In: Castle DJ, McGrath J, Kulkarni J (eds) Women and schizophrenia. Cambridge University Press, Cambridge, 95-110 Flaum M, Swayze VW, O’Leary DS, Yuh WT, Ehrhardt JC, Arndt S, Andreasen NC (1995) Effects of diagnosis, laterality, and gender on brain morphology in schizophrenia. Am J Psychiatry 152: 704-714 Gattaz WF, Behrens S, De Vrie J, Häfner H (1992) Östradiol hemmt Dopamin-vermittelte Verhaltensweisen bei Ratten – ein Tiermodell zur Untersuchung der geschlechtsspezifischen Unterschiede bei der Schizophrenie. Fortschr Neurol Psychiatr 60: 8-16 Goldberg TE, Gold JM; Torrey EF, Weinberger DR (1995) Lack of sex differences in the neuropsychological performance of patients with schizophrenia. Am J Psychiatry 152: 883-888

89

Heinz Häfner Goldstein JM (1988) Gender differences in the course of schizophrenia. Am J Psychiatry 145: 684–689 Goldstein JM (1993) Impact of sampling biases in explaining discrepancies in studies on gender and schizophrenia: a reply. Schizophr Bull 19: 9-14 Goldstein JM (1995a) The impact of gender on understanding the epidemiology of schizophrenia. In: Seeman MV (ed) Gender and psychopathology. American Psychiatric Press, Washington DC, 159-199 Goldstein JM (1995b) Gender and the famial transmission of schizophrenia. In: Seeman MV (ed) Gender and psychopathology. American Psychiatric Press, Washington DC Goldstein JM (1996) Sex and brain abnormalities in schizophrenia: fact or fiction? Harvard Rev Psychiatry 4: 110-115 Goldstein JM, Lewine RRJ (2000) Overview of sex differeces in schizophrenia: where have we been and where do we go from here? In: Castle DJ, McGrath J, Kulkarni J (eds) Women and schizophrenia. Cambridge University Press, Cambridge, 111-143 Gur RE, Gur RC (1990) Gender differences in regional cerebral blood flow. Schizophr Bull 16: 247-254 Gur RC, Mozley LH, Mozley PD, Resnick SM, Karp JS, Alavi A, Arnold SE, Gur RE (1995) Sex differences in regional cerebral glucose metabolism during a resting state. Science 267: 528-531 Häfner H (1996) The epidemiology of onset and early course of schizophrenia. In: Häfner H, Wolpert EM (eds) New research in psychiatry. Hogrefe and Huber Publishers, Seattle Toronto, 33-60 Häfner H (1998) Ist es einzig die Krankheit? In: Möller H-J, Müller N (eds) Schizophrenie – Moderne Konzepte zu Diagnostik, Pathogenese und Therapie. Springer, Wien, 37-59 Häfner H (2000) Gender differences in schizophrenia. In: Frank E (ed) Gender and its effects on psychopathology. American Psychopathological Association Series. American Psychiatric Press, Washington DC, 187-228 Häfner H, an der Heiden W (1997) Epidemiology of schizophrenia. Can J Psychiatry 42: 139-151 Häfner H, an der Heiden W (1999) The course of schizophrenia in the light of modern follow-up studies: the ABC and WHO studies. Eur Arch Psychiatry Clin Neurosci 249 (Suppl 4): IV/14-IV/26 Häfner H, Nowotny B (1995) Epidemiology of early-onset schizophrenia. Eur Arch Psychiatry Clin Neurosci 245: 80-92 Häfner H, Behrens S, de Vry J, Gattaz WF, Löffler W, Maurer K, Riecher-Rössler A (1991) Warum erkranken Frauen später an Schizophrenie? Nervenheilkunde 10: 154-163 Häfner H, Riecher-Rössler A, Hambrecht M, Maurer K, Meissner S, Schmidtke A, Fätkenheuer B, Löffler W, an der Heiden W (1992) IRAOS: an instrument for the assessment of onset and early course of schizophrenia. Schizophr Res 6: 209-223 Häfner H, Maurer K, Löffler W, Riecher-Rössler A (1993a) The influence of age and sex on the onset and early course of schizophrenia. Br J Psychiatry 162: 80-86 Häfner H, Riecher-Rössler A, an der Heiden W, Maurer K, Fätkenheuer B, Löffler W

90

Gender Differences in Schizophrenia (1993b) Generating and testing a causal explanation of the gender difference in age at first onset of schizophrenia. Psychol Med 23: 925-940 Häfner H, Maurer K, Löffler W, Bustamante S, an der Heiden W, Riecher-Rössler A, Nowotny B (1995) Onset and early course of schizophrenia. In: Häfner H, Gattaz WF (eds) Search for the causes of schizophrenia, vol III. Springer, Berlin Heidelberg, 43-66 Häfner H, Maurer K, Löffler W, Nowotny B (1996) Der Frühverlauf der Schizophrenie. ZMP 5: 22-31 Häfner H, Hambrecht M, Löffler W, Munk-Jorgensen P, Riecher-Rössler A (1998a) Is schizophrenia a disorder of all ages? A comparison of first episodes and early course over the life-cycle. Psychol Med 28: 351-365 Häfner H, an der Heiden W, Behrens S, Gattaz WF, Hambrecht M, Löffler W, Maurer K, Munk-Jörgensen P, Nowotny B, Riecher-Rössler A, Stein A (1998b) Causes and consequences of the gender difference in age at onset of schizophrenia. Schizophr Bull 24: 99-113 Häfner H, Löffler W, Maurer K, Riecher-Rössler A, Stein A (1999a) Interview für die retrospektive Erfassung des Erkrankungsbeginns und -verlaufs bei Schizophrenie und anderen Psychosen. Huber, Bern (English translation 2003) Häfner H, Maurer K, Löffler W, an der Heiden W, Stein A, Könnecke R, Hambrecht M (1999b) Onset and prodromal phase and determinants of the course. In: Gattaz WF, Häfner H (eds) Search for the causes of schizophrenia, Vol IV: Balance of the century. Springer, Berlin Heidelberg/Steinkopff, Darmstadt, 35-58 Hallonquist JD, Seeman MV, Lang M, Rector NA (1993) Variation in symptom severity over the menstural cycle of schizophrenics. Biol Psychiatry 33: 207-209 Hambrecht M, Maurer K, Sartorius N, Häfner H (1992) Transnational stability of gender differences in schizophrenia? An analysis based on the WHO study on determinants of outcome of severe mental disorders. Eur Arch Psychiatry Clin Neurosci 242: 6-12 Hambrecht M, Riecher-Rössler A, Fätkenheuer B, Louza MR, Häfner H (1994) Higher morbidity risk for schizophrenia in males: Fact or fiction? Compr Psychiatry 35: 39-49 Harrison G, Croudace T, Mason P, Glazebrook C, Medley I (1996) Predicting the long-term outcome of schizophrenia. Psychol Med 26: 697–705. Hegarty JD, Baldessarini RJ, Tohen M, Waternaux C, Oepen G (1994) One hundred years of schizophrenia: a meta-analysis of the outcome literature. Am J Psychiatry 151: 1409-1416 Hruska RE (1986) Evaluation of striatal dopamine receptors by estrogen: Dose and time studies. J Neurochem 47: 1908-1915 Isohanni M, Rantakallio P, Jones P, Järvelin M-R, Isohanni I, Mäkikyrö T, Moring J (1998a) The predictors of schizophrenia in the 1966 Northern Finland Birth Cohort Study. Schizophr Res 29: 11 Isohanni M, Järvelin M-R, Nieminen P, Jones P, Rantakallio P, Jokelainen J, Isohanni M (1998b) School performance as a predictor of psychiatric hospitalization in adult life. A 28-year follow-up in the Northern Finland 1966 Birth Cohort. Psychol Med 28: 967-974 Jablensky A (1995) Schizophrenia: the epidemiological horizon. In: Hirsch SR, Weinberger DR (eds) Schizophrenia. Blackwell Science, Oxford, 206-252

91

Heinz Häfner Jablensky A, Sartorius N, Ernberg G, Anker M, Korten A, Cooper JE, Day R, Bertelsen A (1992) Schizophrenia: manifestations, incidence and course in different cultures. A World Health Organization ten-country study. Psychological Medicine Monograph (Suppl 20) Cambridge University Press, Cambridge Jones PB, Murray RM, Rodgers B (1995) Childhood risk factors for adult schizophrenia in a general population birth cohort at age 43 years. In: Mednick SA, Hollister JM (eds) Neural development in schizophrenia. Plenum Press, New York, 151-176 Jung E, Krumm B, Biehl H, Maurer K, Bauer-Schubart C (1989) DAS – Mannheimer Skala zur Einschätzung sozialer Behinderung. Beltz, Weinheim Könnecke R, Häfner H, Maurer K, Löffler W, an der Heiden W (2000) Main risk factors for schizophrenia: increased familial loading and pre- and perinatal complications antagonize the protective effect of oestrogen in women. Schizophr Res 44: 81-93 Kraak B, Nord-Rüdiger D (1989) Der Fragebogen zu Lebenszielen und zur Lebenszufriedenheit (FLL). Hogrefe, Göttingen Kraepelin E (1909-1915) Psychiatrie, vol 1–4, 8th ed. Barth, Leipzig Kretschmer E (1921) Körperbau und Charakter, 25th ed. Springer, Heidelberg Kulkarni J, de Castella A, Smith D, Taffe J, Keks N, Copolov D (1996a) A clinical trial of the effects of estrogen in acutely psychotic women. Schizophr Res 20: 247-252 Kulkarni J, Gostt K, de Castella A (1996b) The menstrual cycle in women with schizophrenia. Schizophr Res 18: 254 Kulkarni J, de Castella A, Taffe J, Burger H, Reidel A (1999) Clinical estrogen trials in patients with schizophrenia. Curr Opin Psychiatry 12 (Suppl 1): 184–185 Kulkarni J, de Castella A, Downey M, White S, Taffe J, Fitzgerald P, Burger H (2002) Results of two controlled studies on estrogen: avenue to neuroprotection in schizophrenia. In: Häfner H (ed) Risk and protective factors in schizophrenia. Steinkopff, Darmstadt, 271-282 Lauriello J, Hoff A, Wieneke MH, Blankfeld H, Faustman WO, Rosenbloom M, DeMent S, Sullivan EV, Lim KO, Pfefferbaum A (1997) Similar extent of brain dysmorphology in severly ill women and men with schizophrenia. Am J Psychiatry 154: 819-825 Leboyer M, Filteau M-J, Jay M, Campion D, Rochert T, D’Amato T, Feingold J, Des Lauries A, Widlöcher D (1992) No gender effect on age of onset in familial schizophrenia? Am J Psychiatry 149: 1409 Lehman AF (1983b) The effects of psychiatric symptoms on quality of life assessments among the chronic mentally ill. Evaluation and Program Planning 6: 143-151 Lehman AF (1983a) The well-being of chronic mental patients. Arch Gen Psychiatry 40: 369-375 Lewine RJ (1980) Sex differences in age of symptom onset and first hospitalization in schizophrenia. Am J Orthopsychiatry 50: 316-322 Lewine RRJ (1988) Gender and schizophrenia. In: Nasrallah HA (ed) Handbook of schizophrenia, vol 3. Elsevier, Amsterdam, 389-397 Lewine RRJ, Seeman MV (1995) Gender, brain and schizophrenia: anatomy of differ-

92

Gender Differences in Schizophrenia ences/differences of anatomy. In: Seeman MV (ed) Gender and psychopathology. American Psychiatric Press, Washington DC, 131-158 Lewine RRJ, Burback D, Meltzer HY (1984) Effect of diagnostic criteria on the ratio of male to female schizophrenic patients. Am J Psychiatry 141: 84-87 Lewine R, Haden C, Caudle J, Shurett R (1997) Sex-onset effects on neuropsychological function in schizophrenia. Schizophr Bull 23: 51-61 Loranger AW (1984) Sex difference in age of onset of schizophrenia. Arch Gen Psychiatry 41: 157-161 Löffler W, Häfner H, Fätkenheuer B, Maurer K, Riecher-Rössler A, Lützhøft J, Skadhede S, Munk-Jørgensen P, Strömgren E (1994) Validation of Danish case register diagnosis for schizophrenia. Acta Psychiatr Scand 90: 196-203 Maccoby EE, Jacklin CN (1974) The psychology of sex differences. Stanford University Press, Stanford McEwen BS, Biegon A, Rainbow TC, Paden C. Snyder L. DeGroff V (1981) The interaction of estrogens with intracellular receptors and with putative neurotransmitter receptors: implications for the mechanisms of activation of regulation of sexual behaviour and ovulaton. In: Fuxe K, Gustafsson JA, Wetterberg L (eds) Steroid hormone regulation of the brain. Pergamon Press, New York, 15-29 Mendelson WB, Gillin JC, Wyatt RJ (1977) Sexual physiology and schizophrenia. Acta Scientifica Venezolana 28: 417-425 Mueser KT, Bellack AS, Morrison RL, Wixted JT (1990) Social competence in schizophrenia: premorbid adjustment, social skill, and domains of functioning. J Psychiatr Res 24: 51-63 Opjordsmoen S (1991) Long-term clinical outcome of schizophrenia with special reference to gender differences. Acta Psychiatr Scand 83: 307-313 Parnas J, Cannon TD, Jacobsen B, Schulsinger H, Schulsinger F, Mednick SA (1993) Lifetime DSM-IIIR diagnostic outcomes in offspring of schizophrenic mothers: Results from the Copenhagen High-Risk Study. Arch Gen Psychiatry 50: 707-714 Reite M, Teale P, Goldstein L, Whalen J, Linnville S (1989) Late auditory magnetic sources may differ in the left hemisphere of schizophrenic patients. Arch Gen Psychiatry 46: 565-572 Reite M, Cullum CM, Stocker J, Teale P, Kozora E (1993) Neuropsychological test performance and MEG-based brain lateralization: sex differences. Br Res Bull 32: 325-328 Richman N, Stevenson J, Graham P (1982) Preschool to school: a behavioural study. Academic Press, London Riecher-Rössler A, Häfner H (1993) Schizophrenia and oestrogens – is there an association? Eur Arch Psychiatry Clin Neurosci 242: 323-328 Riecher-Rössler A, Häfner H, Stumbaum M, Maurer K, Schmidt R (1994a) Can estradiol modulate schizophrenic symptomatology? Schizophr Bull 20: 203-214 Riecher-Rössler A, Häfner H, Dütsch-Strobel A, Oster M, Stumbaum M, van GülickBailer M, Löffler W (1994b) Further evidence for a specific role of estradiol in schizophrenia? Biol Psychiatry 36: 492-495 Salokangas RKR, Stengard E, Räkköläinen V, Kaljonen IHA (1987) New schizophrenic patients and their families (English summary). In: Reports of Psychiatria Fennica 78. Foundation for Psychiatric Research in Finland, 119-216

93

Gender Differences in Schizophrenia Seeman M (1983) Interaction of sex, age and neuroleptic dose. Compr Psychiatry 24: 125-128 Seeman MV, Lang M (1990) The role of estrogens in schizophrenia gender differences. Schizophr Bull 16:185-194 Seeman M, Cohen R (1999) Puberty and schizophrenia. In: Lopéz-Ibor JL, Sartorius N, Gaebel W, Haasen C (eds) Psychiatry on new thresholds. Abstracts of the 11th World Congress of Psychiatry, Hamburg, Germany, 6-11 August, 1999, vol I. (Curr Opin Psychiatry 12 [Suppl 1]) Shughrue PJ, Lane MV, Merchenthaler I (1997) Comparative distribution of estrogen receptor and X and B MRNA in the rat central nervous system. J Compr Neurol 388: 507-525 Sumner BEH, Fink G (1995) Oestradiol-17ß in its positive feedback mode significantly increases 5-HT2a receptor density in the frontal, cingulate and piriform cortex of the female rat. J Physiology 483: 52 Sumner BEH, Grant KE, Rosie R, Hegele-Hartung Ch, Fritzemeier K-H, Fink G (1999) Effects of tamoxifen on serotonin transporter and 5-hydroxytryptamine2A receptor binding sites and mRNA levels in the brain of ovariectomized rats with or without acute estradiol replacement. Mol Brain Res 73: 119-128 Walker EF, Weinstein J, Baum K, Neumann CS (1995) Antecedents of schizophrenia: Moderating effects of development and biological sex. In: Häfner H, Gattaz WF (eds) Search for the causes of schizophrenia, vol III. Springer, Berlin Heidelberg, 21-42 Watt NF, Grubb TW, Erlenmeyer-Kimling L (1984) Social, emotional, and intellectual behavior at school among children at high risk for schizophrenia. In: Watt NF, Anthony J, Wynne LC, Rolf LC (eds) Children at high risk for schizophrenia. Cambridge University Press, London, 212-226 Weber I (1996) Lebenszufriedenheit einer Kohorte Schizophrener 15,5 Jahre nach stationärer Aufnahme. Doctoral thesis, Mannheim Faculty of Clinical Medicine of the Ruprecht Karls University of Heidelberg Weiser M, Reichenberg A, Rabinowitz J, Kaplan Z, Mark M, Nahon D, Davidson M (2000) Gender differences in premorbid cognitive performance in a national cohort of schizophrenic patients. Schizophr Res 45: 185-190 WHO-World Health Organization (1980) Follow-Up History Schedule (WHO/ FU-HSD) Geneva WHO-World Health Organization (1988) Psychiatric Disability Assessment Schedule (WHO/DAS). WHO, Geneva Wing JK, Cooper JE, Sartorius N (1974) Measurement and classification of psychiatric symptoms: An instruction manual for the PSE and CATEGO program. Cambridge University Press, London Woolley CS, McEwen BS (1994) Estradiol regulates hippocampal dendritic spine density via an N-Methyl-D-aspartate receptor-dependent mechanism. J Neurosci 14: 7680-7687

94

4 Puberty and Schizophrenia Onset Robin Z. Hayeems and Mary V. Seeman

Introduction Estradiol, the derivative of estrogen that is most active in the brain, protects brain function in a variety of ways. There is evidence that estradiol exerts a positive effect on neuron viability. Though this may not cause an increase in the ultimate number of adult neurons, even a temporary effect may facilitate neuronal connectivity (Purves 1985). Estradiol also enhances neuron growth (Torand-Allerand 1984). Furthermore, estradiol may act independently by altering growth-related genes directly or may act in concert with growth factors (neurotrophins) and their receptors (Toran-Allerand 1996). Woolley and McEwen (1992; 1994a) have repeatedly demonstrated that estradiol enhances synaptic density. In 1992, they reported that synaptic density in the hippocampal CA1 region in the adult female rat was sensitive to estradiol and correlated with levels of ovarian steroids over the 5-day estrous cycle. In a 24-h period when estrogen levels dropped, hippocampal synaptic density decreased by 32%. They further demonstrated that estrogen treatment increased the density of dendritic spines on pyramidal neurons in the hippocampus by 35% (Woolley et al. 1997). Female sex hormones have long been known to affect the pace of myelination (Petropoulos et al. 1972). Failure of myelination to occur in a timely manner during adolescence may result in increased vulnerability to brain pathology (Benes 1989). Furthermore, estrogen seems to protect vulnerable hippocampal neurons from glucose deprivation (Bishop and Simpkins 1995; Goodman et al. 1996). Estrogen acts as an antioxidant, thereby preventing free radicals from fracturing membrane lipids, proteins, and DNA and leading to cell death. Glutamate and β-amyloid are two such toxins that produce free radicals and whose detrimental effects can be attenuated by estrogen (Behl et al. 1995; Goodman et al. 1996).

95

Robin Z. Hayeems and Mary V. Seeman

Estrogens also affect a variety of neurotransmitter systems. The system most commonly associated with schizophrenia is dopamine (DA). Other relevant neurotransmitters modulated by estrogen include serotonin, glutamate, and acetylcholine. Häfner et al. (1991) suggest that estrogen exerts a potent antidopaminergic effect, thereby raising the vulnerability threshold for psychosis in women. In ophorectomized/castrated neonatal and adult rats, they examined the effects of gonadal hormones, estradiol and testosterone, on behavioral changes (catalepsy) induced by the DA antagonist haloperidol and by the DA agonist apomorphine (oral stereotypies, grooming, and sitting behavior). Estradiol attenuated the behavioral changes induced by both haloperidol and apomorphine. This indicated downward regulation of DA neurotransmission by estradiol, a finding which was confirmed by a follow-up ligand binding investigation showing that estradiol pretreatment significantly reduced DA-receptor affinity to striatal 3H-sulpiride binding (Häfner et al. 1991). DiPaolo and her co-workers have extensively studied the relationship between estradiol and dopamine in rat models. In line with the findings of Häfner et al., they have consistently found that the density of dopamine uptake sites of estradiol-treated rats increases by approximately 20% (DiPaolo et al. 1981; Levesque and DiPaolo 1989; 1991; Morisette and DiPaolo 1993a, b). They suggest that estrogens behave in opposition to dopamine: estrogens decrease dopaminergic transmission and induce a compensatory increase in the number of dopamine binding sites. They have also shown that striatal dopamine uptake site density fluctuates during the female estrous cycle: binding site density increases as estrogen levels climb (Morisette and DiPaolo 1993). With respect to other neurotransmitters, Fink et al. (1996) have shown that estradiol modulates the serotonergic (5-HT) system. They suggest that positive psychotic symptoms may result from abnormal dopaminergic activity, whereas negative symptoms may result from abnormal 5-HT function.

Clinical Effects of Estrogen Riecher-Rössler (1994a, b) has examined the psychopathology over different phases of the menstrual cycle in premenopausal women

96

Puberty and Schizophrenia Onset

with schizophrenia. She used the Brief Psychiatric Rating Scale (BPRS), a nurses’ observation scale, and a self-rated symptomatology scale and measured serum estradiol levels. She demonstrated that symptoms improved when estrogen levels were high. Furthermore, Hallonquist et al. (1993), and Gattaz et al. (1994) concur that a significant proportion of women with schizophrenia experience variation in symptoms as a function of estrogen levels during the menstrual cycle. Kendall et al. (1987) further support this notion by presenting data which suggest that the high levels of estrogen during pregnancy protect women from psychosis. They report an increased number of psychiatric hospitalizations for psychosis postpartum relative to immediately antepartum. Seeman (1996) traced the course of women before, during, and after pregnancy. She found that psychotic symptoms worsened during estrogen-low phases, often remitted during pregnancy when estrogen levels increased, and then returned postpartum when estrogen levels fell. Gattaz et al. (1994), studied 65 women hospitalized for schizophrenia and 35 women hospitalized for affective disorder and showed that women with psychosis require lower neuroleptic doses during estrogen-high phases. Seeman (1989) suggests that, in order to maintain remission from psychosis, women need progressively higher doses of neuroleptics after age 40 when estrogen levels begin to fall. In addition, she suggests that compared to men, women’s substantially greater propensity to develop psychosis in middle age (40’s–50’s) may result from the failure of aging ovaries to produce sufficient estrogen (Seeman 1983). Counter-evidence is offered by Davies et al. (1995), who observed no protection against psychosis during pregnancy in his sample of schizophrenia and affective psychosis women and who observed affective rather than schizophrenic psychosis postpartum. Salokongas (1995) reports further counter-evidence to the hypothesis that estrogen acts protectively in schizophrenia. In 1,097 schizophrenic outpatients, he followed daily neuroleptic requirements over a 3-year period. In opposition to Seeman’s findings, he found that in women between the ages of 20 and 59, daily neuroleptic dose was fairly stable; there was no change in the required dose at menopausal age. At the 3-year follow-up point, during menopause, and thereafter, required daily doses decreased in both males and females, more markedly in females than in males. In only one group of women, those who had a late onset and a long duration of illness, were daily

97

Robin Z. Hayeems and Mary V. Seeman

doses required at menopause higher than they had been at a younger age. This study is important as it points out the importance of several intervening variables in the estrogen/psychosis connection.

Study Rationale Given the above, we decided to study the connection between the beginning of psychosis and the beginning of puberty, a time when gonadal hormones steeply rise. Characteristically, the onset of schizophrenia comes within 10 years of puberty onset. We predicted that the beginning of schizophrenia symptoms in women would correlate inversely with the age of menarche. The earlier the onset of menarche, the earlier the protective effects of estrogens can exert their effects. We postulated that early puberty in men would, by contrast to women, correlate with an earlier onset of schizophrenia symptoms. Although some testosterone is converted to estradiol in the brain, we hypothesized that the effect of testosterone itself would be associated with impulsivity, antisocial behavior, risk-taking, head trauma, and substance abuse, all of which could act to bring forward the onset of schizophrenia symptoms in vulnerable individuals.

Methodology Power calculation indicated that 35 men and 35 women would be needed to test this hypothesis in a one-tailed test of significance at the p < .05 level. Recruitment was from First Episode Psychosis Clinics in Toronto, Ontario, and Halifax, Nova Scotia, in Canada. Diagnosis was restricted to DSM-IV schizophrenia and schizoaffective disorder. Subjects were all within the first 10 years of diagnosis and, wherever possible, subjects’ mothers were also interviewed. This was possible in approximately 70% of cases. The percentage of maternal refusal was the same for women and men. Written informed consent was obtained from subjects and participating mothers. Clinical diagnoses were checked by submitting clinical records to a computerized diagnostic system (DTREE) which asks a series of questions about the presence and absence of symptoms (First et al. 1997). The Interview for the Retrospective Assessment of the Onset

98

Puberty and Schizophrenia Onset

of Schizophrenia (IRAOS) was used with subject and mother to date the first appearance of signs and symptoms (Häfner et al.1992). From this information, three data points were established: (1) age at first sign of odd behavior; (2) age at first psychotic symptom; and (3) age at first hospital admission. The same interview was used to date substance abuse behavior prior to the appearance of the first psychotic symptom. The Maturational Timing Questionnaire (MTQ) (Gilger et al. 1991) was used to achieve consensus between subject and mother on onset of various maturational indices. Menarche for women and voice change for men showed the highest concordance between reporters and were subsequently used in the analysis. A Head Injury Questionnaire administered to both subjects and mothers was used to determine whether the individual had suffered a head injury severe enough to cause loss of consciousness between the onset of puberty and the age of schizophrenia onset. Head injury was scored as a dichotomous variable. A Family History Questionnaire was used to inquire about psychotic disorder in first-degree relatives since that can influence schizophrenia onset. The Obstetric Complication Scale (Lewis et al. 1989), a checklist of 17 possible perinatal complications that can also influence schizophrenia onset, was scored on an ordinal scale: 0 for no complications, 1 for one or more equivocal but not definite complications, and 2 for one or more definite complications. This scale was administered to both subject and mother.

Data Analysis Bivariate correlations were performed between all variables to detect significant associations. To compare the predictive hypotheses between puberty and onset age in men and women, the data were grouped according to gender and analyzed using hierarchical linear regression. Separate analyses were run for each of the three dependent variables: first psychiatric hospitalization, first psychotic symptom, and first odd behavior. For each dependent variable, predictor variables were gender, age of puberty, an interaction term (gender × puberty), family history of schizophrenia, and the remaining independent variables (alcohol and drug use, head injury, and obstetric complications). To test the main effects, gender and age of puberty were entered

99

Robin Z. Hayeems and Mary V. Seeman

into the regression analysis as step one. To test the interaction, the product term (gender × age of puberty) was entered as step two. To test the influence of family history on gender and onset age, the family history variable was entered as step three. Finally, to test the contribution of the ancillary variables, alcohol use, drug use, head injury, and obstetric complication were entered as step four. For each step, sample size was adjusted according to the number of missing data points for the variables entered into each respective step.

Results The total sample consisted of 35 women and 45 men. When comparing the DTREE and the clinical diagnosis the diagnostic reliability (schizophrenia versus schizoaffective disorder) was slightly better for men: 85.7% versus 82.6%. There were significantly more women with schizoaffective disorder: 13/35 women versus 6/45 men. The relationship between puberty and schizophrenia onset was the same whether the schizoaffective group was included or excluded. It was also the same whether all subjects or only those with participating mothers were included. Mean ages of first odd behavior, first psychotic symptom, and first hospitalization for females were 19.9 (SD = 5.9), 26.7 (SD = 6.8), and 29.0 (SD = 6.8), respectively, and for males: 18.6 (SD = 5.9), 22.4 (SD = 5.1), and 24.2 (SD = 5.0), respectively. Using ANOVA, males were significantly younger than females for mean age of first hospitalization and mean age of first psychotic symptom (F(1) = 6.2, p = .015, F(1) = 9.0, p = .004, respectively) but not for mean age of first odd behavior. Mean ages of puberty in women, defined as first menstruation, was 12.7 (SD = 1.3). Mean age of puberty in men, defined as age of voice change, was 14.5 (SD = 1.6). An independent samples t-test showed that, prior to onset of the first psychotic symptom, men used alcohol and illicit drugs significantly more often than women [t(76) = 2.01, p = .048; t(76)= 2.35, p = .022)]. Significant head injury prior to onset of the first psychotic symptom occurred in 9.3% of men and 8.8% of women. Pearson Chisquare test showed no significant gender difference in the frequency of head injury [X2(1, n = 77) = .005, p = .94].

100

Puberty and Schizophrenia Onset

Obstetric complications had definitely occurred in 37.1% of women and 44.2% of men and possibly in approximately 50% of both genders. Pearson Chi-square test showed no significant gender difference [X2(2, n = 62) = 2.88, p = .24]. In both sexes, the complication usually occurred during delivery (as opposed to during pregnancy or after childbirth). Complications characteristic of delivery included the umbilical cord wrapped around baby’s neck, the use of forceps, cesarean section, or baby born in an abnormal position. Schizophrenia or schizoaffective disorder was present in a firstdegree relative of three women (8.6%) and a schizophrenic-like illness was suspected in first-degree relatives of three other women (8.6%). Psychotic illness was present in first-degree relatives of four men (9.3%) and suspected in one (2.3%). In both sexes, the subject’s mother was usually the relative reported to have experienced psychotic illness and this may have accounted for some of the maternal refusals to participate. Pearson Chi-square test showed no significant gender difference in the number of first-degree relatives affected with schizophrenia-like illness [X2(2, n = 78) = 1.55, p = .46]. Puberty Onset in Women Bivariate correlations were performed first in order to detect the strength of the association between dependent variables (first odd behavior, first psychotic symptom, and first hospitalization) and independent variables (age of puberty, substance use and head injury prior to onset, obstetric complications, and family history). The inverse correlation between puberty and odd behavior in women did not quite reach statistical significance (r = –.31, p = .07), but those between puberty and first psychotic symptom, as well as puberty and first hospitalization definitely did (r = .55, p = .001; r =–.57, p < .001, respectively). Simple linear regression analysis confirmed the strength and direction of these correlations. There were no meaningful correlations with the other variables. Puberty Onset in Men As in women, bivariate correlations were performed between dependent and independent variables in men. None of the three correlations were significant (r = .18, p = .23; r = .10, p = .49; r = .26, p = .09, respectively). Simple linear regression analysis confirmed

101

Robin Z. Hayeems and Mary V. Seeman

nonsignificant, but positive correlations between age of puberty and each onset definition in males. In other words, the trend was in the opposite direction to that in women. As in women, there were no other meaningful correlations. Age of First Hospitalization Using hierarchical regression analysis to show which factors best predicted age of first hospitalization for schizophrenia, gender itself was entered first and was significant. In step 2, the interaction term (gender × puberty) was entered into the model and it, too, significantly predicted age of first hospitalization. Gender remained as a significant predictor. In step 3, family history was entered and resulted in no significant change in the magnitude of the regression coefficient and, therefore, did not further influence age of first hospitalization. In step 4, the remaining independent variables were entered into the regression model. Neither alcohol use, drug use, head injury, nor obstetric complications contributed significantly to the prediction of age of first hospitalization. Age of First Psychotic Symptom Using hierarchical regression analysis for predicting first psychotic symptom, in step 1, gender was a predictor. In step 2, gender remained significant and the interaction term (gender × puberty) was significant. In step 3, family history proved to be a further significant predictor of age of first psychotic symptom. In step 4, neither alcohol use, drug use, head injury, nor obstetric complications significantly changed R2 nor the magnitude of the regression coefficients. Age at First Odd Behavior Using hierarchical regression analysis for first odd behavior, in step 1, gender was not significant. In step 2, gender became significant, and the interaction term (gender × puberty), was significant. In step 3, family history was not a significant predictor of age of first odd behavior. In step 4, neither alcohol use, drug use, head injury, nor obstetric complications significantly changed R2 nor the magnitude of the regression coefficients. There was one female outlier for age at puberty (age 17.0) and two

102

Puberty and Schizophrenia Onset

male outliers for age of puberty (age 19.0). When the analyses were run without these subjects, the predictive regression model for the three onset definitions was not altered. In steps 3 and 4 of each of the three analyses above, the sample size was reduced as a result of missing data. The family history variable was missing two data points; therefore, step 3 of each analysis was based on n = 78. Missing data for obstetric complications reduced the sample size to n = 62 for step 4 of each analysis. When steps 1 and 2 were repeated for each analysis using the smallest sample size (n = 62 instead of n = 80), the pattern of the results did not significantly change.

Discussion As hypothesized, the results show a strong inverse relation between female puberty and onset age of schizophrenia. There was no significant correlation in men but the trend was, interestingly, in the opposite direction. Bivariate correlations confirmed a significant inverse relationship between age of first menstruation in women and age of both first hospitalization and first appearance of psychotic symptoms. The same relationship was seen with first appearance of odd behavior, but it was weaker. Hierarchical regression strengthened this relationship. The direction and magnitude of the regression coefficients support an inverse relation between age of puberty and age of illness onset in women. The model used confirms that gender significantly predicts onset age and that the effect of gender is mediated by age of puberty. The effect of pubertal age on illness onset was not mediated by any other predictive variables tested in the analysis. Neither alcohol use, drug use, head injury, obstetric complication, nor family history added strength to the predictive model of onset age for women or for men. These findings suggest a protective effect of female hormones in schizophrenia but do not explain why the effect seems to work in the opposite direction for men even though testosterone is partially converted to estrogen in the brain. Female hormones may exert their effects indirectly by counteracting behavioral tendencies that result in illness. Such tendencies as studied in this investigation (substance

103

Robin Z. Hayeems and Mary V. Seeman

abuse and risk-taking, reflected by head trauma) had low group prevalence, did not distinguish men from women, and showed no predictive association with schizophrenia onset age. There are reasons to be cautious when interpreting these data. The sample was small; diagnoses were not elicited by standardized interviews. The interviewer was not blinded. The accuracy of patient and maternal recall can be called into question and external corroboration was not available for every data point. Nevertheless, the results of this study lend credence to the theory that female hormones act on the developing brain to protect its integrity and delay the expression of schizophrenic psychosis. Though not tested in this study, the effect may be relatively specific for schizophrenia since puberty is known to bring forward rather than to delay the expression of most other psychiatric syndromes in women such as depression, anxiety, and psychophysiological disorders.

References Behl C, Widmann M, Trapp T, Holspoer F (1995) 17beta-estradiol protects neurons from oxidative stress-induced cell death in vitro. Biochem Biophys Res Comm 216: 473-482 Benes FM (1989) Myelination of cortical-hippocampal relays during late adolescence. Schizophr Bull 15: 585-593 Bishop J, Simpkins JW (1995) Estradiol enhances brain glucose uptake in ovariectomized rats. Brain Res Bull 36: 315-320 Davies A, McIvor RJ, Kumar C (1995) Impact of childbirth on a series of schizophrenic mothers: a comment on the possible influence of oestrogen on schizophrenia. Schizophr Res 16: 25-31 DiPaolo T, Poyet P, Labrie F (1981) Effect of chronic estradiol and haloperidol treatment on striatal dopamine receptors. Eur J Pharmacol 73: 105-106 DiPaolo T (1994) Modulation of brain dopamine transmission by sex steroids. Rev Neurosci 5: 27-42 Fink G, Sumner BEH, Rosie R, Grace O, Quinn JP (1996) Estrogen control of central neurotransmission: effect on mood, mental state, and memory. Cell Mol Neurobiol 16: 325-344 First MB, Williams JBW, Spitzer RL (1997) DTREE for Windows: the DSM IV Expert computer program. Multi Health Systems Inc, Canada Gattaz WF, Vogel P, Riecher-Rössler A, Soddu G (1994) Influence of the menstrual cycle phase on the therapeutic response in schizophrenia. Biol Psychiatry 35: 137-139 Gilger JW, Geary DC, Eisele LM (1991) Reliability and validity of retrospective self

104

Puberty and Schizophrenia Onset reports of the age of pubertal onset using twin, sibling, and college student data. Adolescence 26: 41-53 Goodman Y, Annadora JB, Cheng B, Mattson MP (1996) Estrogens attenuate and corticosterone exacerbates excitotoxicity, oxidative injury, and amyloid Bpeptide toxicity in hippocampal neurons. J Neurochemistry 66: 1836-1844 Häfner H, Behrens S, De Vry J, Gattaz, WF (1991) Oestradiol enhances the vulnerability threshold for schizophrenia in women by an early effect on dopaminergic neurotransmission. Eur Arch Psychiatr Clin Neurosci 241: 65-68 Häfner H, Riecher-Rössler A, Fatkenheuer B, Maurer K, Meissner S, Löffler W, Patton G (1992) Interview for the retrospective assessment of the onset of schizophrenia (IRAOS). Schizophr Res 6: 209-223 Hallonquist JD, Seeman MV, Lang M, Rector NA (1993) Variation in symptom severity over the menstrual cycle of schizophrenics. Biol Psychiatry 33: 207-209 Kendall RE, Chalmers JC, Platz C (1987) Epidemiology of puerperal psychoses. Br J Psychiatry 150: 662-673 Kirov G, Jones PB, Harvey I, Lewis SW, Toone BK, Rifkin L, Sham P, Murray RM (1996) Do obstetric complications cause the earlier age at onset in male than female schizophrenics? Schizophr Res 20: 117-124 Levesque D and DiPaolo T (1989) Chronic estradiol treatment increases ovariectomized rat striatal D-1 dopamine receptors. Life Sciences 45: 1813-1820 Levesque D, and DiPaolo T (1991) Dopamine receptor reappearance after irreversible receptor blockade: effect of chronic estradiol treatment of ovariectomized rats. Mol Pharmacol 39: 659-665 Lewis SW, Owen MJ, Murray RM (1989) Obstetric complications and schizophrenia: methodology and mechanism in schizophrenia. In: Schulz SC, Tamminga CA (eds) Scientific progress. New York, Oxford University Press Morissette, M and Di Paolo, T (1993) Sex and estrous cycle variations of rat striatal dopamine uptake sites. Neuroendocrinology 58: 16-22 Riecher-Rössler A, Häfner H, Stumbaum M, Maurer K, Schmidt R (1994) Can estradiol modulate schizophrenic symptomatology? Schizophr Bull 20: 203-214 Riecher-Rössler A, Häfner H, Dütsch-Strobel A, Oster M, Stumbaum M, van GulickBailer M, Loffler W (1994) Further evidence for a specific role of estradiol in schizophrenia? Biol Psychiatry 36: 492-494 Salokangas, RKR (1995) Gender and the use of neuroleptics in schizophrenia: further testing of the estrogen hypothesis. Schizophr Res 16: 7-16 Seeman, MV (1983) Interaction of sex, age, and neuroleptic dose. Compr Psychiatry 24: 125-128 Seeman, MV. (1989) Neuroleptic prescription for men and women. Soc Pharmacol 3: 219-236 Seeman MV (1996) The role of estrogen in schizophrenia. J Psychiatry Neurosci 21: 123-127 Toran-Allerand, CD (1984) On the genesis of sexual differentiation of the central nervous system: morphogenetic consequences of steroidal exposure and possible role of alpha-fetoprotein. Progr Brain Res 61: 63-98 Toran-Allerand CD (1996) The estrogen/neurotrophin connection during neural development: is co-localization of estrogen receptors with the neurotrophins and their receptors biologically relevant. Develop Neurosci 18: 36-48

105

Puberty and Schizophrenia Onset Woolley CS, McEwen BS (1992) Estradiol mediates fluctuation in hippocampal synapse density during the estrous cycle in the adult rat. J Neuroscience 12: 2549-2554 Woolley CS and McEwen BS (1994) Estradiol regulates hippocampal dendritic spine density via an N-methyl-D-aspartate receptor-dependent mechanism. J Neuroscience 14: 7680-7687 Woolley CS, Weiland NG, McEwen BS, Schwartzkroin PA (1997) Estradiol increases the sensitivity of hippocampal CA1 pyramidal cells to NMDA receptor mediated synaptic input: correction with dendritic spine density. J Neuroscience 17: 1848-1859

106

5 Clinical Estrogen Trials in Patients with Schizophrenia Jayashri Kulkarni

Historical Background In 1892, Emil Kraepelin postulated links between hormones and dementia praecox and reviewed the endocrine status of his patients. Other early researchers such as Hoskins (1929) studied endocrine changes in schizophrenia in postmortem studies. During this period, the discovery of insulin stimulated further interest in the interaction between behavior and metabolism. Hoskin’s work ultimately discounted the efficacy of treating schizophrenia by insulin-induced hypoglycemia and provided an important early focus on endocrine research in schizophrenia. Between 1940 and 1970, there was considerable interest in the psychoendocrine characterization of schizophrenia. The mimicking of psychotic symptoms by the administration of high doses of steroid hormones by Mason (1975) and the demonstration of elevated levels of the steroid metabolic 17-hydroxycorticosteroid in acutely ill people with schizophrenia by Sachar et al. (1963) represent two important early neuroendocrine studies in schizophrenia. Brambilla and Penati (1978) reviewed the evidence for hypo- and hyperfunction of the adrenal, pituitary, and thyroid glands and of the gonads in patients with schizophrenia. Endocrinopathies were only rarely present, suggesting that endocrine abnormalities in these patients were due to the disease process of schizophrenia rather than being causally related. With increasing clarity of the role of neurotransmitters in the regulation of pituitary hormone release via hypothalamic hormones, neuroendocrine studies in schizophrenia conducted in the early 1980s aimed to use the pituitary gland as the window to the brain, using probes that modified the secretion of anterior pituitary hormones to detect abnormalities in tuberoinfundibular-pituitary func-

107

Jayashri Kulkarni

tion, reflecting similar abnormalities in the mesolimbic system. Biological research in schizophrenia in the 1980s and 1990s largely excluded women from neuroendocrine studies, citing menstrual cycle effects as major confounding factors. More recently, the pioneering research of Häfner (1991), and Seeman and Lang (1990) has brought about a long overdue focus on gender differences in schizophrenia. This, in turn, has led to considerable attention being paid to the role of female hormones or functioning of the hypothalamic – pituitary – gonadal (HPG) axis in patients with schizophrenia, with the resultant “estrogen protection hypothesis” – that is well described in chap. 2 of this book.

Animal Studies The effects of estrogen on the dopamine system are complex and the effects of estrogen on dopaminergic neurotransmission are believed to depend on the dose and length of administration. Studies have demonstrated that, within 24 h of low-dose estradiol being administered to rats, there is a significant decrease in the ratio of high-tolow-affinity agonist states of the striatal dopamine D2 receptors (Gordon et al. 1980; DiPaolo et al. 1982; Joyce et al. 1982); this has been described as an acute hyposensitive state. On the other hand, administration of high doses of estrogen results in the development of striatal dopamine receptor hypersensitivity (Perry et al. 1981; Clopton and Gordon 1983). This can be expressed behaviorally in rats as an increase in dopamine agonist-induced stereotypes or biochemically as an increase in the density of [3H] spiroperidol-binding sites (Gordon et al. 1980; Gordon and Perry 1983). Ferretti et al. (1992) found that estrogen administration had little effect on dopamine D1 receptors, but that D2 receptor density fell in response to low-dose estradiol. Behrens et al. (1992) studied the effects of estradiol and testosterone on cataplexy induced by haloperidol in rats and concluded that estradiol downregulated dopamine transmission. More recently, Fink and colleagues (1998; 1999) have shown that estrogen induces a significant increase in 5-HT2A receptors and the serotonin transporter (SERT) in regions of the rat forebrain that, in humans, are associated with mental state, mood, cognition, memory, emotion, and neuroendocrine control. The precise mechanism of

108

Clinical Estrogen Trials in Patients with Schizophrenia

estrogen-serotonin interaction is not clear. However, the forebrain receives a dense innervation of serotonergic projections from the midbrain raphe neurons, and the 5-HT2A receptors are present at high concentrations in most areas of forebrain, in particular in the anterior and frontal cingulate cortex – regions of the forebrain associated with cognition and emotion. The highest density of these receptors is in laminae IV and VA of these regions of the cortex, where the receptors are located on the apical dendrites of the pyramidal neurons and on glutamatergic interneurons. The net effect of 5-HT stimulation of the 5-HT2A receptors in frontal and cingulate cortex would probably be to increase the rate of firing of the pyramidal neurons, although this might be offset by inhibitory effects mediated by 5-HT1 receptors. Since the functional density of 5-HT receptors is influenced by the concentrations of 5-HT to which they are exposed, the density and degree of activity of serotonin transporters (SERT) may also have a bearing on 5-HT2A receptor activity. However, the estrogen-induced increase in the density of SERT sites occurs in areas of the forebrain (amygdala, lateral septum, and venteromedial nucleus of the hypothalamus) in which estrogen has no apparent effect on the density of the 5-HT2A receptors. The action of estrogen on 5-HT2A receptors and the SERT could be mediated by both genomic and nongenomic mechanisms. Fink et al. (1998) reported that estrogen induced a two- to threefold increase in the amount of 5HT2A receptor messenger RNA in the dorsal raphe nucleus, suggesting that exposure to estrogen stimulated 5HT2A receptor gene transcription. This is in contrast to the finding of an estrogen-induced rise in 5HT2A receptor density without a concordant rise in 5HT2A receptor messenger RNA in cortical neurons. In turn, this is analogous to the action of estradiol on D2 receptor-expressing neurons in the striatum, where there are few estradiol receptors. The action of estradiol in the striatum may be mediated by membrane receptors rather than classical cytoplasmic receptors. Mosselman et al. (1996) cloned a second estrogen receptor and termed it estrogen receptor-β (as distinct from estrogen receptor-α). Shughrue et al. (1997), Fink et al. (1998), and Sumner et al. (1999) have provided data that suggest that any genomic action of estrogen on the 5HT2A receptor and SERT genes may be mediated by estrogen receptor-β.

109

Jayashri Kulkarni

Estrogen, by way of its actions on the 5-HT2A receptor, the SERT, and the D2 receptor, may protect against depressive and psychotic symptoms. This apparent psychoprotective effect of estrogen may have a major biological purpose in that, teleologically, the actions of estrogen may be related to its role as the major endocrine coordinator of events that lead to fertility and reproduction. Estrogen triggers the ovulatory luteinizing hormone-releasing hormone/luteinizing hormone surge which influences mating behavior in nonhuman primates. In order for mating to occur, mood and mental state must be right in both humans and other animals. Hence, estrogen triggers and coordinates several biological events which are necessary for procreation. The significance of these findings for schizophrenia lies in the fact that the 5HT2A receptor is the target for the atypical antipsychotics, such as olanzapine and clozapine, which are particularly effective in the treatment of the negative symptoms of the illness (Fink 1995). Thus, the effect of estrogen on 5-HT2A receptors could be part of the explanation for the findings encompassed in the estrogen protection hypothesis. There is no conflict between the possible involvement of serotonin and the dopamine hypothesis of schizophrenia, since estrogen increased density of both the dopamine D2, and the 5HT2A receptor, such that the two mechanisms could operate in parallel (Fink 1995; Fink et al. 1998).

Clinical Studies Following these epidemiological, clinical, and animal study results, we conducted an open-label pilot study (Kulkarni et al. 1996) in which 11 women of child-bearing age with schizophrenia were given 0.02 mg ethinyl estradiol orally as an adjunct to antipsychotic drug treatment for 8 weeks and their progress compared with a similar group who received antipsychotic drugs only. The group receiving estrogen made a significantly more rapid recovery from acute psychotic symptoms and also reported improvement in their general health status. Subsequent to this, we conducted a dose-finding study for the optimal use of estradiol in women with schizophrenia (Kulkarni et al. 2001). This was a three-arm, double-blind, placebo-controlled, 28day study in which 12 women received 50 µg transdermal estradiol

110

Clinical Estrogen Trials in Patients with Schizophrenia

plus standardized antipsychotic drug, 12 women received 100 µg transdermal estradiol plus standardized antipsychotic drug, and 12 women received placebo plus standardized antipsychotic drug. All of the women had a diagnosis of DSM-IV schizophrenia. Women were excluded from the trial if they had any known endocrine abnormalities, were pregnant or lactating, were currently taking synthetic steroids, including the oral contraceptive pill, or were using illicit drugs. Each subject was enrolled in the trial for 28 days (one menstrual cycle) and received a baseline psychopathology and hormone assessment followed by assessments at days 4, 7, 14, 21, and 28. At each assessment, psychopathology was evaluated using the Positive and Negative Syndrome Scale (PANSS; Kay et al. 1987). The PANSS is comprised of three subscales – the positive symptom subscale, the negative symptom subscale, and the general symptom subscale. Hormone assays for serum estrogen (E2), progesterone (Prog), prolactin (PRL), luteinizing hormone (LH), follicle-stimulating hormone (FSH), and testosterone (Test) were performed. Separate radioimmuno-assay tests were conducted for each individual sample. A Menstrual Cycle Interview (MCI; Kulkarni et al. 1996b) was used to stage the patients’ menstrual cycle phase. All patients were randomized into the active estradiol skin patch treatment groups or an identical placebo patch group. For transdermal delivery adhesive skin patches were utilized that contained either 4 mg estradiol per patch, with a release rate of 50 µg per 24 h, or 8-mg estradiol patches, with a release rate of 100 µg estradiol per 24 h. Both estrogen and placebo patches were changed every 4 days. Placebo patches were also adhesive but had no active substance. All patients received antipsychotic drug treatment, which was administered according to a protocol that indicated doses of between 3 and 6 mg per day of risperidone. The dose of risperidone was dependent on the patient’s clinical state.

Results The group receiving adjunctive 100 µg estradiol showed greatest improvement across the study compared to the other two groups, although the 50-µg adjunct group showed more improvement than the placebo adjunct group. Improvement scores differed significantly

111

Jayashri Kulkarni

between the three groups (p < 0.005). The 100-µg group improved more than the placebo group (p = 0.001) and the 50-µg group (p = 0.001), whereas the latter two groups did not differ significantly. The mean change from baseline for the positive symptom subscale of the PANSS differed between the three groups (p < 0.005). The 100-µg group improved more than the placebo group (p = 0.002) and the 50-µg group (p = 0.001), whereas the latter two groups did not differ significantly from each other. The mean change from baseline for the negative symptom subscale of the PANSS showed no statistically significant difference between the three groups, although there was a significant difference between the 100-µg estrogen and placebo groups (p = 0.039). The mean change from baseline for the general symptom subscale of the PANSS revealed a significant difference between the groups (p = 0.001). The 100-µg group improved more than the 50-µg group (p = 0.001) and the placebo group (p = 0.002), whereas the 50-µg and placebo groups did not differ significantly from each other. Hormone data analysis showed that the 100-µg group had significantly lower mean LH levels than the 50-µg group (p = 0.027), suggesting that 100 µg adjunctive estradiol had greater impact on the pituitary and perhaps, therefore, a direct neuroleptic effect on the dopamine and serotonin systems.

Follow-up of the 100-µg Estradiol Group To further investigate the significant improvement in psychotic symptoms with the addition of 100 µg of estradiol to antipsychotic drug treatment, we studied 36 women with DSM-IV schizophrenia further. In this 28-day, double-blind study, all women were randomized to the active 100-µg estradiol skin patch treatment group or an identical placebo patch group. All patients received antipsychotic drug treatment of risperidone, according to a standardized protocol. Psychopathology was assessed by the PANSS rating scale and serum hormone levels of estrogen, progesterone, prolactin, LH, and FSH were measured. A battery of cognitive tests was also administered.

112

Clinical Estrogen Trials in Patients with Schizophrenia

Results There were no significant differences between the 100-µg and placebo adjunct groups in mean age, menstrual cycle phase, race, illness duration, or antipsychotic drug dose. Mean Baseline PANSS for the two groups were: 86.3 +/– 27.7 for the 100-µg estradiol group and 73.3 +/– 12.3 for the placebo group. This difference is statistically significant (p < 0.05). To allow for correction of this difference in the baseline scores, the change across time was calculated. Change from baseline in the total PANSS and the positive, negative, and general subscales across time are shown (Figs. 1–4). The changes in hormone data across 28 days for the two groups revealed that the estrogen group had significantly higher mean estrogen levels and higher testosterone and lower LH levels. There were no differences in progesterone, FSH, prolactin, or testosterone levels between the estrogen adjunct and placebo adjunct groups.

Conclusion

Mean change from baseline

The results to date supply further evidence suggesting that the addition of 100 µg estradiol transdermally provides greater im-

Fig. 1. Mean change from baseline in total PANSS scores for both groups

113

Mean change from baseline

Jayashri Kulkarni

Fig. 2. Mean change from baseline in PANSS Positive Subscale scores for both groups. The group receiving estrogen (100 µg) had a more significant decrease in psychotic symptoms as measured by the PANSS rating scale

Mean change from baseline

provement in the treatment of psychotic symptoms in women with DSM-IV schizophrenia than standardized antipsychotic drug treatment alone. The impact of administering 100 µg of estradiol transdermally, as indicated by its effect on the pituitary measured by LH assay, suggests that this dose and type of unconjugated estrogen

Fig. 3. Mean change from baseline in PANSS Negative Subscale scores for both groups

114

Mean change from baseline

Clinical Estrogen Trials in Patients with Schizophrenia

Fig. 4. Mean change from baseline in PANSS General Subscale scores for both groups

affects CNS neurotransmitter systems positively. This is consistent with the “estrogen hypothesis” as formulated by Häfner (1991).

Clinical Adjunctive Estrogens Trial in Men with Schizophrenia In one of our studies (Kulkarni et al. 1995), we measured baseline gonadotropin and sex steroid hormones in men and women suffering from schizophrenia compared with age- and sex-matched controls. A major finding was that 19 men with schizophrenia had average testosterone levels of 26.6 mg compared with 17.7 mg in 18 healthy age-matched male controls. The male patients had significantly higher testosterone LH and FSH levels than controls. In effect, the male patients had a “puberty-like” gonadal axis profile. However, the patients had normal secondary sexual characteristics and no endocrine-related illnesses. The patients had all been off neuroleptics for at least 3 months at the time of testing. We also correlated testosterone levels with psychopathology scores and found a significant, positive correlation between testosterone levels and increasing SAPS and SANS score. The exact opposite results were found in the women we studied. Female patients had lower estrogen levels and higher LH

115

Jayashri Kulkarni

levels than age- and cycle phase-matched female controls. In effect, the female patients had a “menopause-like” gonadal axis profile – even though the average age for the patient group was 26.4 years and they were not menopausal. Estrogen levels correlated inversely with psychopathology scores. According to these data from clinical and preclinical studies, estrogen appears to have a potentially positive effect on psychotic symptoms, most likely due to its dopamine-modulating effect. In view of the positive response found in our adjunctive estrogen trials in women, we believe that adding short-term, low-dose estradiol to standardized antipsychotic drug treatment in men with schizophrenia could improve their response to neuroleptics.

Results of a Pilot Study In a pilot study conducted in 1999, 2 mg estradiol valerate was given as an adjunct to six men who were receiving antipsychotic drugs and five men received oral placebo plus their standard antipsychotic medication. Oral estrogen, rather than transdermal estrogen was used to ensure compliance in acutely psychotic men. Side effects of estrogen therapy in men such as gynecomastia, decreased libido, and fluid retention are not reported in studies using less than 1.25 mg estrogen per day for less than 4 weeks. Psychopathology was assessed using PANSS and the Brief Psychiatric Rating Scale (BPRS). High scores on both these scales indicate more severe symptoms. Both groups commenced antipsychotic drugs plus estrogen or placebo on day 1 of the trial. However, the dose and type of antipsychotic drug was not standardized in this small pilot study.

Summary of Pilot Study Results The t-tests results from the pilot study show no significant difference in psychopathology scores between the two groups at the start of the study. At day 5, the adjunctive estrogen group had a significantly lower positive PANSS, BPRS, score (p < 0.05) than the placebo adjunct group. The estrogen group had a significantly higher negative PANSS score (p < 0.05). There were significant differences in mean hormone levels between the two groups at the start of the study. At day 7, the

116

Clinical Estrogen Trials in Patients with Schizophrenia

estrogen group had significantly higher mean estrogen levels, higher LH levels, and higher prolactin levels. There was no difference in testosterone levels of FSH at day 7 between the groups.

Current Study We are currently investigating the effect of adding 2 mg oral estradiol velate to standardized antipsychotic drug treatment in a group of men with schizophrenia for a 2-week, double-blind trial. A target number of 60 patients will be recruited over a 3-year period, with equal numbers being allocated to either the adjunctive estradiol or adjunctive place group, for 2 weeks.

Results A total of 16 men with DSM-IV schizophrenia have been recruited since May 2001. Eight received 2 mg adjunctive estradiol orally and eight received adjunctive oral placebo. All patients received standardized antipsychotic drugs. Demographic data revealed no differences between the two groups in terms of age, race, diagnosis, illness duration, or antipsychotic drug treatment. The mean PANSS data for the estrogen and placebo groups as shown in Table 1 reveals that the estrogen adjunct group was not as well at the start of the trial. To compare the improvement in the two groups across the 14-day trial, changes from baseline measures were calculated (Table 2). The results show that the estrogen adjunct group made a more significant improvement over the 14-day trial as shown by the decrease in PANSS scores from baseline. In particular the most significant differences were seen in the PANSS general subscale symptoms (see Figs. 5–8).

Hormone Levels There were significant changes between the two groups: the estrogen group had significantly higher day 14 estrogen levels (p = .002) and lower testosterone levels (p = 0.04). The change in LH levels from baseline measurements was not significant across time.

117

Jayashri Kulkarni Table 1. Mean PANSS data for the estrogen and placebo groups E2 group

Total mean PANSS ± SD Total mean positive Sx Total mean negative Sx Total mean general Sx

Baseline

Day 7

75.4±9.6

Placebo group Day 14

Baseline

Day 7

Day 14

68.0±13.8 63.0±9.0

67.6±3.6

64.6±8.4

63.2±7.1

21.0±5.7

18.3±6.4

16.3±5.1

21.7±3.5

19.0±3.1

19.2±3.1

16.6±4.8

15.5±2.4

14.3±2.0

13.2±2.9

12.4±2.3

13.0±1.8

37.7±4.3

34.3±6.7

30.3±4.1

32.6±1.7

32.0±5.1

31.5±3.6

(PANSS scores in bold denote significant differences p < 0.05)

Table 2. Mean PANSS change from baseline scores for estrogen and placebo groups

PANSS change from baseline to day 14 Positive symptom change 0-day 14 Negative symptom change 0 –day 14 General symptom change 0 – day 14

E2 group

Placebo group

p value

-12.38+/6.4

-4.38+/6.1

0.04

-3.6+/2.6

-2.5+/4.4

NS

-2.3+/5.1

-0.75+/4.1

NS

-6.7+/4.3

-1.1+/4.1

0.02

This study is in progress and is also a dose-finding study. Patients are closely monitored and no serious adverse effects have been noted:

Conclusions to Date The use of estrogen as a potential treatment in schizophrenia opens up exciting new avenues of preventive and acute treatment of schizophrenia in both men and women. The rapid development of new estrogen compounds, selective estrogen receptor modulators – or “brain estrogens,” has further expanded the area of hormone treatment. Studying the mechanisms by which estrogen potentially

118

Clinical Estrogen Trials in Patients with Schizophrenia

Mean change from baseline Total PANSS

Mean change from baseline

Treatment 2 mg Estradiol N=8 Placebo N=8

Time (days)

Fig. 5. Mean change from baseline for total PANSS scores for the two groups

Mean change from baseline PANSS Positive Subscale

Treatment 2 mg Estradiol N=8

Mean change from baseline

Placebo N=8

Time (days)

Fig. 6. Mean change from baseline for PANSS positive subscale for the two groups

119

Jayashri Kulkarni

Mean change from baseline

Mean change from baseline

PANSS Negative Subscale

Treatment 2 mg Estradiol N=8 Placebo N=8

Time (days)

Fig. 7. Mean change from baseline for PANSS negative subscale for the two groups

Mean change from baseline

Mean change from baseline

PANSS General Subscale

Treatment 2 mg Estradiol N=8 Placebo N=8

Time (days)

Fig. 8. Mean change from baseline for PANSS general subscale for the two groups

120

Clinical Estrogen Trials in Patients with Schizophrenia

represents an option for treating psychotic symptoms may also promote our understanding of the etiological aspects of schizophrenia, in particular, the reasons why schizophrenia is a postpubertal disease.

References Behrens S, Häfner H, De Vry J, Gattaz WF (1992) Estradiol attenuates dopaminemedicated behavior in rats. Implications for sex differences in schizophrenia. Schizophr Res 6: 114 Brambilla F, Penati G (1978) Perspectives in endocrine psychobiology. In: Brambilla F and Bridges P (ed). Perspectives in endocrine psychobiology. Witey, London, 309-422 DiPaolo T, Payet P, Labrie F (1982) Effect of prolactin and estradiol on rat striated dopamine receptors. Life Sci 31: 2921-2929 Ferretti C, Blengio M, Vigna I, Ghia P, Gerazzani E (1992) Effects of estradiol on the ontogenesis of striatial dopamine D1 and D2 receptor sites in male and female rats. Brain Res 571: 212-217 Fink G (1995) The psychoprotective action of estrogen is mediated by central serotonergic as well as dopaminergic mechanisms. In: Takada A and Curzon G (eds). Serotonin in the central nervous system and periphery. Elsevier Science, Amsterdam, 175-187 Fink G, Sumner B, McQueen JK, Wilson H, Rose R (1998) Sex steriod control of mood, mental state and memory. Clin Exp Pharmacol Physiol 25: 764-765 Fink G, Sumner B, Rosie R, Wilson H, McQueen J (1999) Androgen actions on central serotonin neurotransmission: relevance for mood, mental state and memory. Behav Brain Res 105: 53-68 Gordon HH, Borison RL, Diamond BI (1980) Modulation of dopamine receptor sensitivity by estrogen. Biol Psychiatry 15: 389-396 Gordon JH, Perry KO (1983) Pre- and postsynaptic neurochemical alteration following estrogen-induced striatal dopamine hypo-and hypersensitivity. Brain Res Bull 10: 425-428 Häfner H (1991) The epidemiology of beginning schizophrenia. Presented at the WPA Section of Epidemiology and Community Psychiatry Symposium, June 14-16, Oslo Hoskins RG (1920) Endocrine factors in dementia praecox. N Engl J Med 200: 361 Joyce JN, Smith RL, Van Hartesveldt C (1982) Estradiol suppresses then enhances intracaudate dopamine-induced contralateral deviation. Eur J Pharm 81: 117-122 Kraepelin E (1982) Psychiatrie, vol 3, part 2, 8th ed. Translated (1919). Dementia praecox and paraphrenia. Livingstone, Edinburgh Kulkarni J, de Castella A, Riedel A, Taffe J, Fitzgerald P, Burger H (2001) Estrogen – a potential new treatment in schizophrenia. Schizophr Res 48: 137-144 Kulkarni J, Smith D, McKenzie D, Hill C, Keks N, Singh B, Copolov D (1995) Donadotrophin response to naloxone challenge in female and male psychotic patients. Biol Psychiatry 38: 701-703

121

Clinical Estrogen Trials in Patients with Schizophrenia Kulkarni J, de Castella A, Smith D, Taffe J, Keks N, Copolov D (1996a) A clinical trial of the effects of estrogen in acutely psychotic women. Schizoph Res 20: 247-252 Kulkarni J, Gostt K, de Castella A (1996b) The menstrual cycle in women with schizophrenia. Schizophr Res 18: 254 Mason JW (1975) Emotion as reflected in patterns of endocrine investigation. In: Levi L (ed) Emotions: their parameters and measurement. Raven Press, New York, 143-181 Mosselman S, Polman J, Dukema R (1996) ER beta: identification and characterization of a novel human estrogen receptor. FEBS Lett 392: 49-53 Perry KO, Diamond BI, Fields JZ, Gordon JH (1981) Hypophysectomy induced hypersensitivity to dopamine: antagonism by estrogen. Brain Res 226: 211-219 Sachar EJ, Mason JW, Kolmer HS, Arfess KL(1963) Psychoendocrine aspects of acute schizophrenic reactions. Psychosom Med 25: 510 Seeman MV, Lang M (1990) The role of estrogens in schizophrenia gender differences. Schizophr Bull 16: 185-195 Shughrue PJ, Lane MV, Merchenthaler I (1997) Comparative distribution of estrogen receptor and X and B MRNA in the rat central nervous system. J Compr Neurol 388: 507-525 Sumner BEH, Grant DE, Rosie R, Hegele-Hartung Ch, Fritzemeier KH, Fink G (1999) Effects of tamoxifen on serotonin transporter and 5-hydroxytryptamine 2A receptor binding sites and mRNA levels in the brain of ovariectomized rats with or without acute estradiol replacement. Mol Brain Res 73: 119-128

122

6 Hypoestrogenism and Estrogen Replacement Therapy in Women Suffering from Schizophrenia Niels Bergemann, Christoph Mundt, Peter Parzer, Benno Runnebaum and Franz Resch

Estrogens have cerebral effects in addition to the genuinely hormonal functions in the gonadal axis. In the past it could be demonstrated that estrogens have an influence on major neurotransmitter systems and brain regions affecting cognitive, emotional, and vegetative functions (e.g., Fink et al. 1996; 1998; Halbreich 1995; Morisette and DiPaolo 1993; DiPaolo 1994; Gordon et al. 1980). Estrogens also exert neurotrophic and neuroprotective effects which are mediated by nongenomic as well as by direct and indirect genomic pathways (for review, see McEwen and Alves 1999; Lee and McEwen 2001; GarciaSegura 2001; Behl 2001). How estrogen influences the course of various diseases of the nervous system has been the focus of research in the neural sciences in the past few years. These diseases include stroke, Parkinson’s disease, Alzheimer’s disease, depression, and schizophrenia. Generally, gender differences in the incidence and in recovery from neurological damage and mental disorders were an important starting point of research in these fields. In schizophrenia research, the estrogen protection hypothesis has been investigated over the last 20 years. It assumes a protective effect of estrogen in women vulnerable to schizophrenia. Evidence for this hypothesis is based on epidemiologic, neurochemical, animal, and clinical data (cf. chap. 2 and 3 of this volume by RiecherRössler and/or Häfner; Seeman 1996). As early as 1909, Kraepelin reported that first-time hospitalization in men with schizophrenia occurs earlier than in women. However, the peak age of onset of schizophrenia is significantly later for women than for men, regardless of whether one investigates age at first hospitalization, first treatment, or age when psychotic symptoms are first noted (DeLisi et al. 1989; Seeman 1982; 1985; Tsuang

123

Niels Bergemann et al.

et al. 1976; Lewine 1981; Lewine et al. 1981; Forrest and Hay 1971; Loranger 1984; Sartorius et al. 1986; Häfner et al. 1991, 1992, 1998; for review see Angermeyer and Kühn 1988). In men, the age at onset is the mid-20s, and in women the late 20s. In contrast to men, the curve of age at onset in women shows not only a smaller peak in young age than in men, but also a second peak during the 40s (cf. chap. 3 of this volume by Häfner). Thus, it seems plausible that the onset of schizophrenia is determined by changes in steroid hormone production that occur at different times in men and women. Between menarche and menopause women are more likely to have a higher vulnerability threshold to schizophrenia than men because of the protective effect of higher plasma concentrations of estradiol. Since within families men and women do not show differences in age at onset, a genetically determined vulnerability may antagonize the onset-delaying effect of estradiol in women and, to a slightly lesser extent, the degree of pre- and peri-natal brain injury (Albus and Maier 1995; Könnecke et al. 2000). Other reports also indicate that women have a significantly less severe course of schizophrenia than men (Sartorius et al. 1978; Solokangas 1983) which can be attributed to the fact that their premenopausal estrogen plasma level may modulate the illness course and serve as a partial endogenous antipsychotic agent. Thus, young women can be maintained on lower doses of antipsychotics than men, but after age 40, women require higher doses than men. This coincides with menopause and further suggests estrogen mediation (Seeman 1983; cf. DeLisi et al. 1989). Several clinical studies, case reports, or case series show a correlation between low estrogen plasma concentrations and an increase in the risk for schizophrenic symptoms occurring in women: postnatally or after the menopause, physiologically low plasma concentrations of estrogen prevail and, at the same time, there is an increased risk for the exacerbation of schizophrenia (Nott 1982; Kendell et al. 1987; Brockington et al. 1988; Seeman 1997). Further studies have shown more severe psychotic symptoms and a higher risk for the exacerbation of schizophrenic symptoms or hospitalization in low estradiol phases of the menstrual cycle than in high estradiol phases (e.g., Swanson et al. 1964; Glick and Stewart 1980; Hallonquist et al. 1993; Riecher-Rössler et al. 1994; Hendrick et al. 1996; Harris 1997; for review see, e.g., Bergemann et al. 2002). Gattaz et al. (1994) showed a significant correlation between the estradiol

124

Hypoestrogenism and Estrogen Replacement Therapy

plasma concentration and the therapeutically required dose of antipsychotics. It is of particular interest, that the observation of a possible relationship between phases of the menstrual cycle and psychotic disorders dates back many years and was discussed extensively as early as in the nineteenth century. In the second half of the nineteenth and the beginning of the twentieth century, a number of publications reported such a relationship (e.g., Häffner 1912; Mayer 1872; Ross 1909; Schroeter 1874; v. Krafft-Ebing 1878; Jolly 1915; Schaefer 1894). In our own study (Bergemann et al. 2002) the hypothesis of a perimenstrual increase in hospital admissions of women suffering from an exacerbation of schizophrenia was tested in two samples of premenopausal women (sample 1: n = 115; sample 2: n = 170). In both samples there was a significant increase in admissions in the perimenstrual phase – three days before and three days after the first day of the menses (sample 1: p = 0.002, sample 2: p = 0.028, binomial test). Of the patients 37.4% were admitted during the perimenstrual phase of the cycle in sample 1, 31.8% in sample 2. Regarding age, age at onset, and duration of illness, there was no difference between the group admitted to the hospital during the perimenstrual phase and the group admitted during the rest of the menstrual cycle in either of the samples.

Hypoestrogenism in Women with Schizophrenia Although numerous investigations could prove that estrogen has a protective effect in schizophrenia, little work has been done to study the estrogen deficiency syndrome or the hypoestrogenism hypothesis in schizophrenia (Riecher-Rössler 2003; cf. chap. 2 of this volume by Riecher-Rössler), which postulates an association between schizophrenia and a chronic estrogen deficiency in schizophrenic women. Low estrogen levels leading to an elevated rate of menstrual dysfunction, such as amenorrhea and irregular menstruation, have been described in schizophrenic women. Of particular interest are studies on hypoestrogenism carried out in the pre-antipsychotic era. They demonstrated different types of menstrual cycle disorders in schizophrenic psychoses, anomalies of secondary sexual features, and an increased rate of virilization (e.g.,

125

Niels Bergemann et al.

Häffner 1912; for review see Bleuler 1943, Reiss 1958, Diczfalusy and Lauritzen 1961). As early as 1869, Mayer described the interaction between gynecological disorders and/or menstrual cycle disturbances and mental disorders in his frequently cited monograph „Die Beziehungen der krankhaften Zustände in den Sexualorganen des Weibes zu Geistesstörungen” [The relation between diseases of the sexual organs in women and mental illness]. Various menstrual cycle disturbances, anomalies of the genital organs, and masculinization were seen in conjunction with psychotic disorders; thus, genital abnormalities and menstrual cycle disturbances were considered possible causes of psychoses (v. Krafft-Ebing 1903; König and Linzenmeier 1913). The terms uterine or amenorrheal insanity illustrate the hypothesis that “insanity in some few cases actually results de novo from this [amenorrhea] as an exciting or predisposing cause” (Clouston 1906). Based on these assumptions, ovariectomies were even carried out as therapeutic interventions in women with schizophrenia (Hegar 1884; Battey 1877; for review see Wells 1886; Burger 1984). A dysfunction of the gonads was postulated to be the cause of psychotic disorders before the endocrine processes responsible for the menstrual cycle were discovered. Clinical observations led Kraepelin (1909), and Kretschmer (1921) to assume a relation between schizophrenia and a “disturbance in state of sexual hormones”, above all in the sense of a “hypofunction of the gonads” and/or a so-called “hypoestrogenism.” Thus, in later years, a hypofunction of the gonads with subsequent hypoestrogenism was postulated in women with schizophrenia. It is of particular interest that the hypothesis of an estrogen deficiency syndrome was formulated before the introduction of antipsychotics (for review, see Diczfalusy and Lauritzen 1961). Since then, hypoestrogenism and subsequent, irregular menstrual cycles and amenorrhea have usually been attributed to antipsychotic-induced hyperprolactinemia, mediated by hypothalamic-pituitary-gonadal feedback mechanisms (Smith et al. 2002), although irregular menstruation had already been observed in schizophrenic women before antipsychotics were introduced in the therapy of patients suffering from psychoses. Diczfalusy and Lauritzen (1961) reviewed the few studies on estrogen plasma concentrations directly measured in urine and/or blood in schizophrenic women in the preneuroleptic era (between 1933 and 1955). In seven out of eight studies, low levels of estrogen were detected (Georgi and Fels 1933; Saethre 1933; Oesterreicher 1934; Kafka 1935; Sears et al. 1937;

126

Hypoestrogenism and Estrogen Replacement Therapy

Nilsson 1939; Lansbury and Hughes 1939). Only one study showed increased estrogen levels (Sacerdoti and Carry 1955). However, these findings have to be interpreted carefully as the laboratory methods applied were are not comparable to currently available analytic methods. Furthermore, study samples of schizophrenic patients were small. In a more recent study including 32 acutely ill premenopausal women with schizophrenia and a history of regular menstrual cycles, Riecher-Rössler et al. (1994) reported lower than normal estradiol plasma concentrations in the course of the menstrual cycle with narrow fluctuations. The estradiol values at admission ranged from 45 to 502 pmol/l (12–137 pg/ml) with a mean of 176.5 pmol/l (48 pg/ml), and only 4 patients reached the lower normal range of the preovulatory phase of 550 pmol/l (150 pg/ml) throughout the entire cycle. Interestingly, even these narrow fluctuations in the estradiol plasma levels correlated with psychopathology scores measured by various rating scales. In addition, Choi et al. (2001) found extremely low serum estrogen levels in a study on premenstrual symptoms in 24 chronically ill patients suffering from schizophrenia in both the premenstrual and postmenstrual phases of the cycle. In these patients, the values did not reach the lower limit of normal levels in either of the phases, with a mean estrogen level of 13.2 pg/ml in the premenstrual phase and a level of 38.8 pg/ml in the postmenstrual phase. Huber et al. (2001) compared the estradiol levels on admission of 32 patients with various psychotic disorders to a control group of 9 healthy controls and 11 patient controls with various psychiatric disorders, most of them with a diagnosis of affective disorder. About half of the patients were taking antipsychotics of some kind. Upon admission, the blood levels of estradiol in the study group ranged from < 7 pg/ml to 185.3 pg/ml (median 41.1 pg/ml); only one patient had an estradiol concentration above 150 pg/ml, indicated as the lower level of ovulation. Both control groups had higher median estradiol blood levels, with values of 67.7 pg/ml in the healthy controls and 55.2 pg/ml in the patient controls, but only the median of the healthy control group was statistically significantly higher (p = 0.017). More recently, Canuso et al. (2002) reported a study on 16 premenopausal women with schizophrenia and schizoaffective disorder treated with an antipsychotic with either a prolactin-elevating or

127

Niels Bergemann et al.

prolactin-sparing potential. Similar rates of menstrual dysfunction and ovarian hormone values were observed between hyperprolactinemic and normoprolactinemic subjects, and, irrespective of antipsychotic type or prolactin status, most subjects had peak estradiol levels below normal reference values for the periovulatory phase of the menstrual cycle; 11 subjects had peak estradiol levels beneath the lower limit of normal for this phase. Hoff et al. (2001) conducted a study including 22 women with schizophrenia and investigated the association between blood levels of estrogen and neuropsychological performance. In this sample of inpatients the authors found not only a correlation between average estrogen level and cognitive function but also, and more relevant for the present context, abnormally low estrogen levels compared to reference values: only one patient showed a value over 550 pmol/l (150 pg/ml). In our own study (Bergemann et al. 2004a; cf. Bergemann et al. 1996) the serum levels of 17β-estradiol, prolactin, luteinizing hormone (LH), follicle-stimulating hormone (FSH), progesterone, and testosterone were investigated during the entire course of the menstrual cycle. In 75 premenopausal, menstruating women with schizophrenia diagnosed according to DSM-IV and ICD-10 and consecutively admitted to the psychiatric unit at the Department of Psychiatry at the University of Heidelberg, hormone levels were assessed three times during the first menstrual cycle after admission to the hospital: at the follicular (proliferative) phase at day 2–4 (t1), at the periovulatory phase at day 10–12 (t2), and at the luteal phase at day 20–22 (t3). This regimen was chosen because a single measurement of hormone levels at any point during the menstrual cycle would not have provided any evidence – for example, results from one blood analysis at admission would be questionable because most schizophrenic women are hospitalized perimenstrually, i.e., during a phase of the cycle in which estradiol levels are very low anyway (Bergemann et al. 2002; Huber et al. 2001) and this value cannot be considered indicative of overall estradiol levels throughout the entire cycle. The results were compared to normal reference values by the Department of Obstetrics and Gynecology at the University of Heidelberg (Rabe and Runnebaum 1994; Klinga 1994). All participants were Caucasian. The mean age was 32.9 ± 6.8 years (range: 18–46 years), and the mean duration of illness was 7.5 ± 7.0 years. The levels of the hormones investigated in this study during the

128

Hypoestrogenism and Estrogen Replacement Therapy

three phases of the menstrual cycle are shown in Fig.1. The results are presented as box plots (Tukey 1977) showing the median, interquartile range (“boxes”), and the total deviation of the values; extreme values are shown separately.

estradiol pg/ml

300

200

100

0 2-4

10-12 day of menstrual cycle

20-22

LH mE/ml

80

60

40

20

0 2-4

10-12 day of menstrual cycle

20-22

Fig. 1. Course of 17β-estradiol, LH, progesterone, prolactin, FSH, and testosterone during the phases of the menstrual cycle in women with schizophrenia (n = 75; box plots, extreme values as circles) (cf. Bergemann et al. 2004a)

129

Niels Bergemann et al.

Fig. 1. (continued)

We could show that the serum levels of estradiol were generally reduced during the entire menstrual cycle although there were significant changes in the estradiol serum levels between the three phases

130

Hypoestrogenism and Estrogen Replacement Therapy

Fig. 1. (continued)

of the menstrual cycle investigated (analysis of variance for repeated measurements with Huyhn-Feldt correction). In accordance with the normal reference range one would expect estradiol plasma levels of

131

Niels Bergemann et al.

between 40 pg/ml and a maximum of 100 pg/ml in the follicular phase (cycle day 2–4) and values of above at least 150 pg/ml and up to 350 pg/ml during the periovulatory phase (cycle day 10–12). The levels of LH were also low throughout the entire cycle. As would be expected, no significant increase in LH could be seen during the periovulatory phase. With a mean progesterone level of 1.56 ± 3.10 ng/ml (normal reference value: >10 ng/ml) in the luteal phase, it was assumed that ovulation had not occurred and/or that the follicles had not matured sufficiently. The prolactin levels were significantly elevated throughout the entire menstrual cycle, as could be expected under medication with predominantly conventional antipsychotics, and no difference between the mean values and deviation between the three phases could be detected (Bergemann et al. 2004a). Estradiol values of < 100 pg/ml during the periovulatory phase (day 10–12 of the menstrual cycle) and < 30 pg/ml during the follicular phase (day 2–4) were considered hypoestrogenemic. In accordance with this strict definition we found hypoestrogenism in 57.3% of the patients. To rule out a possible effect of hyperprolactinemia on the gonadal axis and the subsequent effect on the estrogen level due to the intake of conventional antipsychotics, serum levels of estradiol in women treated with atypical antipsychotics known to induce only a mild or no increase in prolactin were compared to those in women treated with conventional antipsychotics. For this additional analysis, subgroups of patients were selected from the sample described above who had been treated continuously for at least 12 weeks with clozapine (n = 11) or olanzapine (n = 7). These subgroups were compared to 31 patients treated continuously for at least 12 weeks with one conventional antipsychotic. For each group the hormone values of the periovulatory phase (cycle day 10–12) were taken for comparison. As expected, the mean prolactin serum level in patients receiving conventional antipsychotics (M ± SD: 1,600 ± 1,278 mU/l) was much higher than the upper range of the normal reference values of 410 mU/l. As expected, there is a statistically significant difference in prolactin serum levels between the group receiving conventional antipsychotics and patients receiving the atypical antipsychotic clozapine (M±SD: 467 ± 170 mU/l; p < 0.001) or olanzapine (M ± SD: 654 ± 456 mU/l; p = 0.005). The difference between the prolactin

132

Hypoestrogenism and Estrogen Replacement Therapy

levels in the patients who received clozapine vs. olanzapine was not statistically significant. However, in contrast to the results for prolactin, the mean serum estradiol levels were far below 100 pg/ml in all three groups (Bergemann et al. 2004a). Thus, there is some evidence for a sex hormone dysregulation and hypoestrogenism in schizophrenia. Furthermore, the fact that hypoestrogenism in schizophrenic women occurs with and without antipsychotic-induced hyperprolactinemia supports the hypothesis of primary hypoestrogenism.

Estrogen Replacement Therapy in Women with Schizophrenia The results presented here are not only of particular interest from a theoretical point of view, but also for their therapeutic implications. As already reported, empirical evidence was found in previous studies for a protective effect of estrogen in schizophrenia, which also suggests its use in treatment and prophylaxis. An estrogen replacement therapy could be particularly effective in those women with schizophrenia suffering from hypoestrogenemia. However, only very few studies addressing an estrogen medication in schizophrenia at all have been conducted to date. Mall (1958; 1960) was the first to apply estrogen treatment in women suffering from schizophrenia. He distinguished between premenstrually and post-menstrually admitted patients and describes different psychopathological types of illness. On the basis of quantitative analyses of estrogen in the urine, he also distinguished between “hypofollicular” and “hyperfollicular” psychoses, where most of the “post-menstrual psychoses” belong to the group of “hypofollicular” psychoses. For this group he recommends estrogen replacement therapy. Unfortunately, the author gives only little information on the results of his study. In recent times Kulkarni (1996; 2001, cf. chap. 5 of this volume) showed a beneficial effect of estradiol on the course of illness in the acute phase in various experimental trials. Lindamer et al. (2001) showed in a post hoc analysis that hormone replacement therapy with estrogen in conjunction with antipsychotic medication in postmenopausal women with schizophrenia helped to reduce negative symptoms. In addition, the users of estrogen replacement

133

Niels Bergemann et al.

therapy required a lower dose of antipsychotic medication than controls. Our own study (Bergemann et al. 2004b) was the first clinical intervention trial to empirically investigate the validity of the estrogen hypothesis in the context of relapse prevention in schizophrenic women. The expected therapeutic effect of estrogen in the treatment of schizophrenia was tested under real-life conditions in order to obtain further clinical evidence for the estrogen hypothesis in schizophrenia and to evaluate the practical importance of an adjuvant estrogen replacement therapy in hypoestrogenemic women with schizophrenia. In particular, the question was addressed as to whether antipsychotics plus estrogen (and/or a combination of estrogen and gestagen) is superior to antipsychotic monotherapy in preventing relapse. A more favorable course of illness or a lower antipsychotic drug requirement as well as a better tolerance were expected from an adjuvant therapy with estrogen versus a antipsychotic monotherapy (Bergemann et al. 2004b). A multi-center, randomized, placebo-controlled, double-blind, cross-over study based on an A-B-A-B- (and/or B-A-B-A) design was applied (Fig. 2, Bergemann et al. 2004b). Forty-six schizophrenic women with hypoestrogenism hospitalized for the first time or repeatedly were included in the study. Their average age was 37.9 (SD = 9.8, range 21–60 years) and they had been suffering from schizophrenia for 8.4 years (SD = 7.4 years; ICD-10 F20.0, F20.1, F20.5). Here, too, hypoestrogenism was defined as estradiol serum level < 30 pg/ml on day 2-4 of the menstrual cycle and estradiol serum level < 100 pg/ml on day 10–12. Women with relevant co-morbidity or who were pregnant or lactating were excluded from the trial (cf. Bergemann et al. 2004b). The estrogen replacement was administered as an adjuvant to routine relapse prevention with antipsychotics. A three-phase estrogen-gestagen combination was used (17β-estradiol 4 mg in the follicular phase, plus norethisteroneacetate 1 mg in the periovulatory phase and 17β-estradiol 1 mg in the luteal phase; Trisequens forte®). If any side effects developed, the dose could be halved to 2 mg in the follicular and periovulatory phase. In addition, a placebo of the same appearance and feel was administered. For a basic relapse prevention, various antipsychotics were given without predefining the drug or dose. For 8 months the patients alternately received, in addition to

134

Hypoestrogenism and Estrogen Replacement Therapy

Fig. 2. Study design of a multi-center, randomized, placebo-controlled, double-blind, cross-over study based on an A-B-A-B- (and/or B-A-B-A) design (Bergemann et al. 2004b)

the various antipsychotics, estradiol or placebo for 2 months each according to the randomization schedule for the study groups A-B-A-B or B-A-B-A (cf. Fig. 2). The endocrinological parameters were calculated separately for the age groups 21–45 and 45+ so as to take potential age effects into account. As defined as inclusion criteria the age group 21–45 showed hypoestrogenism at baseline. The age group 45+ displayed a menopausal hormone profile at baseline, with low estradiol levels along with elevated gonadotropins, as expected. A significant effect of the adjuvant hormone replacement therapy on the estradiol levels could be observed in both groups, and high and low levels of estradiol prevailed in the verum and placebo phases, respectively. During the 2-month replacement therapy phases the estradiol levels corresponded more or less to normal; in the placebo phases the levels were considerably below the norm. We did not find any relevant difference either in the psychopathology scales (Brief Psychiatric Rating Scale, BPRS, Overall and Gorham 1962; Positive and Negative Symptom Scale, PANSS, Kay et al. 1986; Beck Depression Inventory, BDI, Beck et al. 1987, German version by Hautzinger et al. 1995; Clinical Global Impression, CGI, Guy 1976; 100 mm analoque scale) or in defined relapse events (BPRS-Score 3: “Thought disorder” 3 points lower; CGI: deterioration to score 7 or 8; 100-mm-scale: 20% lower score) between the

135

Niels Bergemann et al.

verum (estradiol replacement) and placebo phase. Neither the required antipsychotic doses (haloperidol equivalence) nor the data on side effects differed between the two phases. Thus, the results of our study could not confirm the hypothesis that a combined estradiolantipsychotic therapy is superior to relapse prevention with antipsychotic monotherapy. Furthermore, the results could not provide evidence for the hypothesis of an estrogen effect especially on the negative symptoms as via an effect on the serotonergic transmission described by Fink et al. (1996) and as concluded by Lindamer et al. (2001). These results were critically reviewed with respect to the literature and regarding the methodology. One objection might be that the phases of verum and placebo treatment were too short. In response to this concern, we refer to Riecher-Rössler et al. (1994), who showed changes in psychopathology based on physiological fluctuations that were much smaller than the estradiol values in the present study. Furthermore, the women included in the study of Riecher-Rössler et al. – as well as in the studies by Kulkarni et al. (1996; 2001; cf. chap. 5 in this volume) – were in the acute phase of schizophrenia, which differs from the women investigated in our study. It is remarkable that in the open intervention trial by Kulkarni et al. (1996) the superiority of a combined treatment also wanes after 3 weeks. Given these results, it could be assumed that estradiol has an antipsychotic-like effect in the acute phase, in particular before the antipsychotics administered have displayed their antidopaminergic effect. After this period a “ceiling effect” could occur, which means that the potential effect of estradiol is outperformed by the effect of antipsychotics. In this context, studies analyzing low-dose antipsychotic treatment might be more sensible – they could serve as an alternative to trails, in which estrogens in monotherapy versus placebo are tested. The latter of course could present an ethical problem.

Summary and Conclusions Following the work of several authors at the end of the nineteenth and the beginning of the twentieth century a few recent studies provided evidence for primary hypoestrogenism in women suffering from schizophrenia independent of hypoestrogenism secondary to

136

Hypoestrogenism and Estrogen Replacement Therapy

hyperprolactinemia induced by some antipsychotics. This estrogen deficiency could indicate a disturbance in the hypothalamicpituitary-gonadal axis in those patients. The underlying pathogenetic process for this disturbance has not been clarified yet, however, some assumptions relating to this topic have been put forward (Bergemann et al. 2004a). In this context, a highly interesting issue to be addressed in further studies is whether schizophrenic women are suffering from long-term hypoestrogenism. An answer to this imminent question is of particular value as somatic problems secondary to hypoestrogenemia in schizophrenic women might require specific therapy. In clinical routine, taking of the patient history should include questions regarding the consequences of hypostrogenism such as menstrual irregularities and amenorrhea or galactorrhea. As hypoestrogenism could also occur in women with schizophrenia reporting regular menstrual cycles, it would be important to know the serum levels of estrodial and prolactin, and potentially of other relevant hormones. Furthermore, the different endocrinological effects of various antipsychotics have to be taken into consideration. From a clinical point of view, the findings that indicate hypoestrogenism in schizophrenic women are of major interest for the estrogen hypothesis in schizophrenia, which postulates a protective effect of estradiol. Estrogen replacement therapy in these patients is considered to be of high value. For this reason, we conducted a prospective randomized, placebo-controlled, double-blind, crossover study in women suffering from schizophrenia (Bergemann et al. 2004b). In this study, we could not find an advantage of the estrogen replacement as an adjuvant to routine relapse prevention with antipsychotics in our patients, which was contrary to our hypothesis. Although an additional effect of estrogen could not be proven, these results still cannot be regarded as refuting the estrogen protection hypothesis. Further studies should consider and overcome the limitations of our study. Assuming that the influence of hormones is different in different groups of patients, as the results of Albus and Maier (1995) as well as Könecke et al. (2000) indicate, the group of women falling ill in the perimenopause represents a particularly interesting subject for further intervention studies. In this group, estrogens could have played a decisive role in suppressing the symptoms until the onset of schizophrenia; the physiological decrease in estrogen serum levels might

137

Niels Bergemann et al.

play a key role in the etiopathogenesis of schizophrenia in these women. To verify these assumptions, further clinical research into the estrogen protection hypothesis in the form of intervention studies of perimenopausal women hospitalized for the first time could provide further insights. However, gynecological indications may prevail in particular in the patient group in question because of perimenopausal complaints. Estrogen replacement therapy has also been recommended in perimenopausal women, for example, as osteoporosis prevention treatment. In general, more interventional trials should be carried out, ideally with a double-blinded and placebo-controlled design and that take the estrogen status of the patients into consideration. However, due to potential side effects and contraindications the individual risk and benefit of such estrogen treatment should be assessed and taken into account (cf. Notman and Nadelson 2002; Schneider 2002; Writing Group for the Women’s Health Initiative Investigators 2002; Writing Group of the International Menopause Society Executive Committee 2004; Turgeon et al. 2004). Within this framework, the use of compounds with more specific and/or potent estrogenic activity in the brain than in other tissues should be the target of further research (Riecher-Rössler 2002; 2003; Halbreich 2002; Cyr et al. 2002). The application of selective estrogen receptor modulators (SERMS), but also of other estrogenic compounds such as phyto-estrogens, xeno-estrogens and dihydroepiandrosterone, should be investigated as they could minimize the side effects of estrogen therapy. Further research in this area should focus on both the potential therapeutic effects of estrogen and/or other estrogenic compounds in schizophrenia and on studies providing more insight into the currently poorly understood relationship between hypoestrogenism and/or disturbances of the hypothalamic-pituitary-gonadal axis and schizophrenia. For example, further endocrinological studies in schizophrenic patients should contribute to understanding the hormonal status of these patients and also provide results on the interdependency of the various hormones relevant in this context and their role in the pathogenesis of schizophrenia. Furthermore, studies on the effects of estrogens on cognition in the therapy of schizophrenia may be of major interest (Bergemann et al. 2001; Hoff et al. 2001). Here, brain-imaging studies might shed light on the relationship between estrogen, cognitive function, and psychopathology

138

Hypoestrogenism and Estrogen Replacement Therapy

(cf. Resnick et al. 2001; chap. 12 of this volume). Further research regarding the role of estrogen in schizophrenia would contribute to our understanding of the development and the course of schizophrenia to the benefit of patients suffering from this disease.

References Albus M, Maier W (1995) Lack of gender differences in age at onset in familial schizophrenia. Schizophr Res 18: 51-57 Angermeyer MC, Kühn L (1988) Gender differences in age at onset of schizophrenia. An overview. Eur Arch Psychiatry Neurol Sci 237: 351-364 Battey R (1876) Extirpation of the functionally active ovaries for the remedy of otherwise incurable diseases. Trans Am Gynecol Soc 1: 101-120 Beck AT, Steer RA (1987) Beck depression inventory – Manual. San Antonio, The Psychological Corporation Behl C (2001) Estrogens – mystery drug for the brain. The neuroprotective activities of the female sex hormone. Springer, Wien New York Bergemann N, Jaggy S, Maier S, Kaiser D, Auler B, Parzer P, Resch F, Mundt Ch (2001) Estrogens effects on neuropsychological performance in schizophrenic women. Arch Women Ment Health 3 (Suppl 2): 20-21 Bergemann N, Mundt Ch, Resch F, Parzer P, Runnebaum B (1996) The hypoestrogenism hypothesis in female schizophrenia: preliminary hormone screening results. Eur Psychiatry 11 (Suppl 4): 403 Bergemann N, Mundt Ch, Parzer P, Jannakos I, Nagl I, Salbach B, Klinga K, Runnebaum B, Resch F (2004a) Plasma concentrations of estradiol in women suffering from schizophrenia treated with conventional versus atypical antipsychotics. Schizophr Res (in press) Bergemann N, Mundt Ch, Parzer P, Pakrasi M, Eckstein-Mannsperger U, Haisch S, Salbach B, Klinga K, Runnebaum B, Resch F (2004b) No additional effect of estrogen as an adjuvant therapy to antipsychotics for relapse prevention on the psychopathology and the course of illness in women with schizophrenia: results of a placebo-controlled double-blind study (submitted) Bergemann N, Parzer P, Nagl I, Salbach B, Runnebaum B, Mundt Ch, Resch F (2002) Acute psychiatric admission and menstrual cycle phase in women with schizophrenia. Arch Women Ment Health 5: 119-126 Bleuler E (1943) Die spätschizophrenen Krankheitsbilder. Neurologie 15: 259-290 Brockington IF, Kelly A, Hall P, Deakin W (1988) Premenstrual relapse of puerperal psychosis. J Affect Disord 14: 287-292 Burger G (1984) Nerven- und Geisteskrankheiten als Indikationen für eine bilaterale Oophorektomie im späten 19. Jahrhundert. Dissertation, Med. Fakultät, Erlangen-Nürnberg Canuso CM, Goldstein JM, Wojcik J, Dawson R, Brandman D, Klibanski A, Schildkraut JJ, Green AI (2002) Antipsychotic medication, prolactin elevation, and ovarian function in women with schizophrenia and schizoaffective disorder. Psychiatry Res 111: 11-20

139

Niels Bergemann et al. Choi SH, Kang SB, Joe SH (2001). Changes in premenstrual symptoms in women with schizophrenia: a prospective study. Psychosom Med 63: 822-829 Clouston S (1906) Clinical lectures on mental disease. London Cyr M, Carlon F, Morissette M, DiPaolo T (2002) Estrogenic modulation of brain activity: implications for schizophrenia and Parkinson’s disease. J Psychiatry Neurosci 27: 12-27 DeLisi LE, Dauphinais ID, Hauser P (1989) Gender differences in the brain: are they relevant to the pathogenesis of schizophrenia? Compr Psychiatry 30: 197-208 Diczfalusy E, Lauritzen C (1961) Östrogene beim Menschen. Springer, Berlin DiPaolo T (1994) Modulation of brain dopamine transmission by sex steroids. Rev Neurosci 5: 27-42 Fink G, Sumner BEH, McQueen JK, Wilson H, Rosie R (1998) Sex steroid control of mood, mental state and memory. Clin Exp Pharmacol Physiology 25: 764-775 Fink G, Sumner BEH, Rosie R, Grace O, Quinn JP (1996) Estrogen control of central neurotransmission: effect on mood, mental state, and memory. Cell Mol Neurobiol 16: 325-344 Forrest AD, Hay AJ (1971) Sex differences and the schizophrenic experience. Acta Psychiatr Scand 47: 137-149 Garcia-Segura LM, Azcoitia I, DonCarlos LL (2001) Neuroprotection by estradiol. Progr Neurobiol 63: 29-60 Gattaz WF, Vogel P, Riecher-Rössler A, Soddu G (1994) Influence of the menstrual cycle phase on the therapeutic response in schizophrenia. Biol Psychiatry 36: 137-139 Georgi F, Fels E (1933) Follikelhormonbestimmungen im Harn schizophrener Frauen. Z Gesamte Neurol Psychiatrie 147: 746-756 Glick J, Stewart D (1980) A new drug treatment for premenstrual exacerbation of schizophrenia. Compr Psychiatry 21: 281-287 Gordon JH, Borison RL, Diamond BI (1980) Modulation of dopamine receptor sensitivity by estrogen. Biol Psychiatry 15: 389-396 Guy U (ed) (1976) ECDEU Assessmant manual for psychopharmacology, revised. DHEU Publication No. (ADM): 76-338. Rockville, National Institute of Mental Health Häffner R (1912) Beziehungen zwischen Menstruation und Nerven- und Geisteskrankheiten auf Grund der Literatur und klinischer Studien. Z Gesamte Neurol Psychiatrie 9: 154-223 Häfner H, an der Heiden W, Behrens S, Gattaz F, Hambrecht M, Löffler W, Maurer K, Munk-Jørgensen P, Nowotny B, Riecher-Rössler A, Stein A (1998) Causes and consequences of the gender differences in age at onset of schizophrenia. Schizophr Bull 24: 99-113 Häfner H, Behrens S, De Vry J, Gattaz F (1991) Oestradiol enhences the vulnerability threshold for schizophrenia in women by an early effect on dopaminergic neurotransmission. Evidence from an epidemiological study and from animal experiments. Eur ArchPsychiatr Clin Neurosci 241: 65-68 Häfner H, Riecher-Rössler A, Maurer K, Fätkenheuer B, Löffler W (1992) First onset and early symptomatology of schizophrenia. A chapter of epidemiologicyl and neurobiological research into age and sex differences. Eur Arch Psychiatr Clin Neurosci 242: 109-118

140

Hypoestrogenism and Estrogen Replacement Therapy Halbreich U, Rojansky N, Palter S, Tworek H, Hissin P, Wang K (1995) Estrogen augments serotonergic activity in postmenopausal women. Biol Psychiatry 37: 434-441 Hallonquist JD, Seeman MV, Lang M, Rector NA (1993) Variation in symptom severity over the menstrual cycle of schizophrenics. Biol Psychiatry 33: 207-209 Harris AH (1997) Menstrually related symptom changes in women with schizophrenia. Schizophr Res 27: 93-99 Hautzinger M, Bailer M, Worall H, Keller F (1995) Beck-Depressions-Inventar (BDI). Testhandbuch, 2nd ed. Huber, Bern Hegar A (1884) Castration als Mittel gegen nervöse und psychische Leiden. Arch Gynäkologie 24: 318-323 Hendrick V, Altshuler LL, Burt VK (1996) Course of psychiatric disorders across the menstrual cycle. Harvard Rev Psychiatry 4: 200-207 Hoff AL, Kremen WS, Wieneke MH, Lauriello J, Blankfeld HM, Faustman WO, Csernansky JG, Nordahl TE (2001) Association of estrogen levels with neuropsychological performance in women with schizophrenia. Am J Psychiatry 158: 1134-1139 Huber TJ, Rollnik J, Wilhelms J, von zur Mühlen A, Emrich HM, Schneider U (2001) Estradiol levels in psychotic disorders. Psychoneuroendocrinology 26: 27-35 Jolly Ph (1915) Menstruation und Psychose. Arch Psychiatrie Nervenkrankheiten 55: 637-686 Kafka V (1935) Hormone bei Schizophrenie. Wien Med Wochenschr 85: 30-31 Kay SR, Fiszbein A, Opler LA (1987) The positive and negative syndrome scale (PANSS) for schizophrenia. Schizophr Bull 13: 261-276 Kendell RE, Chalmers JC, Platz C (1987) Epidemiology of puerperal psychoses. Br J Psychiatry 150: 662-673 Klinga K (1994) Determination of hormones and hormone receptors. In: Runnebaum B, Rabe T (eds) Gynecological endocrinology and reproductive medicine, vol 1. Springer, Berlin Heidelberg New York, 45-54 König C, Linzenmeier G (1913) Ueber die Bedeutung gynäkologischer Erkrankungen und den Wert ihrer Heilung bei Psychosen. Arch Psychiatrie Nervenkrankheiten 51: 1002-1054 Könnecke R, Häfner H, Maurer K, Löffler W, an der Heiden W (2000) Main risk factors for schizophrenia: increased familial loading and pre- and peri-natal complications antagonize the protective effect of oestrogen in women. Schizophr Res 44: 81-93 Kraepelin E (1909) Psychiatrie, Bd 1–4. Barth, Leipzig Krafft-Ebing R v (1878) Untersuchungen über das Irresein zur Zeit der Menstruation. Arch Psychiatrie Nervenkrankheiten 8: 65-107 Krafft-Ebing R v (1903) Lehrbuch der Psychiatrie auf klinischer Grundlage für praktische Ärzte und Studierende, 7th ed. Ferd. Enke, Stuttgart Kretschmer E (1921) Körperbau und Charakter. Springer, Berlin Kulkarni J, deCastella A, Riedel A, Taffe J, Fitzgerald P, Burger H (2001) Estrogen – a potential new treatment in schizophrenia. Schizophr Res 48: 137-144 Kulkarni J, deCastella A, Smith D, Taffe J, Keks N, Copolov D (1996) A clinical trial of the effects of estrogen in acutely psychotic women. Schizophr Res 20: 247-252

141

Niels Bergemann et al. Lansbury J, Hughes J (1939) Estrin excretion in psychotic patients. Am J Psychiatry 95: 1119-1125 Lee SJ, McEwen BS (2001) Neurotrophic and neuroprotective actions of estrogens and their therapeutic implications. Annu Rev Pharmacol Toxicol 41: 569-591 Leidenberger FA (1992) Klinische Endokrinologie für Frauenärzte. Berlin, Springer Lewine RRJ (1981) Sex differences in schizophrenia. Psychol Bull 90: 432-444 Lewine RRJ, Strauss JS, Gift TE (1981) Sex differences in age at first hospital admission for schizophrenia: a fact or artifact? Am J Psychiatry 138: 440-444 Lindamer LA, Buse DC, Lohr JB, Jeste DV (2001) Hormone replacement therapy in postmenopausal women with schizophrenia: positive effect on negative symptoms? Biol Psychiatry 49: 47-51 Loranger AW (1984) Sex difference in age of onset of schizophrenia. Arch Gen Psychiatry 41: 157-161 Mahé V, Dumaine A (2001) Oestrogen withdrawl associated psychoses. Acta Psychiatr Scand 104: 323-331 Mall G (1958) On the hormonal treatment of pre- and postmenstrual ovarian psychoses. In: Reiss M (ed) Psychoendocrinology. Grune & Stratton, Orlando, 96-103 Mall G (1960) Diagnostik und Therapie ovarieller Psychosen. Zentralbl Gesamte Neurol Psychiatrie 155: 250 Mayer CEL (1872) Menstruation im Zusammenhange mit psychischen Störungen. Beiträge zur Geburtshülfe und Gynäkologie 1: 111-135 Mayer CEL (1869) Die Beziehungen der krankhaften Zustände in den Sexualorganen des Weibes zu Geistesstörungen. Berlin McEwen BS, Alvers SE (1999) Estrogen actions in the central nervous system. Endocrine Rev 20: 279-307 Melkersson KI, Hulting A-L, Rane AJ (2001) Dose requirement and prolactin elevation of anttipsychotics in male and female patients with schizophrenia or related psychoses. Br J Clin Pharmacol 51: 317-324 Morissette M, DiPaolo T (1993) Sex and estrous cycle variations of rat striatal dopamine uptake sites. Neuroendocrinology 58: 16-22 Nilsson P (1939) Der Prolan- und Follikulingehalt des Harns bei schizophrenen Frauen mit in der Genitalsphäre lokalisierten taktilen Halluzinationen. Upsala Läkareförenings Förhandlingar 44: 495-504 Notman MT, Nadelson CC (2002). The hormone replacement therapy controversy. Arch Women Ment Health 5: 33-35. Nott PN (1982). Psychiatric illness following childbirth in Southampton: a case register study. Psychol Med 12: 557-561 Oesterreicher W (1934) Die Ausscheidung des gonadotropen Hormons des Hypophysenvorderlappens und des weiblichen Sexualhormons (Follikelhormon) bei Manien, Depressionen und Schizophrenien im geschlechtsreifen Alter. Wien Klin Wochenschr 46: 1385-1388 Overall JE, Gorham DR (1962) The brief psychiatric rating scale. Psychol Reports 10: 799-812 Rabe T, Runnebaum B (1997) Hormones. In: Runnebaum B, Rabe T (eds) Gynecological endocrinology and reproductive medicine, vol 1. Springer, Heidelberg Berlin New York, 1-43

142

Hypoestrogenism and Estrogen Replacement Therapy Reiss M (ed) (1958) Psychoendocrinology. Grune & Stratton, Orlando Riecher-Rössler A (2003) Oestrogens and schizophrenia. Curr Opin Psychiatry 16: 187-192 Riecher-Rössler A, Häfner H, Stumbaum M, Maurer K, Schmidt R (1994) Can estradiol modulate schizophrenic symptomatology? Schizophr Bull 20: 203-214 Ross SM (1909): Menstruation in its relationships to insanity. J Mental Sci 55: 270-280 Sacerdoti G, Carry E (1955) Funzionalità epatica e reattivita cutanea agli estrogeni in schizofreniche con iperestrogenismo. Rivista di Patologia Nervosa e Mentale 76: 781-790 Saethre H (1933) Titrierung von Sexualhormonen bei Geisteskranken. Klin Wochenschr 12: 1409-1410 Salokangas RR (1983) Prognostic implications of the sex of schizophrenic patients. Br J Psychiatry 142: 145-151 Sartorius N, Jablensky A, Shapiro R (1978) Cross-cultural differences in the shortterm prognosis of schizophrenic psychoses. Schizophr Bull 4: 102-113 Sartorius N, Jablensky A, Korten A, Ernberg G, Anker M, Cooper JE, Day R (1986) Early manifestations and first-contact incidence of schizophrenia in different cultures. Psychol Med 16: 909-928 Schaefer A (1894) Einfluss der Psychose auf den Menstruationsvorgang. Allg Z Psychiatrie Psychisch-gerichtliche Medizin 50: 976-996 Schneider HPG (2002) The view of the International Menopause Society on the Women’s Health Initiative. Climacteric 5: 211-216 Schroeter R (1874) Die Menstruation in ihren Beziehungen zu den Psychosen. Z Psychiatrie 30: 551-572 (part 1); 31: 234-250 (part 2) Sears HA, Morter RA, Simonsen M, Williams C (1937) Blood estrin level in schizophrenia. Am J Psychiatry 93: 1293-1303 Seeman MV (1997) Psychopathology in women and men: focus on female hormones. Am J Psychiatry 154: 1641-1647 Seeman MV (1996) The role of estrogen in schizophrenia. J Psychiatry Neurosci 21: 123-127 Seeman MV (1985) Sex and schizophrenia. Can J Psychiatry 30: 313-315 Seeman MV (1983) Interaction of sex, age, and neuroleptic dose. Compr Psychiatry 24: 125-128 Seeman MV (1982) Gender differences in schizophrenia. Can J Psychiatry 27: 107-112 Smith S, Wheeler MJ, Murray R, O’Keane V (2002) The effects of antipsychotic induced hyperprolactinaemia on the hypothalamic-pituitary-gonadal axis. J Clin Psychopharmacol 22: 109-114 Swanson DW, Barron A, Floren A, Smith JA (1964) The use of norethynodrel in psychotic females. Am J Psychiatry 120: 1101-1103 The North American Menopause Society (2003) Estrogen and progestogen use in peri- and postmenopausal women: September 2003 position statement of The North American Menopause Society. Menopause 10: 497-506 Tsuang MT, Dempsey GM, Rauscher FA (1976) A study of “atypical schizophrenia”: comparison with schizophrenia and affective disorder by sex, age of admission, precipitant, outcome, and family history. Arch Gen Psychiatry 33: 1157-1160

143

Hypoestrogenism and Estrogen Replacement Therapy Tukey JW (1977) Exploratory data analysis. Reading Mass., Addison-Wesley Turgeon JL, McDonnel DP, Martin KA, Wise PM (2004) Hormone therapy: physiological complexity belies therapeutic simplicity. Science 304: 1269-1273 Wells S (1886) Castration in mental and nervous diseases. A symposium by Sir T. Spencer Wells, Dr. Alfred Hegar and Robert Battey MD. Am J Med Sci 8: 455-471 Writing Group for the Women’s Health Initiative Investigators (2002) Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized controlled trial. J Am Med Assoc 288: 321-333 Writing Group of the International Menopause Society Executive Committee (2004) Guidelines for the hormone treatment of women in the menopause transition and beyond. Climacteric 7: 8-11

144

7 The Effect of Estrogens on Depression Linda S. Kahn and Uriel Halbreich

Introduction: Gender Differences in Prevalence of Depression and Responses to Psychotropic Medication Women suffer disproportionately from several major psychiatric disorders. According to worldwide epidemiological studies, the prevalence of major depressive disorder (MDD) among women is 1.5–3 times higher than that of men (Kessler 2000; Weissman et al. 1996). According to The World Health Organization (WHO), unipolar MDD is the number one cause of disease burden for reproductiveaged women (18–44 years old) in both developed and developing countries, followed by schizophrenia (Murray and Lopez 1996). In developing regions, suicide is the fourth leading cause of disease burden for women (Murray and Lopez 1996) – which suggests that serious mental disorders such as depression are not adequately diagnosed and treated. In the USA, the National Comorbidity Survey (NCS) reports that the lifetime prevalence of MDD is 21% for women, compared to 13% for men (Kessler et al. 1994). For dysthymia (a chronic low-grade form of depression), the lifetime prevalence is 8% for women, and 4.8% for men (Kessler et al. 1994). In the USA, the lifetime prevalence for any affective disorder among women is reported to be 24% for women and 15% for men (Kessler et al. 1994). These reports confirm earlier studies (Weissman et al. 1991) reporting gender differences in prevalence of unipolar MDD of 7% in women as compared to 2.8% in men and a prevalence of dysthymia of 4.1% in women and 2.2% in men. Women are also more vulnerable to developing other types of affective disorders, including anxious depression (van Valkingburg 1984), seasonal affective disorders (Parry 1989), and phobias (Kessler 1995), compared to men. The gender disparity in affective disorders is most evidenced between menarche and 55 years of age, suggesting that gonadal hormones play a major role in the onset and course of depressions in women (Archer 1999; Burvill 1995).

145

Linda S. Kahn and Uriel Halbreich

The first onset of mood symptoms or exacerbations of pre-existing affective disorders may be triggered by specific hormonal and lifecycle conditions (e.g., menarche, postpartum, peri-menstrual, perimenopause) (Stewart and Boydell 1993; Halbreich 1996, 1997; Appleby et al. 1994; Kumar and Robson 1984; O’Hara and Swain 1996). Evidence exists suggesting a significant association between certain reproductive-related depressions (RRDs), particularly premenstrual dysphoric disorder (PMDD), and other affective disorders, including postpartum depression (PPD) and MDD (Halbreich and Endicott 1985). The consumption of psychotropic medications, particularly anxiolytics and antidepressants (Halbreich 1997), is higher among women than men, with a ratio of greater than 2:1 (Ohayon and Caulet 1995; Rawson and D’Arcy 1991; Skegg et al. 1977). Compared to men, women taking psychotropic drugs experience more frequent side effects (about twice as often than men), and they also experience more adverse responses to psychotropic drugs (Hamilton and Yonkers 1996; Halbreich 1997). It has been suggested that women are more frequently given questionable and high-risk prescriptions than men, particularly for psychotropic drugs (Jensvold and Hamilton 1996). Drug consumption, including psychotropic drugs, increases with age and is higher among women (Halbreich 1997; Jensvold and Hamilton 1996). Among adults over 65 years of age, 16.7% of women, compared to 10.1% of men, take psychotropic drugs (Jensvold and Hamilton 1996). Because it is more common for women to be taking multiple medications (e.g., hormone replacement therapy, HRT, oral contraceptives, OCs, anti-inflammatories, thyroid hormones, etc.), they may also be at higher risk of adverse drug-drug reactions (Seeman 2000). Women differ from men in their pharmacokinetic and pharmacodynamic responses to psychotropic medications. Gender differences exist in key pharmacokinetic variables, including absorption, distribution and bioavailability, hepatic function and changes in hepatic enzymes, metabolism, and clearance and elimination (Jensvold 1996; Yonkers et al.; Halbreich 2000). Because women’s bodies contain proportionately more adipose tissue than men’s, and because psychotropic drugs (especially antipsychotics, antidepressants, and anxiolytics) are lipophilic, these drugs remain longer in women’s bodies after discontinuance and may cause side effects during periods of rapid weight loss (Seeman 2000). In addition, liver pathways and

146

The Effect of Estrogens on Depression

renal clearance are slower in younger women than in men and further contribute to higher psychotropic blood levels for longer periods (Halbreich 1997). Other important gender-related pharmacokinetic factors include gonadal hormones, ovulation-related cyclicity, and exogenous hormones such as OCs or HRT – all of which may significantly alter protein binding (Yonkers et al. 2000; Hamilton and Yonkers 1996). Central nervous system (CNS) distribution of medications may be accelerated during periods of more rapid blood flow to the CNS (Yonkers et al. 2000). In general, women of reproductive age and postmenopausal women on estrogen replacement therapy have greater cerebral blood flow (CBF) than men, contributing to improved site delivery of psychotropic medications (Halbreich 1997). It is likely that also pharmacodynamic properties contribute to a gender difference in drug-response. Pharmacodynamic processes include receptor distribution and action, neural structure/plasticity, enzymes, and messenger systems (Jensvold and Hamilton 1996). Pharmacodynamics may also include drug-drug interactions and the behavioral impact of drug administration. Although less is known about sex differences in the pharmacodynamic effects of medications, these differences may be as significant as pharmacokinetic aspects, especially if therapeutic compounds interact with testosterone, estrogen, or progesterone (Yonkers et al. 2000).

Role of Estrogen in Symptom Formulation The brain and gonadal hormones interact in a bi-directional manner. Peripheral gonadal hormones exert both organizational/genomic and activational/nongenomic actions on the CNS (McEwen 1991). The organizational/genomic effects are trophic and permanent and control neural architecture and future activity. They occur during the brain’s early development, and their influences, or their absence, are responsible for gender differences in brain and behavior. Activational/nongenomic effects of gonadal hormones occur mostly during postnatal life and continue throughout the life cycle, adding to and supporting gender-differentiated brain functions. They are reversible and entail alterations of normal electrical and biochemical functions and structure, as well as many functions implicated in the continuous regulation of mood and behavior that are believed to be

147

Linda S. Kahn and Uriel Halbreich

impaired in mental disorders and to be influenced by psychotropic medications. These activities include receptor potentiation by direct or modulatory effects, enzyme induction or inhibition, potentiation of intracellular second-messenger processes, and other transitory effects. Due to their selective distribution in the brain, the effects of receptors, binding sites, and other targeted sites for actions of gonadal hormones cannot be generalized. In addition, different metabolic or synthetic analogues might cause opposing behavioral effects. For example, several progesterone metabolites and progestins are anxiolytic while some others are anxiogenic. Timing is also a key factor – and cyclic administration might cause different effects than continuous use (Halbreich 2000). In addition, priming with estrogen might alter the effects of progesterone.

Estrogen Cyclicity and the Course of Pre-existing Mood Disorders The role of gonadal steroids in the precipitation and course of mood disorders in women has been well documented (Archer 1999; Rubinow et al. 1998a, b; Yonkers 1998). The cyclicity of hormonal fluctuations throughout reproductive life is one of the dominant characteristics of women’s biology. This is especially evidenced by the significant hormonal changes during pregnancy, followed by the abrupt postpartum withdrawal of estrogen. During the last trimester of pregnancy, hormonal rates are relatively stable, and depression is rare – but the rates jump to 20% postpartum, corresponding to the sharp drop in estrogen (Archer 1999; Studd and Smith 1994). The inherent cyclical hormonal instability, combined with the multifaceted interactions between hormones, neurotransmitters, and other CNS processes, as well as the influences of behavior and mood, might explain why women appear more vulnerable to affective disorders than men (Halbreich 2000). Menstrually related processes (presumably hormonal) have been linked to exacerbation of a variety of related disorders, including late luteal phase exacerbation of MDD, bulimia nervosa or epilepsy, panic attacks, agoraphobia, alcoholism, puerperal psychosis-like episodes, periodic psychosis, violent criminal acts, and various other conditions (Abramowitzet al. 1982; Backstrom et al. 1984; Brockington et al. 1988a, b; Endicott and Halbreich 1988; Friedman et al. 1982;

148

The Effect of Estrogens on Depression

Gladis and Walsh 1987; Hatotani et al.1979; d’Orban and Dalton 1980; Luggin et al. 1984; Price et al. 1987). That such a broad spectrum of diagnostic conditions and symptom clusters may be expressed or exacerbated during the late luteal phase of the menstrual cycle suggests that they may be activated by a similar trigger or conditions, depending upon the individual’s specific vulnerability (Halbreich 1997).

The Relevant Influences of Estrogen on Mood-related CNS Processes Estrogen has multiple effects on neural functions putatively implicated in multiple processes involving the regulation of mood, behavior, and cognition (Halbreich and Lumley 1993). It enhances monoamine activity and increases serotonergic (5-HT) postsynaptic responsivity (Halbreich et al. 1995), increasing both the number of serotonergic receptors as well as neurotransmitter transport and uptake (McEwen et al. 1997; Matsumoto et al. 1985). Estrogen also increases 5-HT synthesis and 5-HTIAA levels (Dickinson and Curzon 1986). It upregulates 5-HT1 receptors and downregulates 5-HT2 receptors and decreases MAO activity (Chakravorty and Halbreich 1997). The cumulative effect of estrogen on serotonergic function is as a 5-HT agonist (Halbreich 1997). In selective brain regions estrogen acts as a cholinergic agonist. It accelerates the activity of acetylcholine transferase in the preoptic area, amygdala, horizontal diagonal nucleus, frontal cortex, and area CA1 of the hippocampus (McEwen et al. 1997; 1998; Wolley and McEwen 1992). It increases the number of muscarinic receptors in the medial, lateral, and ventromedial hypothalamus, but decreases their number in the medial preoptic area. Estrogen also increases electrical firing of neurons in the hypothalamus in response to acetylcholine. Estrogen selectively increases norepinephrine (NE) activity in the brain. Its effect on the enzyme tyrosine hydroxylase is mixed. Some investigators have reported increased activity in some areas, whereas others have reported decreased activity in other areas. Estrogen increases NE turnover, while its effects on plasma levels of the NE metabolite 3-methoxy-4-hydroxyphenylglycol (MHPG) are mixed. Increased NE activity may be caused by decreased NE reuptake and decreased NE metabolism due to inhibition of monoamine oxidase and decreased

149

Linda S. Kahn and Uriel Halbreich

catechol-o-methyl transferase (COMT) activity (Luine 1985). Estrogen’s effect on α2-andrenoceptor binding is mixed, and it increases β-adrenoceptor binding. It decreases the binding of the partially related imidazoline receptors (Piletz and Halbreich 2000). Estrogen decreases D2 receptor sensitivity and probably other dopaminergic receptors as well. Estrogen’s influence is not limited to monoamines. It acts as a γ-aminobutyric acid (GABA) adjunct agonist by increasing binding of GABA agonists and their upregulation of GABA receptors and also decreases activity of glutamic acid decarboxylase in the hypothalamus. Its effects on endorphins is mixed: it increases overall endorphic activity but decreases their levels in the mesolimbic hypothalamus (see review by McEwen et al. 1997) Estrogen’s neurostructural effects point to a broad and extensive impact on the CNS, with significant implications for human cognitive abilities (McEwen et al. 1997). Estrogen induces dendritic spines, creating new synapses in the ventromedial hypothalamus. It also increases dendritic spinal density on pyramidal neurons in the hippocampus (McEwen et al. 1997; 1998; McEwen and Woolley 1994). Estrogen stimulates expression of neurotrophic factors (NGF, BDNF) (Singh et al. 1994; Singh et al. 1995) and stimulates axonal regeneration and synaptogenesis (Matsumoto et al. 1985). It decreases cAMP while increasing PCK. It maintains the viability of neurons and inhibits β-amyloid toxicity (Simpkins et al. 1994). Estrogen also increases cerebral blood flow and glucose transport (McEwen et al. 1998; McEwen and Woolley 1994; Ohkura et al. 1994; Birge 1997). The overall neurostructural and neurophysiological effects of estrogen can be summarized as preventative of neurodegeneration (neuroprotective), probably neuroregenerative, and selectively stimulating of neurotransmission. Some of its inhibitory effects might also be viewed as contributing to mental well-being, and thus pertinent to current hypotheses of the pathophysiology of depressions and their treatment (Archer 1999; Halbreich 1997; Garlow et al. 1999; Young and Korszun 1998; Young et al. 2000).

150

The Effect of Estrogens on Depression

The Role of Estrogen in the Symptoms, Mechanisms, and Treatment of MDD and Reproductive-Related Disorders in Women Major Depressive Disorder There is growing evidence of differences between women and men in the clinical manifestations of MDD (Kornstein and Schatzberg 2000). Depressed women present more frequently with atypical symptoms: increased appetite, weight gain, as well as anxiety and somatic symptoms (Kornstein 1997). A recent large clinical study reported that women with chronic major depressive disorder are more severely depressed and experience greater functional impairment than men (Kornstein and Schatzberg 2000; Kornstein 1997; Kornstein et al. 1995). Women were also more likely to report a younger age at onset and a family history of clinical depression (Kornstein and Schatzberg 2000; Kornstein and McEnany 2000). A study of a female twin population found a high concordance in the subphenotypes of depression, especially for monozygotic (MZ) twins (Kendler et al. 1996), with differences in comorbidity of depression. Women with atypical depression had high rates of bulimia and increased obesity, compared to those with severe “typical” depression who had comorbid anxiety and panic as well as longer episodes and impairment. This suggests a partial overlap between vulnerability to atypical depression and to obesity and overeating under stress (Stunkard et al. 1990). The propensity of women to “atypical depression”, characterized by increased sleep and appetite, motor retardation, increased lethargy, anxiety, hyperactivity to external events, mood lability and interpersonal sensitivity (Stewart and Boydell 1993; West and Dally 1959), resembles the more prevalent dysphoric subtypes of premenstrual syndromes (PMS) (Halbreichet al. 1982; Wikander et al. 1998). Compared with men, women are more likely to have reported a stressful life event in the 6 months leading up to an episode of MDD (Kornstein 1997). Women’s depressions are also more frequently triggered by seasonal influences (Leibenluft and Hardin 1995) – and the female-to-male ratio for seasonal affective disorder is greater than 3:1 (Kornstein 1997; Leibenluft and Hardin 1995). Episodes of MDD among women may also be produced by reproductive-related (RRD) hormonal changes occurring during pre-menstruation, pregnancy, postpartum, and perimenopause, as well as those brought on by HRT,

151

Linda S. Kahn and Uriel Halbreich

particularly when it is administered sequentially (Parry 1989; Kornstein 1997). Research on the pathophysiology of RRDs suggests that several systems and processes may be abnormal among depressed women, including hypothyroidism (Brayshaw and Brayshaw 1986; Roy-Byrne et al. 1987; Casper et al. 1989; Halbreich et al. 1988), abnormal, mostly decreased serotonegic functions (extensive reviews: Halbreich and Tworek 1993; Lepage et al. 1991) dysregulation of NE systems and related processes (Halbreich et al. 1993; Gurguiset al. 1998a, b), anxiety symptoms in response to lactate infusions (Sandberg et al.1993), CO2 inhalation (Harrison et al. 1989) and cholecystokinin (CCK4) injections (LeMeledo et al. 1995) as well as a dysregulated dopaminegic system (Halbreich et al. 1976), which is recently drawing renewed interest. That some of these abnormalities also occur during nonsymptomatic periods suggest the existence of a trait or an indication of vulnerability. Whether these abnormalities are distinguishable from each other and represent different biological subtypes that might be associated with different phenotypes remains unclear. The consistent increased variance of biological values in women as compared to men (Halbreich and Lumley 1993) combined with the diversified symptoms of some RRDs, especially during the late luteal phase (Halbreich 1997), might support the notion of diversified biological traits that might be related to diversified genetic factors. Treatment Several studies have documented that women may respond better to serotonergic agonists, especially SSRIs, than to NE agonists, to which men respond somewhat better (Hamilton and Yonkers 1996; Kornstein and McEnany 2000; Kornstein and Schatzberg 2000; Steiner et al.1993; Yonkers et al. 1992, 1998). One explanation for this difference in treatment response is that female gonadal steroids may enhance serotonin activity, thus augmenting SSRI efficacy (Kornstein and McEnany 2000). Therefore, based on side-effect profile and patient compliance, the SSRIs are a first choice for treatment, especially for women. Estrogen has been tested both as monotherapy (Klaiber et al. 1979; Michael et al. 1970) and adjunct therapy (Prange 1972; Shapira et al.1985; Schneider et al. 1997) for treatment of MDD in reproductive-age women.

152

The Effect of Estrogens on Depression

Klaiber et al. (1979) tested conjugated estrogen as monotherapy for severe MDD in 40 female inpatients. Although the 23 women in the estrogen group showed significantly greater improvement than those in the placebo group, post-treatment depression scores in the estrogen group were still moderate to severe – raising doubts as to whether estrogen as monotherapy is efficacious. The dosages administered in the study were significantly higher than the usual ERT dosage (up to 25 mg), suggesting that estrogen might entail a significant benefit only when administered in high dosages (Klaiber et al. 1979). However, at least four studies reported no improvement in depression in response to treatment with estrogen (Prange 1972; Shapira et al. 1985; Schneider et al. 1977; Coope and Poller 1975; Coope 1981). Other studies have suggested that estrogen might improve overall sense of well-being among nondepressed women (a “mental tonic effect”), but it does not appear to be an efficacious treatment for MDD (Archer 1999). Estrogen has also been evaluated as an adjunct therapy for MDD. Prange et al. (1972) added estrogen to a tricyclic antidepressant (TCA) and noted that improvement in the TCA plus high-dose estrogen group was initially greater than in the low-dose estrogen and TCA or a TCA alone group. Two weeks later, this benefit was lost, and the high-dose estrogen plus TCA patients did not experience the same improvement in mood, compared to the low-dose estrogen plus TCA group or the TCA alone group. Cholinergic adverse effects were very high. A small study of 11 women treated with imipramine and then randomized to conjugated estrogen (in doses up to 3.75 mg per day) or placebo as an adjunct treatment reported similar results (Shapira et al. 1985). Compared to placebo augmentation, estrogen augmentation yielded no benefit. In a placebo-controlled trial of older women on fluoxetine for MDD, those who were also taking estrogen (ERT) showed slightly greater improvement than those on fluoxetine alone (Schneider et al. 1997). However, women on ERT without fluoxetine did less well than women not on ERT. Similar results were reported by Schneider’s group (Schneider et al. 1998) in a post hoc analysis of sertraline with or without HRT. Even though estrogen as monotherapy probably does not improve mood of depressed women, it might be a potent adjunct to SSRI for postmenopausal women diagnosed with depression (Schneider et al. 1997; Schneider et al. 1998).

153

Linda S. Kahn and Uriel Halbreich

Depression and the Menarche Before puberty, the prevalence of mood disorders is evenly distributed between boys and girls (Epperson et al. 1999). Beginning at menarche, however, the prevalence of depressions and dysphorias becomes significantly higher among girls (Epperson et al. 1999; Steiner et al. 2000). Among 14- to 18-year-olds, females are twice as likely to be depressed as males (Lewinsohn et al. 1998). In the USA, the prevalence of MDD is 20.6% among females between 15 and 24 years of age, compared to 10.5% among males in the same age group (Kessler and Walters 1998). Much remains unknown regarding the etiology and mechanisms underlying the apparent female-specific increased risk for depressions emerging during menarche. Angold et al. (1998) tried to pinpoint the age at which the gender disparity in depression prevalence emerges, and whether pubertal timing affects unipolar depression rates among girls. In their study of 4,509 11- to 13-year-olds, they noted that girls experienced consistently higher rates of unipolar depression than boys beginning around age 13. Using Tanner staging, they determined that this gender disparity develops during midpuberty in girls, as demarcated by Tanner stage III. More recently, Angold et al. (1999) tried to determine whether an association exists between depression among adolescent girls and pubertal morphological status, as measured by Tanner stages, or hormonal levels. They measured testosterone, estrogen, FSH and LH levels, and rates of depression among 465 girls. Their results point to a direct link between pubertal increases in testosterone, and estrogen in particular, and the prevalence of depression in females. Increases in gonadal steroid levels alone do not cause depression in adolescent girls; but may render some girls more vulnerable to depression, depending upon specific environmental or psychosocial cognitive variables. Other hypotheses on the emergence of mood disorders during menarche have focused on the serotonergic system. Because gonadal hormones influence the production and function of 5-HT receptors during menarche, alterations may result, leading to an increased vulnerability to mood disorders (Steiner et al. 2000). Disturbed sleep patterns or menstrual cycles, or other disruptions of biological rhythms, combined with psychosocial stresses, might trigger depression in vulnerable individuals (Steiner et al. 2000).

154

The Effect of Estrogens on Depression

There is much that is still unknown regarding why depressions emerge during menarche, and their pathophysiology. Once these questions are answered, knowledge about the mechanisms of gender differences in depressions will be significantly furthered. For female adolescents, no accepted specific pharmacological therapy exists – especially involving possible hormonal interventions for depression. Premenstrual Dysphoric Syndrome (Dysphoric PMS or PMDD) Approximately 50%–80% of women will experience at least a few premenstrual symptoms that may vary from mild to severe (Woods, Most, and Dery 1982), while the prevalence of severe dysphoric premenstrual symptoms (or PMDD) ranges from 3% to 9% (Woods et al. 1982; Rivera-Tovar and Frank 1990; Johnson 1987; Merikangas et al. 1993; Ramcharan et al. 1992). Patients with PMS have reported up to 300 different premenstrual complaints (Halbreich et al. 1982), with depressed mood, mood swings, anxiety/tension, anger/irritability, low interest, decreased concentration, poor energy and changes – mostly increase – in sleep and appetite, and physical symptoms most commonly attributed to PMS (Hurt et al. 1992). Since the 1930s and the seminal work of Frank (Frank 1931), gonadal steroids have been considered a major causal factor in PMS. Hypotheses included lack of estrogen, lack of progesterone, and differences in their ratios. At one time, progesterone suppository treatment was given much publicity, based on the hypothesis that PMS was caused by insufficient progesterone (Dalton 1964). Progesterone suppository treatment, however, remains controversial. It is most likely that many women with PMS have increased and not decreased levels of progesterone (Halbreich et al. 1986; Freeman et al. 1985; 1990). Absolute plasma levels of gonadal hormones are probably comparable between women with and without PMS – although there might be differences in cyclicity or rate of fluctuations as well as an increased sensitivity to “normal” fluctuations (Halbreich et al. 1986; 1988; Schmidt et al. 1998). Hormonal changes during the early luteal phase might be equally important as those that occur during the symptomatic period. The pivotal role of gonadal hormone fluctuations in the pathophysiology of PMS is further underscored by reports that women with PMS are asymptomatic during anovulatory cycles and that ovulation suppression alleviates

155

Linda S. Kahn and Uriel Halbreich

PMS symptoms (Halbreich et al. 1991). It is probable that the interactions between gonadal hormones, neurotransmitters, and other processes putatively involved in regulation of mood and behavior, combined with individual vulnerability, distinguishes women with PMS from those without PMS. Differences between these two groups in a variety of variables, including personality, cognition, serotonergic processes, thyroid functions and NE-related receptors (even among women with PMS when they do not have any symptoms) have been widely documented (for review see Halbreich 1995). This strongly suggests that the pathophysiology of PMS involves a kindling effect caused by the multifaceted interactions and accumulated impacts of vulnerability, hormonal changes, and brain processes, combined with environmental and psychosocial factors. Treatment There are a number of efficacious treatments currently used for PMS. Antidepressant medications, primarily SSRIs, are a treatment of choice for women with severe dysphoric PMS (Steiner et al. 1995; Yonkers et al. 1996). The administration of SSRIs can be limited to the luteal phase (Halbreich and Smoller 1997). With continuous or intermittent SSRI treatment, the reported success rate approaches 60%–70%. Suppression of ovulation, mostly with GnRH analogues (Muse et al. 1984), but also with danazole is widely considered the most effective treatment for a broad spectrum of physical and dysphoric PMS symptoms. Because it induces “pharmacological menopause,” however, it is not a treatment of choice for most women. Adding cyclic estrogen-progesterone replacement may undermine the effectiveness of GnRH analogues by simulating the physiological hormonal fluctuation. Estrogen has been used as a treatment modality for PMS mood symptoms or premenstrual dysphoric disorder (PMDD). Studies that have examined the effectiveness of treatment with transdermal β-estradiol followed by 7 days of norethisterone, as compared to placebo (Magos et al. 1986; Watson et al. 1989), found 17β-estradiol more effective than placebo. However an often-cited study of PMS treatment with conjugated estrogens (Dahr and Murphy 1990) reported no difference between estrogen and placebo. Because the sample size for this study was quite small (11 women), the lack of significant difference may be due to a lack of power.

156

The Effect of Estrogens on Depression

We have found that estradiol transdermal patches can be efficacious if administered in dosages high enough to suppress ovulation and in a continual, cyclic fashion to promote withdrawal and endometrial shedding (Halbreich 1999). We have had good experience with transdermal 17β-estradiol (0.2 mg) for women with PMS. This treatment suppresses ovulation, eliminates hormonally related fluctuations, and offers the CNS benefits of estrogen – without the bothersome side effects of pharmacological menopause. Postpartum Depressions Within the first 6-months postpartum, between 10% and 22% of adult women will experience an episode of major or minor depression (Appleby et al. 1994; Kumar and Robson 1984; O’Hara and Swain 1996; Cooper et al. 1996; Troutman and Cutrona 1990; O’Hara et al. 1990; Reighard and Evans 1995; Roy et al. 1993; Whiffen 1988; Demyttenaere et al. 1995; Terry et al. 1996; Cox et al. 1982); and this rate is even higher (26%) among adolescent mothers (Troutmann and Cutrona 1990). Data suggest that the risk of MDD during the puerperium may be higher than in non-postpartum periods (Whiffen 1992). One large study (n = 6,000) estimated that the 2-month prevalence for postpartum MDD was 15% (Cooper et al. 1996) – surpassing the 12-month prevalence rate for MDD found in a US epidemiological survey (Kessler et al. 1994). The onset of PPD usually occurs within the first 4 weeks after delivery (Kumar and Robson 1984; O’Hara et al. 1990; Carothers and Murray 1990; Zelkowitz and Milet 1995). Most women experience postpartum blues, the most mild of postpartum dysphorias, and this is considered normal. Postpartum blues, characterized by dysphoria, mood lability, irritability, crying, anxiety, insomnia, and poor appetite, occur in most women, begin during the first week postpartum, peak on the fifth day, and resolve shortly thereafter. Unlike postpartum blues, PPD symptoms last beyond the second week postpartum and mirror those of MDD, except for the unique timing and context. Although the most widely accepted pathobiological explanation for PPD focuses on the abrupt and substantial hormonal withdrawal immediately following delivery, this hypothesis has not yet been fully proven. A recent report that simulation of this hormonal process causes symptoms in vulnerable women (Bloch et al. 2000)

157

Linda S. Kahn and Uriel Halbreich

strongly suggests its validity. Postpartum changes in gonadal hormones are also linked to dysfunction of the thyroid system, “positive” or “negative” changes in the hypothalamo-pituitary-adrenal (HPA) system, rapid postpartum withdrawal of endorphins, and a gonadal hormones-associated decrease in serotonergic activity, and probably also in some other neurotransmitters systems. As with other mental disorders, the pathobiology of postpartum disorders may also involve vulnerability to affective disorders and to disrupted homeostasis. It is likely that the abrupt withdrawal of gonadal hormones and their influence on a broad spectrum of other biological systems involved in the regulation of mood and behavior results in disrupted homeostasis. Almost all women have “blues” shortly following delivery, but there is probably a strong stabilizing mechanism that restores the euthymic state. In some women, that mechanism may be deficient; they continue to be dysphoric with increased severity for longer periods. In these cases, treatment is warranted. Estrogen may be an effective treatment for postpartum MDD. Gregoire et al. (1996) treated 64 women with MDD with either transdermal 17β-estradiol or with placebo. The 17β-estradiol-treated women had a significantly greater response, both statistically and clinically, than the placebo group. However, because 32 women in this trial were also being treated with antidepressants, the effectiveness of 17β-estradiol as monotherapy for PPD cannot be assessed. It is likely that estrogen can act as a potent adjunct for women with PPD also undergoing a treatment with antidepressant drugs. Whether it modulates the effects of abrupt estrogen withdrawal postpartum remains unknown, as does its long-term effects. Perimenopausal Depressions The perimenopause is the transitional period leading up to the menopause. During perimenopause, FSH levels increase while estrogen and progesterone levels decrease – until menopause, when follicules no longer mature, ovulation is absent, menstrual cycles cease, along with their associated estradiol and progesterone secretion, and LH and FSH levels are continuously high due to the lack of gonadal feed-back mechanism. Most women usually experience perimenopause during their late 1940 to early 1950. It is marked by irregular menstrual cycles and may also include symptoms such as hot flashes, night sweats, insomnia

158

The Effect of Estrogens on Depression

and other sleep disturbances, vaginal dryness, as well as possible mental sluggishness, irritability, and decreased libido. At one time, the menopause, and especially the perimenopause, were associated with severe depression. This belief was furthered by Kraepelin, who described “Involutional melancholia” (1907) as a very severe depression with hypochondrial delusions among women with no prior history of depression. This was later classified as a form of manic-depressive illness. Epidemiological studies during the 1970s questioned and then dismissed the prevalence of new-onset depressions during the perimenopause and menopause (Weissman and Klerman 1977; Weissman 1979). Based on recent epidemiological studies, roughly 10% of women experience mood symptoms or changes which can be directly linked to perimenopause (Schmidt et al. 1997). Mood symptoms tend to be more commonly reported among perimenopausal women than menopausal women and the ratio of affective disorders is higher (from 2:1 to 3–4 : 1) among women than men during midlife. This suggests a strong relationship between the perimenopause and the onset of mood symptoms. Further support for this suggestion is provided by Angst et al. (1980), who reported a second peak of onset of bipolar affective disorder in women aged 45–55 but not in men of the same age. It is conceivable that a relationship exists between the hormonal instability or substantial hormonal changes during perimenopause and exacerbations of bipolar disorder and other affective disorders (Halbreich 1997). Treatment with Estrogen There have been few studies of female depressions during midlife that have focused only on perimenopausal women (Thompson and Oswald 1977; Schmidt et al. 2000). Most studies of female depressions at midlife have included both depressed perimenopausal and postmenopausal women (Coope and Poller 1975; Coope 1981; Aylward et al. 1974; Montgomery et al. 1987; Strickler et al. 1977). Several researchers have documented beneficial effects from estrogen administration (Schmidt et al. 2000; Aylward et al. 1974; Montgomery et al. 1987). Aylard et al. (1974) found that estrone sulfate improved mood, and that this correlated with an increase in free tryptophan levels. Montgomery’s group (Montgomery et al. 1987) compared 17β-estradiol alone or 17β-estradiol in conjunction with

159

Linda S. Kahn and Uriel Halbreich

testosterone, to placebo, and noted improvement in both the periand postmenopausal women in all treatment groups, with significantly greater improvement in the active treatment groups. In both peri- and postmenopausal women, the benefit from hormonal treatment was lost by 4 months – perhaps because either hormonal levels were diminished at this point or because placebo-treated patients continued to improve. Schmidt et al. (2000) administered transdermal 17β-estradiol (Estraderm 0.05 mg/d) to 34 women in a recent randomized, doubleblind trial. After 3 weeks, a positive response was noted among women undergoing estradiol treatment. After 6 weeks, a full or partial therapeutic response was reported by 80% of subjects receiving estradiol, compared to 22% of those on placebo. Those trials not showing benefit of estrogen over placebo were using conjugated equine estrogen (Coope 1981; Montgomery et al. 1987; Strickler et al. 1977; Coope 1981) and estrone sulfate (Thomson and Oswald 1977). Several, but not all, of these trials included women with comorbid psychiatric illnesses (Strickler et al. 1977; Coope 1981) who were permitted to continue psychotropic medication throughout the trials. Thus, the results are difficult to interpret due to the possibility that the underlying psychopathology among these women played a greater role in precipitating and exacerbating mood symptoms that did the hormonal changes brought on by menopause. Several other studies also reported a high nonspecific response rate with significant improvement over baseline in both the estrogen and placebo conditions (Coope 1981; Thompson and Oswald 1977). Menopausal Depressions Natural menopause is generally defined as starting 1 year after a woman’s last menstrual period and usually occurs by 55 years of age (the mean age is 51) in industrial societies. Surgical menopause occurs after a bilateral oophorectomy. During menopause, sex hormones’ feedback inhibition ceases, resulting in decreased gonadal estrogen and progesterone secretions and gonadotropin (LH and FSH) levels. Postmenopause estrogen levels drop to less than 20 ng/l (compared to between 50–250 ng/l in normally menstruating women). Generally, the main indication of menopausal state occurs when levels of follicle-stimulating hormone (FSH) increase to over 20 ng during the perimenopausal period and over 40 ng during menopause (Kouri and Halbreich 1998).

160

The Effect of Estrogens on Depression

After menopause, estrone becomes the most active form of estrogen. It is derived from the peripheral conversion of adrenal androstanedione (Kouri and Halbreich 1998), while estradiol is produced mainly in the ovaries. A number of multisystemic changes accompany the decrease in circulating estrogen during menopause. These occur not only in the reproductive system, but also in bone mineral density, in the cardiovascular system, as well as in the CNS (Kouri and Halbreich 1998). Although hormonal decrease is associated with decreases in selected cognitive functions, depression is not necessarily increased. Estrogen for Treating Dysphorias Among Perimenopausal and Menopausal Women Although estrogen is most often prescribed for alleviation of the vasomotor and physical symptoms of perimenopause, evidence for its therapeutic efficacy in ameliorating dysphoric symptoms during menopause is not clear. Over 20 placebo-controlled studies have evaluated the therapeutic efficacy of estrogen for depressive symptoms or depressive disorders during the menopause. Estrogen for Treating Dysphorias Among Surgically Menopausal Women Estrogen may be more beneficial than placebo for improving mood in women who have undergone surgical menopause (Ditkoff et al. 1991; Sherwin and Gelfand 1985; Dennerstein and Burrows 1979). The estrogen preparations used in these positive studies included conjugated estrogens (Ditkoff et al.; Lobo 1991), ethinyl estradiol and estradiol valerate (Sherwin and Gelfand 1985), while two negative trials with smaller sample sizes employed either estrogen sulfate (Coppen et al. 1977) or conjugated estrogen (George et al. 1973). Sherwin et al. (1985) compared placebo, estradiol valerate, testosterone enanthate, or a combination of the two preparations in oophorectomized women who had mood baseline levels equivalent to a hysterectomized but nonoophorectomized control group – the placebo group reported a deterioration in mood whereas the three hormonal treatment groups reported improved mood. Although the testosterone-only group experienced the most significant decrease in mood symptoms, they also reported increased hostility scores. Ditkoff et al. (1991) undertook a trial which excluded women

161

Linda S. Kahn and Uriel Halbreich

suffering from either substantial mood or physical symptoms. The researchers found that women taking conjugated estrogens still showed a more improved mood than the placebo group. It is suggested that estrogen is superior to placebo in alleviating depressed mood among women who have undergone surgical menopause. Estrogen appears to effectively lift mood and improve well-being. Testosterone probably amplifies this effect. Estrogen for Treating Dysphorias Among women with Naturally Occurring Postmenopause Most reports (Brincat et al. 1984; Deerman et al. 1995; Fedor-Freybergh 1977; Furuhjelm et al. 1984; Wiklund et al. 1993) have demonstrated that estrogen (mostly parenteral and oral estradiol) is more effective than placebo in alleviating physical and emotional symptoms of menopause. 17β-estradiol administered parenterally with testosterone was significantly better than placebo for all menopausal symptoms evaluated (somatic as well as psychological), with the exception of aches and pains (Brincat et al. 1984). Results of a large study (n = 223) support these findings. Wiklund et al. (1993) compared transdermal estradiol to placebo in a group of mildly symptomatic women. A dose of 50 µg 17β-estradiol proved more efficacious than placebo for psychological well-being, anxiety, and depressive symptoms, as well as vasomotor symptoms. Another study raised questions regarding estrogen’s relative benefits over placebo. Patterson (1982) compared placebo to administration of a synthetic estrogen, mestranol, followed by addition of norethisterone. Active treatment was superior to placebo only for night sweats and flushing but not for mood symptoms. A high nonspecific response rate (placebo response) was found among postmenopausal women with MDD (Saletu et al. 1995), who were given either 50 µg of transdermal 17β-estradiol or placebo and in minimally symptomatic women receiving either conjugated estrogen or placebo (Campbell 1976; Campbell and Whitehead 1977). Gerdes et al. (1979) administered HRT with conjugated estrogen daily and 14 days of medroxyprogesterone was compared with clonidine, and a no treatment group, and showed improvement of the HRT group on a variety of personality measures and on a self-report mood measure. Long-term estrogen therapy may provide greater mood-enhancing effects than short-term administration. In one longitudinal study,

162

The Effect of Estrogens on Depression

women continuing on estrogen for 1 year reported greater improvement in mood and sexual functioning, compared to those who discontinued treatment (Maoz and Durst 1980). Palinkas and BarrettConnor (1992) conducted a cross-sectional epidemiological study of nearly 1200 women and reported more frequent depressive symptoms in the 50-60 group than in the over 60 group. The authors suggest that symptomatic women sought treatment and thus showed higher initial rates in the 50- to 60-year-old hormone replacement group. Over time and with treatment, these women felt better and sought treatment less.

Summary Although much is known about the multiple actions and interactions of estrogens, less is known about its specific role in each dysphoric state. Several compelling facts suggest that estrogen or its absence play a major role in the onset and course of depressions in women. (1) From menarche through perimenopause, a significant gender difference exists in the prevalence of depressions. This distinction does not exist among prepubescent children, nor between postmenopausal women and older men. (2) The gender difference in prevalence of depressions is especially apparent during periods when estrogen levels are high but unstable. This is caused by cyclicity during the menstrual cycle and increases in estrogen levels during pregnancy, and sharp withdrawals in the postpartum period. (3) During periods of estrogen change (premenstrual, postpartum, and perimenopause), depressive symptoms are also more prevalent – although the direct effects of estrogen cannot be clearly distinguished from other factors, such as environmental influences, etc. (4) During menopause, when estrogen is almost absent, there is no increase in the prevalence of first-episode depression. It is plausible that hormonal fluctuations or lack of estrogen increase the risk of depression among vulnerable women. Based on current knowledge, treatment of depression with estrogen may: (1) stabilize and restore disrupted homeostasis – as during postpartum, premenstrual, or perimenopausal conditions; (2) act as a psychomodulator to offset vulnerability to dysphoric mood when estrogen levels are significantly decreased, as in the case of postmenopausal women.

163

Linda S. Kahn and Uriel Halbreich

There is evidence demonstrating estrogen’s beneficial effects in alleviating mood symptoms during the menopause and improving well-being and sexual drive. Based on reports from placebo-controlled studies, the most consistent effects of estrogen were found in surgically menopausal women or those with mild symptoms. Transdermal β-estradiol is the preparation most highly associated with positive mood changes, even though conjugated estrogen remains the most widely prescribed. Surgically menopausal women appear to benefit most from estrogen therapy for prevention and treatment of mood disturbances. Among these women, the type of estrogen preparation administered appears to matter less and the addition of testosterone or other androgens might be beneficial. However, there are situations that might warrant changing to a preparation that delivers 17β-estradiol: (1) If a woman fails to respond to conjugated estrogens or to other preparations that predominantly deliver estrone, and (2) if there are no contraindications to parenteral or transdermal estrogen. At present no definitive support exists for the use of estrogen as monotherapy for severe depression in menopausal women.

References Abramowitz ES, Baker AH, Fleiscler S (1982) Onset of depressive psychiatric crisis and the menstrual cycle. Am J Psychiatry 139: 475-478 Angold A, Costello EJ, Erkanli A, Worthman CM (1999) Pubertal changes in hormone levels and depression in girls. Psychol Med 29: 1043-53 Angold A, Costello EJ, Worthman CM (1998) Puberty and depression: the roles of age, pubertal status and pubertal timing. Psychol Med 28: 51-61 Angst J, Frey R, Lohmeyer R, Zerben-Rubin E (1980) Bipolar manic depressive psychoses: results of a genetic investigation. Hum Genet 55: 237-254 Appleby L, Gregoire A, Platz C (1994) Screening women for high risk postnatal depression. J Psychosom Res 38: 539-545 Archer JS (1999) Relationship between estrogen, serotonin, and depression. Menopause 6: 71-78 Aylward M, Holly F, Parker RI (1974) An evaluation of clinical response to piperazineoestronesulphate (“Harmogen”) in menopausal patients. Curr Med Res Opin: 417 Backstrom T, Landgren S, Zetterland B (1984) Effects of ovarian steroid hormones on brain excitability and their relation to epilepsy seizure, variation during the menstrual cycle. In: Porter RJ, Mattson RH, Ward AAJ (eds) Advances in epileptology. Raven Press, New York, 269-277

164

The Effect of Estrogens on Depression Birge S (1997) The role of estrogen in the treatment of Alzheimer’s disease. Neurology 48 (Suppl 7): S36-S41 Bloch M, Schmidt PJ, Danaceau M, Murphy J, Nieman L, Rubinow DR (2000) Effects of gonadal steroids in women with a history of postpartum depression. Am J Psychiatry 157: 924-930 Brayshaw ND, Brayshaw DD (1986) Thyroid hypofunction in premenstrual syndrome. N Engl J Med 315: 1486 Brincat M, Studd JW, O’Dow W, Magos A, Cardozo LD, Wardle PJ (1984) Subcutaneous hormone implants for the control of climetric symptoms. Lancet: 160-168 Brockington IF, Kelly A, Hall P (1988a) Premenstrual relapse puerperal psychosis. J Affect Disord 14: 287-292 Brockington IF, Margison FR, Schofield EM, Knight RJE (1988b) The clinical picture of the depressed form of puerperal psychosis. J Affect Disord 15: 29-37 Burvill PW (1995) Recent progress in the epidemiology of major depression. Epidemiol Rev 17: 21-31 Campbell S, Whitehead M (1977) Oestrogen therapy and the menopausal syndrome. Clin Obset Gynaecol 4: 31-47 Campbell S (1976) Double blind psychometric studies on the effects of natural estrogens on post-menopausal women. In: Campbell S (ed) The management of menopause and postmenopausal years. MTP, Lancaster, 149-158 Carothers AD, Murray L (1990) Estimating psychiatric morbidity by logistic regression: application to post-natal depression in a community sample. Psychol Med 20: 695-702 Casper RF, Patel-Christopher A, Powell A (1989) Thyrotropin and prolactin responses to thyrotropin-releasing hormone in premenstrual syndrome. J Clin Endocrinol Metab 68: 608-612 Chakravorty SG, Halbreich U (1997) The influence of estrogen on monoamine oxidase activity. Psychopharmacol Bull 33: 229-233 Coope I (1981) Is oestrogen therapy effective in the treatment of menopausal depression. J R Coll Gen Pract 31: 134-140 Coope I (1981) The menopause. Is oestrogen therapy affective in the treatment of postmenopausal depression. J R Coll Gen Pract 31: 134-140 Coope J, J.M. T, Poller L (1975) Effects of “natural oestrogen” replacement therapy on menopausal symptoms and blood clotting. BMJ 4: 139-143 Cooper PJ, Murray L, Hooper R, West A (1996) The development and validation of a predictive index for postpartum depression. Psychol Med 26: 627-634 Coppen A, Bishop M, Beard RJ (1977) Effects on piperazine oestrone sulphate on plasma tryptophan, oestrogens, gonadotrophins and psychological functioning in women following hysterectomy. Curr Med Res Opin 4: 29-36 Cox JL, Connor Y, Kendell RE (1982) Prospective sudy of the psychiatric disorders of childbirth. Br J Psychiatry 140: 111-117 Dahr V, Murphy BEP (1990) Double-blind randomized crossover trial of luteal phase estrogens (Premarin®). Psychoneuroendrocrinology 15: 489-493 Dalton K (1964) The premenstrual syndrome. Heinemann, London Demyttenaere K, Lenaerts H, Nijs P, Van Assche FA (1995) Individual coping style and psychological attitudes during pregnancy predict depression levels during pregnancy and during postpartum. Acta Psychiatr Scand 91: 95-102

165

Linda S. Kahn and Uriel Halbreich Dennerstein L, Burrows GD (1979) Affect and the menstrual cycle. J Affect Disord 1: 77-92 Derman RJ, Dawood MY, Stone S (1995) Quality of life during sequential hormone replacement therapy – a placebo-controlled study. Int J Fertil Menopausal Stud 40: 73-78 Dickinson SL, Curzon G (1986) 5-Hydroxytryptamine-mediated behavior in male and female rats. Neuropharmacology 25: 771-776 Ditkoff EC, Crary WG, Cristo M, Lobo R (1991) Estrogen improves psychological function in asymptomatic postmenopausal women. Obstet Gynecol 78: 991-995 d’Orban PT, Dalton J (1980) Violent crime and the menstrual cycle. Psychol Med 10: 353-359 Endicott J, Halbreich U (1988) Clinical significance of premenstrual dysphoric changes. J Clin Psychiatry 49: 486-489 Epperson CN, Wisner KL, Yamamoto B (1999) Gonadal steroids in the treatment of mood disorders. Psychosom Med 61: 676-697 Fedor-Freybergh (1977) The influence on oestrogen on wellbeing and mental performance in climacteric postmenopausal women. Acta Obstet Gynecol Scand 64: 2-10 Frank R (1931) The hormonal causes of premenstrual tension. Arch Neurol Psychiatry 26: 1053-1057 Freeman E, Rickels K, Sondheimer SJ, Polansky M (1990) Ineffectiveness of progesterone suppository treatment for premenstrual syndrome: a placebo-controlled, randomized, double-blind study. JAMA 264: 349-353 Freeman EW, Sondheimer SJ, Rickels K, Weinbaum PJ (1985) PMS treatment approaches and progesterone therapy. Psychosomatics 26: 811-816 Friedman RC, Hurt SW, Charkin J (1982) Sexual histories and premenstrual affective syndrome in psychiatric inpatients. Am J Psychiatry 139: 1484-1486 Furuhjelm M, Karlgren E, Carlstrom K (1984) The effect of estrogen therapy on somatic and psychical symptoms in postmenopausal women. Acta Obstet Gynecol Scand 63: 655-661 Garlow SJ, Musselman DJ, Nemeroff CB (1999) The neurochemistry of mood disorders: clinical studies. In: Charney DS, Nestler EJ, Bunney BS (eds) Neurobiology of mental illness. Oxford University Press, New York George GCW, Utian WH, Beumont PJV, Bearwood CJ (1973) Effect of exogenous oestrogens on mionor psychiatric symptoms in postmenopausal women. S Afr Med J 47: 2387-2388 Gerdes LC, Sonnendecker EWW, Polakow ES (1979) Psychological changes effected by estrogen-progestogen and chlonidine treatment in climacteric women. Am J Obstet Gynecol 1: 77-92 Gladis MM, Walsh BT (1987) Premenstrual exacerbation of binge eating in bulimia. Psychiatry 144: 1592-1595 Gregoire AJ, Kumar R, Everitt B, Henderson AF, Studd JW (1996) Transdermal oestrogen for treatment of severe postnatal depression. Lancet 347: 930-3 Gurguis GN, Yonkers KA, Blakeley JE, Phan SP, Williams A, Rush AJ (1998a) Adrenergic receptors in premenstrual dysphoric disorder. II. Neutrophil beta2-adrenergic receptors: Gs protein coupling, phase of menstrual cycle and prediction of luteal phase symptom severity. Psychiatry Res 79: 31-42

166

The Effect of Estrogens on Depression Gurguis GN, Yonkers KA, Phan SP, Blakeley JE, Williams A, Rush AJ (1998b) Adrenergic receptors in premenstrual dysphoric disorder: I. Platelet alpha2 receptors: Gi protein coupling, phase of menstrual cycle, and prediction of luteal phase symptom severity. Biol Psychiatry 44: 600-609 Halbreich U, Alt IH, Paul L (1988) Premenstrual changes. Impaired hormonal homeostasis. Neurol Clin 6: 173-94 Halbreich U, Ben-David M, Assael M, Bornstein R (1976) Serum-prolactin in women with premenstrual syndrome. Lancet 02: 654-656 Halbreich U, Carson S, Rojansky N, Lane J, Alt E (1988) 20% of women who reported premenstrual dysphoric changes might have subclinical hypothyroidism. Proc Annual Meeting of the American College of Neuropsychopharmacology (ACNP). San Juan, PR Halbreich U, Endicott J, Goldstein S, Nee J (1986) Premenstrual changes and changes in gonadal hormones. Acta Psychiatr Scand 74: 576-586 Halbreich U, Endicott J, Schacht S, Nee J (1982) The diversity of premenstrual changes as reflected in the premenstrual assessment form. Acta Psychiatr Scand 65: 46-65 Halbreich U, Endicott J (1985) Relationship of dysphoric premenstrual changes to depressive disorders. Acta Psychiatr Scand 71: 331-338 Halbreich U, Lumley LA (1993) The multiple interactional biological processes that might lead to depression and gender differences in its appearance. J Affect Disord 29: 159-173 Halbreich U, Piletz JE, Carson S, Halaris A, Rojansky N (1993) Increased imidazoline and alpha 2 adrenergic binding in platelets of women with dysphoric premenstrual syndromes. Biol Psychiatry 34: 676-686 Halbreich U, Rojansky N, Palter S, Tworek H, Hissin P, Wang K (1995) Estrogen augments serotonergic activity in postmenopausal women. Biol Psychiatry 37: 434-441 Halbreich U, Rojansky N, Palter S (1991) Elimination of ovulation and menstrual cyclicity (with danazol) improves dysphoric premenstrual syndromes. Fertil Steril 56: 1066-1069 Halbreich U, Smoller JW (1997) Intermittent luteal phase sertraline treatment of dysphoric premenstrual syndrome. J Clin Psychiatry 58: 399-402 Halbreich U, Tworek H (1993) Altered serotonergic activity in women with dysphoric premenstrual syndromes. Int J Psychiatry Med 23: 1-27 Halbreich U (2000) Gonadal hormones, reproductive age, and women with depression. Arch Gen Psychiatry 57: 1163-1164 Halbreich U (1995) Menstrually related disorders: what we do know, what we only believe that we know, and what we know that we do not know. Crit Rev Neurobiol 9: 163-175 Halbreich U (1997a) Menstrually related disorders – towards interdisciplinary international diagnostic criteria. Cephalalgia 17 (Suppl 20): 1-4 Halbreich U (1997b) Premenstrual dysphoric disorders: a diversified cluster of vulnerability traits to depression. Acta Psychiatr Scand 95: 169-176 Halbreich U (1996) Premenstrual Syndromes. In: Halbreich U (ed) Balliere’s clinical psychiatry: International practice and research. Bailliere Tindall, London, 667-686

167

Linda S. Kahn and Uriel Halbreich Halbreich U (1997c) Psychiatric disorders in women. In: Weller M, van Kammen D (eds) Progress in clinical psychiatry. Saunders, London, 96-115 Halbreich U (1997d) Role of estrogen in postmenopausal depression. Neurology 48: S16-19 Halbreich U (1999) RU-486, estrogen and leuprolide as treatment for PMS [abstract]. 30th Annual Meeting, ISPNE, Orlando, Florida Hamilton JA, Yonkers KA (1996) Sex differences in pharmacokinetics of psychotropic medications: Part II: Effects on selected psychotropics. In: Jensvold M, Halbreich U, Hamilton JA (eds) Psychopharmacology of women: sex, gender, and hormonal considerations. American Psychiatric Press, Washington DC, 43-71 Hamilton JA, Yonkers KA (1996) Sex differences in pharmacokinetics of psychotropic medications: Part I: Physiological basis for effects. In: Jensvold M, Halbreich U, Hamilton JA (eds) Psychopharmacology of women: sex, gender, and hormonal considerations. American Psychiatric Press, Washington DC, 11-41 Harrison WM, Endicott J, Nee J, Glick H, Rabkin JG (1989) Characteristics of women seeking treatment for premenstrual syndrome. Psychosomatics 30: 405-411 Hatotani N, Nishikubo M, Iitayama I (1979) Periodic psychosis in the female and the reproductive process. In: Zichella L, Pancheri P (eds) Psychoneuroendocrinology in reproduction. Elsevier North Holland, New York, 55-68 Hurt SW, Schnurr PP, Severino SK, et al. (1992) Late luteal phase dysphoric disorder in 670 women evaluated for premenstrual complaints. Am J Psychiatry 149: 525-530 Jensvold MF, Hamilton JA (1996) Sex and gender effects in psychopharmacology: contributory factors and implications for pharmacotherapy. In: Halbreich U (ed) Psychiatric issues in women. Bailliere Tindall, London, 647-665 Jensvold MF (1996) Nonpregnant reproductive-age women, Part I: the menstrual cycle and psychopharmacology. In: Jensvold MF, Halbreich UH, Hamilton JA (eds) Psychopharmacology and women. American Psychiatric Press, Washington DC, 139-169 Johnson SR (1987) The epidemiology and social impact of premenstrual symptoms. Clin Obstet 30: 369-384 Kendler KS, Eaves LJ, Walters EE, Neale MC, Heath AC, Kessler RC (1996) The identification and validation of distinct depressive syndromes in a populationbased sample of female twins. Arch Gen Psychiatry 53: 391-399 Kessler R (2000) Gender and mood disorders. In: Goldman MB, Hatch MC (eds) Women and health. Academic Press, San Diego, 989-996 Kessler RC, McGonagle KA, Zhao S, et al. (1994) Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the United States. Results from the National Comorbidity Survey. Arch Gen Psychiatry 51: 8-19 Kessler RC, Walters EE (1998) Epidemiology of DSM-III-R major depression and minor depression among adolescents and young adults in the National Comorbidity Survey. Depress Anxiety 7: 3-14 Kessler RC (1995) Epidemiology of psychiatric comorbidity. In: Tsuang MT, Tohan M, Zahner GEP (eds) Textbook in psychiatric epidemiology. Wiley, New York, 179-197

168

The Effect of Estrogens on Depression Klaiber EL, Broverman DM, Vogel W, Kobayashi Y (1979) Estrogen therapy for severe persistent depressions in women. Arch Gen Psychiatry 36: 550-554 Kornstein SG, McEnany G (2000) Enhancing pharmacologic effects in the treatment of depression in women. J Clin Psychiatry 61 (Suppl 11): 18-27 Kornstein SG, Schatzberg AF, Thase ME, et al. (2000) Gender differences in chronic major and double depression. J Affect Disord 60: 1-11 Kornstein SG, Schatzberg AF, Thase ME, et al. (2000) Gender differences in treatment response to sertraline versus imipramine in chronic depression. Am J Psychiatry 157: 1445-1452 Kornstein SG, Schatzberg AF, Yonkers KA, et al. (1995) Gender differences in presentation of chronic major depression. Psychopharmacol Bull 31: 711-718 Kornstein SG (1997) Gender differences in depression: implications for treatment. J Clin Psychiatry 58 (Suppl 15): 12-18 Kouri E, Halbreich U (1998) Psychotrophic effects of hormonal replacement therapy. Drugs of Today 34: 251-257 Kumar R, Robson KM (1984) A prospective study of emotional disorders in childbearing women. Br J Psychiatry 144: 35-47 Leibenluft E, Hardin TA (1995) Gender differences in seasonal affective disorder. Depression 3: 13-19 LeMeledo JM, Bradwein J, Koszycki D, Bichet D (1995) Premenstrual dysphoric disorder and response to cholecystokinin-tetrapeptide. Arch Gen Psychiatry 52: 605-606 Lepage P, Steiner M (1991) Gender and serotonergic dyregulation: implications for late luteal phase dysphoric disorder. In: Cassano GB, Akiskal HS (eds) Serotonin related psychiatric syndromes: Clinical and therapeutic links. Roy Soc Med Service Congress Symp Series No 165, London Lewinsohn PM, Rohde P, Seeley JR (1998) Major depressive disorder in older adolescents: prevalence, risk factors, and clinical implications. Clin Psychol Rev 18: 765-794 Luggin R, Bernsted L, Petersson B, Jacobsen AT (1984) Acute psychiatric admission related to the menstrual cycle. Acta Psychiatr Scand 69: 461-465 Luine VN (1985) Estradiol increases choline acetyltransferase activity in specific basal forebrain nuclei and projection areas of female rats. Exp Neurol 89: 484490 Magos AL, Brincat M, Studd JWW (1986) Treatment of the premenstrual syndrome by subcutaneous oestradiol implants and cyclical oral norethisterone: Placebo controlled study. BMJ 292: 1629-1633 Maoz B, Durst N (1980) The effects of oestrogen therapy on the sex life of postmenopausal women. Maturitas 2: 327-336 Matsumoto A, Arai Y, Osanai M (1985) Estrogen stimulates neuronal plasticity in the deafferented hypothalamic arcuate nucleus in aged female rats. Neurosci Res 2: 412-418 McEwen BS, Alves SE, Bulloch K, Weiland NG (1998) Clinically-relevant basic sciences studies of gender differences and sex hormones effects. Psychopharmacol Bull 34: 251-259 McEwen BS, Alves SE, Bulloch K (1997) Ovarian steroids and the brain: Implications for cognition and aging. Neurology 48 (Suppl 7): S8-S15

169

Linda S. Kahn and Uriel Halbreich McEwen BS, Woolley CS (1994) Estradiol and progesterone regulate neuronal structure and synaptic connectivity in adult as well as developing brain. Exp Gerontol 29: 431-436 McEwen BS (1991) Non-genomic and genomic effects of steroids on neural activity. Trends Pharmacol Sci 12: 141-147 Merikangas KR, Foeldenyl M, Angst J (1993) The Zurich Study. XIX. Patterns of menstrual disturbances in the community: results of the Zurich Cohort Study. Eur Arch Psychiatry Clin Neurosci 243: 23-32 Michael C, Kantor H, Shore H (1970) Further psychometric evaluation of older women – the effect of estrogen administration. J Gerontol 25: 337-341 Montgomery JC, Appleby L, Brincat M, et al. (1987) Effect of oestrogen and testosterone implants on psychological disorders in the climacteric. Lancet 1: 297299 Murray CJL, Lopez AD (1996) The global burden of disease: a comprehensive assessment of mortality and disability from diseases, injuries, and risk factors in 1990 and projected to 2020. Global burden of disease and injury series. The Harvard School of Public Health, on behalf of the World Health Organization and the World Bank, Boston, 43 Muse K, Cetel N, Futterman L, Yen SC (1984) The premenstrual syndrome: effects of ‘medical ovariectomy’. N Eng J Med 311: 1345-1349 O’Hara MW, Swain AM (1996) Rates and risk of postpartum depression – a metaanalysis. Int Rev Psychiatry 8: 37-54 O’Hara MW, Zekoski EM, Philipps LH, Wright EJ (1990) Controlled prospective study of postpartum mood disorders: Comparison of childbearing and nonchildbearing women. J Abnorm Psychol 99: 3-15 Ohayon MM, Caulet M (1995) Insomnia and psychotropic drug consumption. Progr Neuropsychopharmacol Biol Psychiatry 19: 421-431 Ohkura T, Teshima Y, Isse K (1994) Estrogen increases cerebral and cerebellar blood flow in postmenopausal women. Menopause 2: 13-18 Palinkas LA, Barrett-Connor E (1992) Estrogen use and depressive symptoms in postmenopausal women. Obstet Gynecol 80: 30-36 Parry BL (1989) Reproductive factors affecting the course of affective illness in women. Psychiatr Clin North Am 12: 207-220 Paterson MEL (1982) A randomized double-blind, cross-over study into the effect of sequential mestranol and norethisterone on climateric symptoms and biochemical parameters. Maturitas 4: 83-94 Piletz JE, Halbreich U (2000) Imidazoline and alpha(2a)-adrenoceptor binding sites in postmenopausal women before and after estrogen replacement therapy. Biol Psychiatry 48: 932-939 Prange AJ (1972) Estrogen may well affect response to antidepressant. JAMA 219: 143-144 Price RA, Kidd KK, Weissman MM (1987) Early onset (under age 30 years) and panic disorder as markers etiologic homogeneity in major depression. Arch Gen Psychiatry 44: 434-440 Ramcharan S, Love EJ, Fick GH, Goldfien A (1992) The epidemiology of premenstrual symptoms in a population-based sample of 2650 urban women: Attributable risk and risk factors. J Clin Epidemiol 45: 377-392

170

The Effect of Estrogens on Depression Rawson NSB, D’Arcy C (1991) Sedative hypnotic drug use in Canada. Health Reports 3: 33-57 Reighard FT, Evans ML (1995) Use of the Edinburgh Postnatal Depression Scale in a Southern, rural population in the United States. Progr Neuropsychophamacol Biol Psychiatry 19: 1219-1224 Rivera-Tovar AD, Frank E (1990) Late luteal phase dysphoric disorder in young women. Am J Psychiatry 147: 1634-1636 Roy A, Gang P, Cole K, Rutsky M, Reese L, Weisbord J (1993) Use of Edinburgh Postnatal Depression Scale in a North American population. Progr Neuropsychophamacol Biol Psychiatry 17: 501-504 Roy-Byrne PP, Rubinow DR, Hoban MC, Grover GN, Blank D (1987) TSH and prolactin responses to TRH in patients with premenstrual syndrome. Am J Psychiatry 144: 480 Rubinow DR, Schmidt PJ, Roca CA (1998) Estrogen-serotonin interactions: implications for affective regulation. Biol Psychiatry 44: 839-50 Rubinow DR, Schmidt PJ, Roca CA (1998) Hormone measures in reproductive endocrine-related mood disorders: diagnostic issues. Psychopharmacol Bull 34: 289-290 Saletu B, Brandstatter N, Metkam M, et al. (1995) Double-blind, placebo-controlled, hormonal, syndromal and EEG mapping studies with trandermal oestradiol therapy in menopasual depression. Psychophamacology 122: 321-329 Sandberg D, Endicott J, Harrison W, Nee J, Gorman J (1993) Sodium lactate infusion in late luteal phase dysphoric disorder. Psychiatr Res 46: 79-88 Schmidt PJ, Nieman L, Danaceau MA, et al. (2000) Estrogen replacement in perimenopause-related depression: a preliminary report. Am J Obstet Gynecol 183: 414-420 Schmidt PJ, Nieman LK, Danaceau MA, Adams LF, Rubinow DR (1998) Differential behavioral effects of gonadal steroids in women with and in those without premenstrual syndrome. N Engl J Med 338: 209-216 Schmidt PJ, Roca CA, Bloch M, Rubinow DR (1997) The perimenopause and affective disorders. Semin Reprod Endocrinol 15: 91-100 Schneider LS, Small G, Clary C (1998) Estrogen replacement therapy and antidepressant response to sertraline [abstract]. Annual Meeting of the American Psychiatric Association, Toronto, Ontario Schneider LS, Small GW, Hamilton SH, Bystritsky A, Nemeroff CB, Meyers BS (1997) Estrogen replacement and response to fluoxetine in a multicenter geriatric depression trial. Fluoxetine Collaborative Study Group. Am J Geriatr Psychiatry 5: 97-106 Schneider MA, Brotherton PL, Hailes J (1977) The effect of exogenous oestrogens on depression in menopausal women. Med J Aus 2: 162-163 Seeman MV (2000) Mental illness in women. In: Goldman MB, Hatch MC (eds) Women and health. Academic Press, San Diego, 989-996 Shapira B, Oppenheim G, Zohar J, Segal M, Malach D, Belmaker RH (1985) Lack of efficacy of estrogen supplementation to imipramine in resistant female depressives. Biol Psychiatry 20: 576-579 Sherwin BB, Gelfand MM, Brender W (1985) Androgen enhances sexual motivation

171

Linda S. Kahn and Uriel Halbreich in females: a prospective, crossover study of sex steroid administration in the surgical menopause. Psychosom Med 47: 339-351 Sherwin BB, Gelfand MM (1985) Sex steroids and affect in the surgical menopause: A double-blind, cross-over study. Psychoneuroendoctrinology 10(3): 325-335 Simpkins JW, Singh M, Bishop J (1994) The potential role for estrogen replacement therapy in the treatment of the cognitive decline and neurodegeneration associated with Alzheimer’s disease. Neurobiol Aging 15: S195-S197 Singh M, Meyer EM, Simpkins JW (1994) Ovarian steroid deprivation results in a reversible learning impairment and compromised cholinergic function in Sprague-Dawley rats. Brain Res 644: 305-312 Singh M, Meyer EM, Simpkins JW (1995) The effect of ovariectomy and oestradiol replacement on brain-derived neurotrophic factor messenger ribonucleic acid expression in cortical and hippocampal brain regions of female Sprague-Dawley rats. Endocrinology 136: 2320-2324 Skegg DCG, Doll R, Perry J (1977) Use of medicines in general practice. Br Med J 1: 1561-1563 Steiner M, Born L, Marton P (2000) Menarche and mood disorders in adolescence. In: Steiner M, Yonkers K, Eriksson E (eds) Mood disorders in women. Martin Dunitz, London, 247-268 Steiner M, Steinberg S, Stewart D, et al. (1995) Fluoxetine in the treatment of premenstrual dysphoria. Canadian fluoxetine/premenstrual dysphoria collaborative studygroup. N Eng J Med 332: 1529-1534 Steiner M, Wheadon D, Kreider M, Bushnell W (1993) Antidepressant response to paroxetine by gender (Abstract #462). Ann Meeting American Psychiatric Association, San Francisco Stewart DE, Boydell KM (1993) Psychologic distress during menopause: associations across the reproductive life cycle. Int J Psychiatry Med 23: 157-162 Strickler RC, Borth R, Cecutti A, et al. (1977) The role of oestrogen replacement in the climacteric syndrome. Psychol Med 7: 631-639 Studd JWW, Smith RNJ (1994) Estrogens and depression in women. Menopause 1: 3-7 Stunkard AJ, Fernstrom MH, Price A, Frank E (1990) Direction of weight change in recurrent depression: consistency across episodes. Arch Gen Psych 47: 857-860 Terry DJ, Mayocchi L, Hynes GJ (1996) Depressive symptomatology in new mothers: a stress and coping perspective. J Abnorm Psychol 105: 220-231 Thompson J, Oswald I (1977) Effect of oestrogen on the sleep, mood, and anxiety of menopausal women. BMJ 19: 1317-1319 Troutman BR, Cutrona CE (1990) Nonpsychotic postpartum depression among adolescent mothers. J Abnorm Psychol 99: 69-78 Van Valkingburg C, Akiskal HS, Puzantian V, Rosenthal T (1984) Anxious depression: clinical, family history, and naturalistic outcome – comparisons with panic and major depressive disorders. J Affect Disord 6: 67-82 Watson NR, Studd JW, Savvas M, Garnett T, Baber RJ (1989) Treatment of severe premenstrual syndrome with oestradiol patches and cyclical oral norethisterone. Lancet 2: 730-732 Weissman MM, Bland RC, Canino GJ, et al. (1996) Cross-national epidemiology of major depression and bipolar disorder. JAMA 276: 293-299

172

The Effect of Estrogens on Depression Weissman MM, Klerman GL (1977) Sex differences and the epidemiology of depression. Arch Gen Psychiatry 34: 98-111 Weissman MM, Livingston BM, Leaf PJ (1991) Affective disorders. In: Robins LN, Regier DA (eds) Psychiatric disorders in America. Free Press, New York Weissman MM (1979) The myth of involutional melancholia. Jama 242: 742-744 West ED, Dally PJ (1992) Effect of iproniazid in depressive syndromes. BJM 1: 1491-1494 Whiffen VE (1002) Is postpartum depression a distinct diagnosis? Clin Psychol Rev 12: 485-508 Whiffen VE (1988) Vulnerability of postpartum depression: a prospective multivariate study. J Abnorm Psychol 97: 467-474 Wikander I, Sundblad C, Andersch B et al. (1998) Citalopram in premenstrual dysphoria: is intermittent treatment during luteal phases more effective than continuous medication throughout the menstrual cycle? J Clin Psychopharmacol 18: 390-398 Wiklund I, Karlberg J, Mattson L (1993) Quality of life menopausal women on a regimen of transdermal estradiol therapy: A double-blind placebo-controled study. Am J Obstet Gynecol 168: 824-830 Wolley CS, McEwen BS (1992) Estradiol mediates fluctuation in hippocampal synapse density during the estrous cycle in the adult rat. J Neurosci 12: 25492554 Woods NF, Most A, Dery GK (1982) Prevalence of perimenstrual symptoms. Am J Public Health 72: 1257-1264 Yonkers KA, Halbreich U, Freeman E, Brown C, Pearlstein T (1996) Sertraline in the treatment of premenstrual dysphoric disorder. Psychopharmacol Bull 32: 41-46 Yonkers KA, Kando JC, Cole JO, Blumenthal S (1992) Gender differences in pharmacokinetics and pharmacodynamics of psychotropic medication. Am J Psychiatry 149: 587-595 Yonkers KA, Kando JC, Hamilton JA, Halbreich U (2000) Gender differences in treatment of depression and anxiety. In: Halbreich U, Montgomery SA (eds) Pharmacotherapy for mood, anxiety, and cognitive disorders. American Psychiatric Press, Washington DC, 59-74 Yonkers KA, Zlotnick C, Allsworth J, Warshaw M, Shea T, Keller MB (1998) Is the course of panic disorder the same in women and men? Am J Psychiatry 155: 596-602 Yonkers KA (1998) Assessing unipolar mood disorders in women. Psychopharmacol Bull 34: 261-266 Young E, Korszun A (1998) Psychoneuroendocrinology of depression. Hypothalamicpituitary-gonadal axis. Psychiatr Clin North America 21: 309-323 Young EA, Midgley AR, Carlson NE, Brown MB (2000) Alteration in the hypothalamic-pituitary-ovarian axis in depressed women. Arch Gen Psychiatry 57: 1157-1162 Zelkowitz P, Milet TH (1995) Screening for post-partum depression in a community sample. Can J Psychiatry 40: 80-86

173

8 Estrogens and Other Hormones in the Treatment of Premenstrual Syndromes Leslie Born and Meir Steiner

Introduction Premenstrual Syndrome (PMS) can be defined as a pattern of emotional, behavioral, and physical symptoms that occur premenstrually and remit after menses (World Health Organization 1996). These symptoms typically include minor mood changes such as irritability, tension, or depressed mood, and physical complaints such as breast tenderness, bloating, cramps, or headache. Premenstrual dysphoric disorder (PMDD) denotes PMS with prominent mood symptoms (irritability, tension, dysphoria, or affective lability) severe enough to markedly affect work, social activities, or relationships with others (American Psychiatric Association 1994). The etiology of PMS and PMDD is still largely unknown, though there is accumulating evidence that these disorders are primarily biological phenomena. For example, investigators have shown the elimination of premenstrual complaints with suppression of ovulation (Muse et al. 1984; Schmidt et al. 1998) or surgical menopause (Casper and Hearn 1990; Casson et al. 1990) and an absence of symptoms during nonovulatory cycles (Hammarbäck et al. 1991). Additionally, there is recent, convincing evidence of the heritability of premenstrual symptoms (Kendler et al. 1998). The role of the female sex hormones in premenstrual symptomatology has long been considered of central importance. That both estrogen and progesterone are important in the manifestation of premenstrual symptoms has been shown in studies of postmenopausal hormone replacement therapy (Hammarbäck et al. 1985; Henshaw et al. 1996) and in a recent animal model of PMS (Ho et al. 2001). Women who received estrogen alone did not show any cyclical worsening in mood or physical symptoms (Hammarbäck et al. 1985).

175

Leslie Born and Meir Steiner

It was previously thought that premenstrual complaints were induced by hormone deficiency, i.e., a decrease in serum levels of estradiol, progesterone, or estradiol and progesterone, in the late luteal phase. Severe PMS or PMDD was regarded as a manifestation of steroid withdrawal. This premise has been challenged by the finding of no differences in the levels or patterns of secretion of progesterone, estradiol, follicle-stimulating hormone, luteinizing hormone, testosterone-estradiol-binding globulin, dehydroepiandrosterone sulfate, dihydrotestosterone, prolactin or cortisol in women with or without PMS (Rubinow et al. 1988). By the late 1990s, there was some discussion in the literature as to whether the hormonal changes occurring during the late luteal phase are in fact relevant to the onset of premenstrual symptoms. One line of thinking posited that premenstrual symptoms begin with the high mid-cycle peak of progesterone and/or estradiol and manifest with a delay of a few days or a week, independently of the subsequent changes in serum hormonal levels (Eriksson et al. 2000a). Another group examined whether the luteal phase is necessary for the expression of PMS and experimentally showed a follicular phase onset of PMS (Schmidt et al. 1991; Rubinow et al. 1998). The hypothalamic-pituitary-gonadal axis in women with severe PMS and PMDD is, contrary to previous beliefs, apparently functioning normally, with normal estrogen and progesterone levels (Schmidt et al. 1998). The current consensus seems to be that normal ovarian function, rather than simple hormone imbalance, is the cyclic trigger for biochemical events within the central nervous system and other target tissues which unleash premenstrual symptoms in vulnerable women (Roca et al. 1996). That is, in susceptible women there is an enhanced responsiveness to the interaction of gonadal steroids with neurotransmitters, in particular the serotonergic system (Steiner et al. 2002). In general, two treatment strategies are being used in an attempt to change this interaction. One is the suppression of ovulation, which stops the cyclical hormonal fluctuations. The second – and more current – approach targets correction of possible changes in neurotransmitter sensitivity by using psychotropic medications, notably serotonin reuptake inhibitors. It is generally recognized that women with PMS or PMDD are likely to be offered one of these two treatment pathways, i.e., hormonal or psychopharmacological management, depending on the setting in which they are seen. The focus

176

Estrogens and Other Hormones in the Treatment of Premenstrual Syndromes

of this review is on hormonal treatment strategies, i.e., the use of estrogen and other hormones to treat PMS.

Estradiol Estrogen has been considered a major malefactor of PMS for over 60 years (Frank 1931). Yet, studies of estrogen in the treatment of PMS are relatively few. Unlike progesterone, which has been long known to have sedative properties, it was thought that estrogen increased energy (Dhar and Murphy 1991). It is now known that estrogen exerts multiple effects on the central nervous system through neurotransmitter receptor activation in estrogen-responsive neurons, leading to elevated mood, increased activity, and antidepressant effects (Halbreich and Kahn 2001). Two factors influence interpretation across studies of estrogen preparations for the treatment of PMS. One is the form of estrogen administration, the other is whether the estrogen is administered alone (unopposed) or combined with a progestin. Trials of estrogen for the treatment of PMS have included implant, transdermal patch, and oral modes of estrogen delivery. Oral estrogens are extensively metabolized in the liver (first-pass effect) and circulate primarily as estrone sulfate. This results in limited oral potency. Estrogen drug products administered by nonoral routes are not subject to true first-pass metabolism, but do undergo significant hepatic uptake, metabolism, and enterohepatic recycling. The transdermal delivery systems provide controlled, sustained nearconstant release of effective estradiol, predictable absorption, lack of gastrointestinal adverse effects, and are associated with a slower metabolism to less active estrogens than oral preparations (Scott et al. 1991). Consequently, lower dosages of estradiol need to be given transdermally to achieve therapeutic plasma levels. Transdermal administration may also render a greater therapeutic benefit for mood compared with oral estrogens, as the former provides minimal fluctuation of plasma estrogen levels and higher levels of estradiol, with a close physiological estrone to estradiol ratio (Halbreich and Kahn 2001). The results of the few studies of estrogen treatment for PMS are consistent with this line of thinking – as are studies of estrogen treatment of menopausal women. A much higher proportion of positive outcomes has been noted for parenteral or

177

Leslie Born and Meir Steiner

transdermal administration of estrogen, compared with compounds that predominantly yield estrone (Yonkers et al. 2000). The interpretation of estrogen studies in the treatment of PMS is further complicated by whether the intervention involved addition of continuous or cyclic administration of progestins, or unopposed estrogen. The cyclical addition of a progesterone analogue to estradiol implants or transdermal estradiol patches to induce endometrial shedding may lead to a relapse of premenstrual-like complaints (Magos et al. 1986; Watson et al. 1989). A combination of estradiol patches and a progesterone-releasing intrauterine device as a treatment for premenstrual dysphoria has been proposed, but remains to be evaluated in controlled trials (O’Brien et al. 1995; see chap. 14). Studies in which women were administered continuous (i.e., during both the follicular and luteal phases) estradiol at doses sufficient to suppress ovulation have consistently demonstrated the superiority of estrogen to placebo in the treatment of both physical and psychological symptoms of prospectively diagnosed PMS. This includes transdermal 17β-estradiol (Watson et al. 1989) and Estraderm TTS (100-µg dose) (Smith and Zamblera 1995), and subcutaneous estradiol implants (Magos et al. 1986). A single trial of luteal phase oral conjugated estrogen (Premarin) showed an aggravation, rather than a resolution, of mental or physical premenstrual symptoms in a sample of 11 women (Dhar and Murphy 1990). In one case, a 100-mg estradiol/ 100-mg testosterone implant with cyclical norethisterone was successfully used to treat recurrent premenstrual depressive symptoms (Crammer 1986). Following treatment, the subject’s elevated plasma cortisol levels returned to normal, and she remained completely well for the next 8 years. Selective estrogen receptor modulators (SERMs), “compounds that selectively interact with the estrogen receptor to produce targeted tissue-dependent agonist or antagonist effects,” have been indicated for the treatment and/or prevention of breast and endometrial cancer, osteoporosis, and coronary heart disease (Halbreich and Kahn 2000). In women of reproductive age, there is most likely an antagonistic effect between SERM and estrogen. A greater understanding, though, of the mechanisms of estrogen effects on the CNS portends the development of a SERM with a sustained estrogen agonist effect on the brain and a possible treatment for PMS.

178

Estrogens and Other Hormones in the Treatment of Premenstrual Syndromes

Progesterone Progesterone is perhaps the most widely known, albeit polemical treatment for PMS. The rationale for progesterone therapy has been the ostensible relationship between PMS symptoms and cyclic changes in plasma progesterone levels and the presumed low levels of progesterone in women with PMS (Dalton 1990). Until the early 1990s, progesterone pessaries or suppositories were among the most widely prescribed treatments for PMS in North America and the United Kingdom (Wyatt et al. 2001). Yet, there is neither consistent evidence that plasma progesterone, allopregnanolone, or pregnanolone levels are altered in women with PMS (Rubinow et al. 1988; Schmidt et al. 1994; Wang et al. 1996; Bicikova et al. 1998; Monteleone et al. 2000; Girdler et al. 2001; Freeman et al. 2002) nor any indication that progesterone or progesterone receptors are mediators of PMS (Chan et al. 1994). Notwithstanding, there is recent evidence to suggest progesterone and the neurosteroids may have an important role in the pathogenesis of these disorders (Epperson et al. 1999; 2002; Sundstrom-Poromaa et al. 2002). Progesterone may have anxiolytic properties through the action of its metabolites at the GABA-A receptors by binding to the GABA receptor complex at or near the barbituate binding site (Majewska 1992). This had led some investigators to suggest that there may be subgroup of women who might benefit from progesterone (Rapkin et al. 1997; Freeman et al. 1995) and may help to explain the clinical satisfaction in some women with PMS who were treated with progesterone. A recent meta-analysis of 10 placebo-controlled trials investigating the efficacy of progesterone for the treatment of premenstrual symptoms has unequivocally shown that progesterone is no more beneficial than placebo for the treatment of PMS (Wyatt et al. 2001). Only one study of 23 women found that oral micronized progesterone (at a dose of 3 × 100 mg/day) was superior to placebo for premenstrual mood symptoms (Dennerstein et al. 1985). The most commonly reported side effects for oral micronized progesterone were fatigue or sedation and dizziness (Wyatt et al. 2001). The (slight) efficaciousness and side effects of oral micronized progesterone may be due to a higher concentration of progesterone metabolites, in particular pregnanolone with its notable anxiolytic and hypnotic properties (Freeman et al. 1993).

179

Leslie Born and Meir Steiner

There are two case reports on the successful use of natural progesterone organogel for treatment of premenstrual dysphoria in adolescents, administered in a 25-mg bid dose to the dorsal hand during the late luteal phase (Schaller et al. 2000). Organogel is a matrix for the transdermal transport of drugs (Willimann et al. 1992). The utility of several progestins in the treatment of premenstrual mood symptoms has also been tested, and although the results are not as consistent as those of the trials using progesterone, they are generally negative (Wyatt 2001). The progestin most commonly evaluated was dydrogesterone: the results of two prospective placebo-controlled studies showed no significant difference between placebo and dydrogesterone in the rating of premenstrual symptoms (Williams et al. 1983; Dennerstein et al. 1986). Taken together, the results of several open trials suggested some relief for somatic symptoms; yet interpretation of some of these findings is hindered by the lack of a prospective diagnosis for PMS (Wyatt 2001). Some, but not all, studies with medroxyprogesterone suggest that this compound may be somewhat better than progesterone or dydrogesterone for premenstrual depression or other psychological symptoms and breast discomfort (Jordheim 1972; West 1990; Helberg et al. 1991). Again, it should be noted that in two trials there was no prospective diagnosis of PMS (Jordheim 1972; Helberg et al. 1991). West noted breakthrough bleeding in about three-fourths of the cycles treated with medroxyprogesterone. Two placebo-controlled trials of norethisterone in the treatment of PMS reported significant improvement for breast symptoms only (Ylöstalo et al. 1981; West 1990).

Oral Contraceptives Oral contraceptives (OCs) suppress ovulation while maintaining menstruation through periodic steroid withdrawal. OCs typically contain a combination of ethinylestradiol and a progestin. Premenstrual symptoms have been prospectively documented in women taking oral contraceptives (Bancroft and Rennie 1993; Marriott and Faragher 1986; Sveindóttir and Bäckström 2000; Freeman et al. 2001a). Years ago, there was a common perception that OCs caused mood disorders. A recent review of the literature, however, has found

180

Estrogens and Other Hormones in the Treatment of Premenstrual Syndromes

limited support for OC-induced mood deterioration (Yonkers et al. 2000). The extent to which OCs alleviate, precipitate, or exacerbate premenstrual mood and behavior symptoms remains unresolved (Kahn and Halbreich 2001). The use of oral contraceptives in reducing PMS symptoms has been somewhat controversial, with clinical trials few in number. The more recent findings are, at best, suggestive of some benefit, mostly for the physical symptoms such as increased appetite, cravings, and acne (Freeman et al. 2001b). One group (Graham and Sherwin 1992; 1993) found OCs to be effective in treating breast pain and bloating, but to be no more beneficial than placebo in reducing premenstrual mood symptoms. Another group (Walker and Bancroft 1990) found breast tenderness to be the only symptom showing a clear difference between OC users and nonusers, with monophasic OC users experiencing less discomfort than triphasic users or OC nonusers. Bancroft and Rennie (1993) found that in comparison to nonusers, OC users (monophasic or triphasic) experienced significantly less improvement in negative mood, in particular irritability. Some investigators, on the other hand, have reported OCs to be an effective option for treatment of PMS mood symptoms (Roy-Byrne et al. 1984; Bäckström et al. 1992). Although both monophasic and triphasic preparations have been shown to be beneficial, women taking combined preparations report less severe premenstrual negative moods than those taking triphasic preparations (Bancroft et al. 1987; Bäckström et al. 1992; Bancroft and Rennie 1993). In one study, a monophasic desogestrel pill was more effective in reducing negative mood changes than monophasic or triphasic levonorgestrel pills, while the triphasic preparation was more effective in relieving physical symptoms (Bäckström et al. 1992).

Androgens The influence of androgens on behavior and mood in women is gaining attention. Androgen excess in women has been suggested to be associated with depression, irritability, and compulsive behavior (Eriksson et al. 2000b). Many researchers regard irritability as the cardinal symptom of PMDD (Steiner et al. 1980; Endicott et al. 1999). Ovarian production of androstendione, the major ovarian androgen, is influenced by luteinizing hormone and, hence, fluctuates during

181

Leslie Born and Meir Steiner

the menstrual cycle, with higher levels in the luteal than in the follicular phase. While there are mixed reports on (abnormal or normal) serum levels of testosterone and premenstrual complaints (Eriksson et al. 2000b), a strong correlation between irritability and free testosterone levels has been found in a group of women diagnosed with PMDD (Steiner et al. 1997). Several studies of spironolactone, a testosterone (and aldosterone) antagonist, have reported a significant reduction in premenstrual psychological and physical symptoms compared with placebo (O’Brien et al. 1979; Burnet et al. 1991; Hellberg et al. 1991; Wang et al. 1995). The beneficial effect of spironolactone may at least partly be due to the antiandrogenic properties of the drug (Burnet et al. 1991; Hellberg et al. 1991). One group noted, however, that a significant difference in androgen levels from follicular to the luteal phases may be an important determinant of response to spironolactone therapy in the treatment of PMS (Burnet et al. 1991). Women who received an oral contraceptive with a spironolactonelike progestin and antiandrogenic properties for the treatment of PMDD showed, after three cycles of use, only a slight improvement in psychological symptoms compared with a placebo group (Freeman et al. 2001b).

Ovulation Suppressants Research has shown that gonadotropin-releasing hormone (GnRH) agonists can reversibly suppress the menstrual cycle, and this is often called medical ovariectomy or medical menopause. GnRH agonists have proven to be very successful in relieving mostly physical symptoms in clinical trials. Intranasal buserelin (Hammarbäck and Bäckström 1988; Hussain et al. 1992) or intramuscular leuprolide (Mezrow et al. 1994; Brown et al. 1994; Freeman et al. 1997; Di Carlo et al. 2001) are the most appropriate GnRH treatments for clinical use. Unfortunately, the long-term use of GnRH agonists is limited by the occurrence of side effects that mimic menopause and the potential for hypoestrogenism and osteoporosis. Preliminary evidence suggests that “add-back” therapy with estrogen and progesterone replacement therapy may prevent some of these side effects (Leather et al. 1993; Mezrow et al. 1994). A recent pilot study of five

182

Estrogens and Other Hormones in the Treatment of Premenstrual Syndromes

patients with severe PMS who received depot leuprolide with addback therapy showed stable bone mineral density more than 36 months later (Mitwally et al. 2002). Notwithstanding, two groups have found a reduced clinical response of GnRH analogue therapy with the addition of estrogen and progesterone (Leather et al. 1999, Schmidt et al. 1998). Tibolone (a synthetic add-back compound with estrogenic, androgenic and progestogenic properties) in combination with a GnRH agonist has been proven to provide long-term medical treatment for women suffering from premenstrual physical and psychological symptoms (Di Carlo et al. 2001). Treatment with tibolone alone has also proved to be effective in reducing mood and somatic PMS symptoms (Taskin et al. 1998). Danazol, a gonadotropin inhibitor (known as Danocrine and Cyclomen), has also been effective in clinical trials (Sarno et al. 1987; Hahn et al. 1995; O’Brien and Abukhalil 1999) and in the treatment of premenstrual mastalgia (O’Brien and Abukhalil 1999). Some common side effects with the use of danazol, however, include decrease in breast size, irregular menstrual periods, weight gain, redness of skin, and nervousness.

Summary and Recommendations To date, no single intervention has proven to be equally effective in treating all women with PMS or PMDD. The optimal treatment plan begins with lifestyle modifications, followed by pharmacotherapy. Most, but not all studies, suggest the symptoms of PMS are effectively reduced by inhibition of ovulation. However, for the long-term management of PMS, all of these strategies are somewhat problematic concerning tolerability and adverse health effects. First line therapy is still low-dose selective serotonin reuptake inhibitors, which have demonstrated efficacy with minimal side effects (Dimmock et al. 2000; Born and Steiner 2001). The average age of onset of severe PMS or PMDD is around 26 years. Further evidence suggests that symptoms gradually worsen over time and most likely recur when treatment is halted (Endicott et al. 1999). Thus, therapeutic goals must be set to ensure maximal safety and efficacy for the patient. Modification of the menstrual cycle should be considered only after all other treatment options have failed (Eriksson et al. 2002).

183

Leslie Born and Meir Steiner

The adverse effects associated with ovulation suppression (in particular, the GnRH agonists, estrogen, and danazol) are serious enough to question use of these methods as first-line treatments for severe PMS or PMDD. Moreover, the utility of a treatment that is advised for up to 6 months only, for example, GnRH agonists, is questionable as PMS and PMDD symptoms will persist until the menopause. Some authors have suggested use of a GnRH agonist as a temporary strategy to assist with decision-making for the most appropriate therapeutic option (e.g., Lyall et al. 1999). Symptom charting will help to track efficacy, symptom response to dosing changes, symptoms on termination of therapy, and real versus perceived side effects. Investigators have yet to reach a consensus on how to define efficacy. Clinically, the easiest way to define efficacy is by the reduction of luteal symptoms, i.e., the luteal symptoms remit significantly, or the follicular to luteal difference is less than 30% (Born and Steiner 2001). For women reporting premenstrual somatic complaints but not psychological symptoms, hormonal intervention may be of limited value. However, further studies are warranted to clarify the explicit role of various forms of hormonal treatment for the management of PMS and PMDD.

References American Psychiatric Association (1994) Premenstrual dysphoric disorder. In: Diagnostic and statistical manual of mental disorders, 4th ed. American Psychiatric Association, Washington DC, 715-718 Bäckström T, Hansson-Malmström Y, Lindhe B-Å, Cavalli-Björkman B, Nordenström S (1992) Oral contraceptives in premenstrual syndrome: a randomized comparison of triphasic and monophasic preparations. Contraception 46: 253-268 Bancroft J, Rennie D (1993) The impact of oral contraceptives on the experience of perimenstrual mood, clumsiness, food craving and other symptoms. J Psychosom Res 37: 195-202 Bancroft J, Sanders D, Warner P, Loudon N (1987) The effects of oral contraceptives on mood and sexuality: a comparison of triphasic and combined preparations. J Psychosom Obstet Gynaecol 7: 1-8 Bicikova M, Dibbelt L, Hill M, Hampl R, Starka L (1998) Allopregnanolone in women with premenstrual syndrome. Horm Metab Res 30: 227-30 Born L, Steiner M (2001) Current management of premenstrual syndrome and premenstrual dysphoric disorder. Curr Psychiatr Reports 3: 463-469 Brown CS, Ling FW, Andersen RN, Farmer RG, Arheart KL (1994) Efficacy of depot

184

Estrogens and Other Hormones in the Treatment of Premenstrual Syndromes leuprolide in premenstrual syndrome: effect of symptom severity and type in a controlled trial. Obstet Gynecol 84: 779-786 Burnet RB, Radden HS, Easterbrook EG, McKinnon RA (1991) Premenstrual syndrome and spironolactone. Aust N Z J Obstet Gynaecol 31: 366-368 Casper RF, Hearn MT (1990) The effect of hysterectomy and bilateral oophorectomy in women with severe premenstrual syndrome. Am J Obstet Gynecol 162: 105-109 Casson P, Hahn PM, Van Vugt DA, Reid RL (1990) Lasting response to ovariectomy in severe intractable premenstrual syndrome. Am J Obstet Gynecol 162: 99-105 Chan AF, Mortola JF, Wood SH, Yen SS (1994) Persistence of premenstrual syndrome during low-dose administration of the progesterone antagonist RU 486. Obstet Gynecol 84: 1001-1005 Crammer JL (1986) Premenstrual depression, cortisol and oestradiol treatment. Psychol Med 16: 451-455 Dalton K (1990) The aetiology of premenstrual syndrome is with the progesterone receptors. Med Hypotheses 31: 323-327 Dennerstein L, Morse C, Gotts G, Brown J, Smith M, Oats J, Burrows G (1986) Treatment of premenstrual syndrome a double-blind trial of dydrogesterone. J Affect Disord 11: 199-205 Dennerstein L, Spencer-Gardner C, Gotts G, Brown JB, Smith MA, Burrows GD (1985) Progesterone and the premenstrual syndrome: a double blind crossover trial. Br Med J 290: 1617-1621 Dhar V, Murphy BE (1990) Double-blind randomized crossover trial of luteal phase estrogens (Premarin) in the premenstrual syndrome (PMS). Psychoneuroendocrinology 15: 489-493 Dhar V, Murphy BE (1991) The premenstrual syndrome and its treatment. J Steroid Biochem Mol Biol 9: 275-281 Di Carlo C, Palomba S, Tommaselli GA, Guida M, Di Spiezio Sardo A, Nappi C (2001) Use of leuprolide acetate plus tibolone in the treatment of severe premenstrual syndrome. Fertil Steril 75: 380-384 Dimmock PW, Wyatt KM, Jones PW, O’Brien PMS (2000) Efficacy of selective serotonin-reuptake inhibitors in premenstrual syndrome: a systematic review. Lancet 356: 1131-1136 Endicott J, Amsterdam J, Eriksson E, Frank E, Freeman E, Hirschfeld R, Ling F, Parry B, Pearlstein T, Rosenbaum J, Rubinow D, Schmidt P, Severino S, Steiner M, Stewart DE, Thys-Jacobs S (1999) Is premenstrual dysphoric disorder a distinct clinical entity? J Womens Health Gender-based Med 8: 663-679 Epperson CN, Haga K, Mason GF, Sellers E, Gueorguieva R, Zhang W, Weiss E, Rothman DL, Krystal JH (2002) Cortical ϒ-aminobutyric acid levels across the menstrual cycle in healthy women and those with premenstrual dysphoric disorder. Arch Gen Psychiatry 59: 851-858 Epperson CN, Wisner KL, Yamamoto B (1999) Gonadal steroids in the treatment of mood disorders. Psychosom Med 61: 676-697 Eriksson E, Endicott J, Andersch B, Angst J, Demyttenaere K, Facchinetti F, Lanczik M, Montgomery S, Muscettola G, O’Brien PMS, Studd J, Sundblad C, Young AH (2002) New perspectives on the treatment of premenstrual syndrome and premenstrual dysphoric disorder. Arch Women Ment Health 4: 111-119

185

Leslie Born and Meir Steiner Eriksson E, Sundblad C, Landén M, Steiner M (2000b) Behavioural effects of androgens in women. In: Steiner M, Yonkers KA, Eriksson E (eds) Mood disorders in women. Martin Dunitz, London, 233-245 Eriksson E, Sundblad C, Yonkers KA, Steiner M (2000a) Premenstrual dysphoria and related conditions: symptoms, pathophysiology and treatment. In: Steiner M, Yonkers KA, Eriksson E (eds) Mood disorders in women. Martin Dunitz, London, 269-293 Frank R (1931) The hormonal causes of premenstrual tension. Arch Neurol Psychiatry 26: 1053-1057 Freeman EW, Frye CA, Rickels K, Martin PA, Smith SS (2002) Allopregnanolone levels and symptom improvement in severe premenstrual syndrome. J Clin Psychopharmacol 22: 516-520 Freeman EW, Kroll R, Rapkin A, Pearlstein T, Brown C, Parsey K, Zhang P, Patel H, Foegh M, for the PMS/PMDD Research group (2001b) Evaluation of a unique oral contraceptive in the treatment of premenstrual dysphoric disorder. J Womens Health Gender-based Med 10: 561-569 Freeman EW, Purdy RH, Coutifaris C, Rickels K, Paul SM (1993) Anxiolytic metabolites of progesterone: correlation with mood and performance measures following oral progesterone administration to healthy female volunteers. Neuroendocrinology 58: 478-484 Freeman EW, Rickels K, Sondheimer SJ, Polansky M (2001a) Concurrent use of oral contraceptives with antidepressants for premenstrual syndromes. J Clin Psychopharmacol 21: 540-542 Freeman EW, Rickels K, Sondheimer SJ, Polansky M (1995) A double-blind trial of oral progesterone, alprazolam, and placebo in treatment of severe premenstrual syndrome. J Am Med Assoc 274: 51-57 Girdler SS, Straneva PA, Light KC, Pedersen CA, Morrow AL (2001) Allopregnanolone levels and reactivity to mental stress in premenstrual dysphoric disorder. Biol Psychiatry 49: 788-797 Halbreich U, Kahn LS (2001) Role of estrogen in the aetiology and treatment of mood disorders. CNS Drugs 15: 797-817 Halbreich U, Kahn LS (2000) Selective oestrogen receptor modulators- current and future brain and behaviour applications. Exp Opin Pharmacother 1: 1385-1398 Hammarbäck S, Bäckström T, Holst J, von Schoultz B, Lyrenas S (1985) Cyclical mood changes as in the premenstrual tension syndrome during sequential estrogen-progestagen postmenopausal replacement therapy. Acta Obstet Gynecol Scand 64: 393-397 Hammarbäck S, Ekholm UB, Bäckström T (1991) Spontaneous anovulation causing disappearance of cyclical symptoms in women with the premenstrual syndrome. Acta Endocrin (Cophenh) 125: 132-137 Henshaw C, Foreman D, Belcher J, Cox J, O’Brien S (1996) Can one induce premenstrual symptomatology in women with prior hysterectomy and bilateral oophorectomy? J Psychosom Obstet Gynecol 17: 21-28 Ho HP, Olsson M, Westberg L, Melke J, Eriksson E (2001) The serotonin reuptake inhibitor fluoxetine reduces sex steroid-related aggression in female rats: an animal model of premenstrual irritability? Neuropsychopharmacology 24: 502-510

186

Estrogens and Other Hormones in the Treatment of Premenstrual Syndromes Kendler KS, Karkowski LM, Corey LA, Neale MC (1998) Longitudinal populationbased twin study of retrospectively reported premenstrual symptoms and lifetime major depression. Am J Psychiatry 155: 1234-1240 Lyall H, Campbell-Brown M, Walker JJ (1999) GnRH analogue in everyday gynecology: is it possible to rationalize its use? Acta Obstet Gynecol Scand 78: 340-345 Magos AL, Brincat M, Studd JW (1986) Treatment of the premenstrual syndrome by subcutaneous oestradiol implants and cyclical oral norethisterone: placebo controlled study. Br Med J 292: 1629-1633 Majewska MD (1992) Neurosteroids: endogenous bimodal modulators of the GABA receptor Mechanism of action and physiological significance. Progr Neurobiol 38: 379-395 Marriott A, Faragher EB (1986) An assessment of psychological state associated with the menstrual cycle in users of oral contraception. J Psychosom Res 30: 41-47 Mezrow G, Shoupe D, Spicer D, Lobo R, Leung B, Pike M (1994) Depot leuprolide acetate with estrogen and progestin add-back for long-term treatment of premenstrual syndrome. Fertil Steril 62: 932-937 Mitwally MFM, Gotlieb L, Casper RF (2002) Prevention of bone loss and hypoestrogenic symptoms by estrogen and interrupted progestogen add-back in longterm GnRH-agonist down-regulated patients with endometriosis and premenstrual syndrome. Menopause 9: 236-241 Monteleone P, Luisi S, Tonetti A, Bernardi F, Genazzani AD, Luisi M, Petraglia F, Genazzani AR (2000) Allopregnanolone concentrations and premenstrual syndrome. Eur J Endocrin 142: 269-73. Muse KN, Cetel NS, Futterman LA, Yen SC (1984) The premenstrual syndrome: effects of medical “ovariectomy”. N Engl J Med 311: 1345-1349 O’Brien PM, Craven D, Selby C, Symonds EM (1979) Treatment of premenstrual syndrome by spironolactone. Br J Obstet Gynaecol 86: 142-147 O’Brien PMS, Abukhalil IEH, Henshaw C (1995) Premenstrual syndrome. Curr Obstet Gynaecol 5: 30-35 O’Brien PMS, Abukhalil IEH (1999) Randomized controlled trial of the management of premenstrual syndrome and premenstrual mastalgia using luteal phase-only danazol. Am J Obstet Gynecol 180: 18-23 Taskin O, Gokdeniz R, Yalcinoglu A, Buhur A, Burak F, Atmaca R, Ozekici U (1998) Placebo-controlled cross-over study of effects of tibolone on premenstrual symptoms and peripheral beta-endorphin concentrations in premenstrual syndrome. Hum Reprod 13: 2402-2405 Rapkin AJ, Morgan M, Goldman L, Brann DW, Simone D, Mahesh VB (1997) Progesterone metabolite allopregnanolone in women with premenstrual syndrome. Obstet Gynecol 90: 709-714 Roca CA, Schmidt PJ, Block M, Rubinow DR (1996) Implications of endocrine studies of premenstrual syndrome. Psychiatr Annuals 26: 576-580 Roy-Byrne P, Rubinow DR, Gold PW, Post RM (1984) Possible antidepressant effect of oral contraceptives: case report. J Clin Psychiatry 45: 350-352 Rubinow DR, Hoban MC, Grover GN, Galloway DS, Roy-Byrne P, Anderson R, Merriam GR (1988) Changes in plasma hormones across the menstrual cycle in

187

Leslie Born and Meir Steiner patients with menstrually related mood disorder and in control subjects. Am J Obstet Gynecol 158: 5-11 Rubinow DR, Schmidt PJ, Roca CA (1998) Estrogen-serotonin interactions: implications for affective regulation. Biol Psychiatry 44: 839-850. Sarno AP Jr, Miller EJ Jr, Lundblad EG (1987) Premenstrual syndrome: beneficial effects of periodic, low-dose danazol. Obstet Gynecol 70: 33-36 Schaller JL, Briggs B, Briggs M (2000) Progesterone organogel for premenstrual dysphoric disorder. J Am Acad Child Adolesc Psychiatry 39: 546-547 Schmidt PJ, Nieman LK, Danaceau MA, Adams LF, Rubinow DR (1998) Differential behavioral effects of gonadal steroids in women with and in those without premenstrual syndrome. N Engl J Med 338: 209-216 Schmidt PJ, Nieman LK, Grover GN, Muller KL, Merriam GR, Rubinow DR (1991) Lack of effect of induced menses on symptoms in women with premenstrual syndrome. N Engl J Med 324: 1174-1179 Schmidt PJ, Purdy RH, Moore RH Jr, Paul SM, Rubinow DR (1994) Circulating levels of anxiolytic steroids in the luteal phase in women with premenstrual syndrome and in control subjects. J Clin Endocrin Metab 79: 1256-1260 Scott RT Jr, Ross B, Anderson C, Archer DF (1991) Pharmacokinetics of percutaneous estradiol: a crossover study using a gel and a transdermal system in comparison with oral micronized estradiol. Obstet Gynecol 77: 758-764 Smith RN, Studd JW, Zamblera D, Nigel Holland EF (1995) A randomized comparison over 8 months of 100µg and 200µg twice weekly doses of transdermal oestradiol in the treatment of severe premenstrual syndrome. Br J Obstet Gynaecol 102: 475-484 Steiner M, Coote M, Wilkins A, Dunn E (1997) Biological correlates of irritability in women with premenstrual dysphoria. Eur Neuropsychopharmacol 7 (Suppl 2): S172 Steiner M, Dunn E, Born L (2002) Female-specific mood disorders. In: D’haenen H, den Boer JA, Willner P (eds) Biological psychiatry. Wiley, London, 849-859 Steiner M, Haskett RF, Carroll BJ (1980) Premenstrual tension syndrome: The development of research diagnostic criteria and new rating scales. Acta Psychiatr Scand 62: 177-190 Sundstrom Poromaa I, Smith S, Gulinello M (2003) GABA receptors, progesterone and premenstrual dysphoric disorder. Arch Women Ment Health 6: 23-41 Sveindóttir H, Bäckström T (2000) Prevalence of menstrual cycle symptom cyclicity and premenstrual dysphoric disorder in a random sample of women using and not using oral contraceptives. Acta Obstet Gynecol Scand 79: 405-413 Walker A, Bancroft J (1990) Relationship between premenstrual symptoms and oral contraceptive use: a controlled study. Psychosom Med 52: 86-96 Wang M, Hammarbäck S, Lindhe B-Å, Bäckström T (1995) Treatment of premenstrual syndrome by spironolactone: A double-blind placebo-controlled study. Acta Obstet Gynecol Scand 74: 803-808 Wang M, Seippel L, Purdy RH, Backstrom T (1996) Relationship between symptom severity and steroid variation in women with premenstrual syndrome: study on serum pregnenolone, pregnenolone sulfate, 5 alpha-pregnane-3, 20-dione and 3 alpha-hydroxy-5 alpha-pregnan-20-one. J Clin Endocrin Metab 81: 1076-1082

188

Estrogens and Other Hormones in the Treatment of Premenstrual Syndromes Watson NR, Studd JW, Savvas M, Garnett T, Baber RJ (1989) Treatment of severe premenstrual syndrome with oestradiol patches and cyclical oral norethisterone. Lancet 2: 730-732 West CP (1990) Inhibition of ovulation with oral progestins- effectiveness in premenstrual syndrome. Eur J Obstet Gynecol Reproduct Biol 34: 119-128 Willimann H, Walde P, Luisi PL, Gazzaniga A, Stroppolo F (1992) Lecithin organogel as matrix for transdermal transport of drugs. J Pharm Sci 81: 871-874 World Health Organization. International statistical classification of diseases and related problems, vol 10 (ICD-10) (1992) World Health Organization Press, Geneva Wyatt K, Dimmock P, Jones P, Obhrai M, O’Brien S (2001) Efficacy of progesterone and progestogens in management of premenstrual syndrome: systematic review. Br Med J 323: 776-780 Ylostalo P, Kauppila A, Puolakka J, Ronnberg L, Janne O (1982) Bromocriptine and norethisterone in the treatment of premenstrual syndrome. Obstet Gynecol 59: 292-298 Yonkers KA, Bradshaw KD, Halbreich U (2000) Oestrogens, progestins and mood. In: Steiner M, Yonkers KA, Eriksson E (eds) Mood disorders in women. Martin Dunitz, London, 207-232

189

9 Estrogens and Perinatal Disorders Alain Gregoire

Introduction Pregnancy and childbirth are accompanied by important fluctuations in levels of circulating hormones including prolactin, cortisol, betaendorphin, and human chorionic gonadotrophin. The most dramatic changes, however, probably occur in the levels of estrogens and progesterone that are produced by the placenta and rise steadily during pregnancy (Figs. 1, 2). Estriol levels increase by 100-fold and

Fig. 1. Increase in maternal plasma estrogens during pregnancy

191

Alain Gregoire

Fig. 2. Increase in maternal plasma progesterone during pregnancy

those of estradiol by 1,000-fold. Loss of the placenta at delivery leads to a sudden and rapid drop in these levels, the scale of which exceeds any other physiological event, prepregnancy levels being reached by day 5 postpartum. Estrogens freely cross the blood-brain barrier, and cerebrospinal fluid (CSF) levels appear to correlate well with puerperal circulating levels (Backström et al. 1976). Because the perinatal period is also associated with a variety of changes in mental state, it provides a unique opportunity for studying the neuroendocrinology of such changes, both normal and pathological, and for exploring potential clinical implications.

Changes in Mental State During Pregnancy and Postpartum The postpartum period is associated with a well-documented increased incidence of mental disorders. The most dramatic of these is puerperal psychosis, which represents a 21-fold increase in risk of

192

Estrogens and Perinatal Disorders

psychotic illness, the highest at any time in a woman’s life (Kendell et al. 1987), resulting in up to one admission to a psychiatric hospital per 500 births. In most cases, puerperal psychosis is affective in nature: there is a pronounced disturbance of mood, which can be consistently low or high, or can fluctuate unpredictably between the two. Delusions and/or hallucinations may not always be congruent with mood or may be bizarre. Apparently organic features are also more commonly seen, such as visual or olfactory hallucinations. Disorientation and cognitive disorganization are more common in puerperal than nonpuerperal women (Dean and Kendell 1981; Wisner et al. 1994). The prominence of cognitive disturbance is intriguing, as various aspects of cognitive function are increasingly associated with estradiol (see chapter 11). Abnormal beliefs and experiences generally involve issues relating to the child or the maternal role. Functioning is invariably impaired, usually severely, and although the ability to care for the basic physical needs of the child may be preserved, more complex parenting functions are frequently compromised. Puerperal psychosis usually begins in the first few days after delivery; the incidence decreases sharply over the subsequent 1 or 2 weeks. The genetic characteristics appear to be those of bipolar disorder, probably with an additional familial predisposition to puerperal onset (Craddock and Jones 1999). There is recurrence in 50% of women in the first year and it is sometimes associated with the menstrual cycle (Brockington et al. 1988), which has been suggested as resulting from an association between falling estrogen levels and increased dopaminergic activity (Lovestone 1992). Clinical services caring for this group of women are placing increasing emphasis on prediction and prevention in addition to assertive treatment programs. In the first few days after childbirth, as many as 60%–70% of women experience a disturbance of mood and cognition known as “the blues”. This is characterized by tearfulness, unpredictable mood swings, anxiety, difficulty concentrating, a general feeling of confusion, and lack of self-confidence (Kennerley and Gath 1989). The peak incidence is on the fourth or fifth day, and generally these feelings last for 2 or 3 days. Rates are increased in women with a history of premenstrual syndrome, low mood in pregnancy, and poor family or marital relationships, but social class and life events do not seem to be related to the blues. Although clinicians and mothers often attribute the symptoms to a difficult or exhausting delivery,

193

Alain Gregoire

being in hospital, or suffering perineal pain, there is no evidence that these factors are more common in women who experience the blues than in those who do not. Although transient, they cause distress to the mother, which can be compounded by a lack of understanding and support from those around her. However, recognition of the problem, both by professionals and by women and their families, is growing. Depressive illness is the most common major complication (physical or psychiatric) of the postnatal period. The prevalence of depressive illness in the postpartum period is 10%–15% and the incidence appears to be increased in the first 3 months (Cox et al. 1993; Cooper et al. 1988). Symptoms do not appear to differ significantly from nonpuerperal depressive disorders (O’Hara et al. 1990), although this finding is not consistent (Dean and Kendell 1981), which may reflect heterogeneity in the disorder. Social class, parity, and adverse obstetric factors appear to have little influence on the incidence of postnatal depression; contradictory results emerge from the association with stressful life events, but social and marital support appear to be important factors. However, the ability of all such factors to predict the occurrence of postnatal depression is limited (Cooper et al. 1996). Again, this may be due to the heterogeneous etiology of depression, and at present there is evidence of a subgroup of women with specific vulnerability to postpartum depressive illness (Cooper and Murray 1995). Although severe mental illnesses such as schizophrenia undoubtedly have a significant impact on parenting function (Gregoire 2000), the effect of having a child during schizophrenia appears to be variable. One of the factors identified as predicting this differential effect is the nature of the schizophrenic illness, with less severe or typical illnesses being more associated with acute episodes of postnatal illness and more severe schizophrenic illnesses being unaffected by childbirth (Davies et al. 1995).

The Relationship between Estrogens and Mental State The variations in the epidemiology of mood disorders and schizophrenia between the genders and over the life-cycle are strongly suggestive of a relationship with gonadal and reproductive function. Numerous studies have demonstrated a higher lifetime prevalence

194

Estrogens and Perinatal Disorders

of depression in women compared to men, with a higher annual prevalence beginning after the age of 10 years (Weissman et al. 1988; Kessler et al. 1993) but gradually disappearing after the age of 45 years. These differences apply to community samples as well as clinical populations and are therefore unlikely to be due to gender differences in presentation. The timing of the differences coincides almost precisely with the reproductive period in women’s lives and the fluctuations in gonadal steroids that characterize it. The distribution of the onset of schizophrenia between genders and across age groups is also suggestive of an association with changes in reproductive hormones (Fig. 3; Häfner et al. 1993). Stevens (2002) postulated that this could be the result of the influence of estrogens and testosterone on anterior forebrain structures which might be associated with the development of schizophrenic symptoms. Such influences on mental state are entirely compatible with what we know about the estrogenic effects on brain structures and neurotransmitter systems involved in mood and cognition. Estrogen receptors are widely distributed in the brain, including the hippocampus, amygdala, raphe nuclei, cingulate cortex, and nucleus accumbens (McEwen et al. 1997; Sumner and Fink 1995). They exert a modulating influence on a number of key neurotransmitter systems, notably serotonergic, dopaminergic, noradrenergic, gammaamino butyric, cholinergic, and glutamate systems (McEwen et al. 1997). Estrogen binds to both membrane and intracellular receptors that exert a wide range of influences including modulation of other receptors and transcription of genes. Details of these effects are beginning to emerge from animal studies. Interspecies differences in neuroendocrinology demand caution in generalizing any such findings to humans, but they remain intriguing nevertheless. Estrogen increases the density of 5-hydroxytryptamine (5-HT) (Biegon and McEwen 1982) and 5-HT2A (Sumner and Fink 1995; Fink and Sumner 1996) receptors in a tissue-specific manner. Furthermore, 5-HT2 receptor changes that occur in response to antidepressants appear to be modulated by estrogen (Kendell et al. 1982). Thus, the serotonergic effects documented to date are consistent with an overall 5-HT agonist effect of estradiol which parallels the central effects of many antidepressants. Despite these in vitro and animal findings, studies that have

195

Alain Gregoire

(a)

(b)

Fig. 3. Sex-specific age distributions at first onset of schizophrenia of a broad definition (ICD-9: 295, 297, 298.3 and 298.4.) (a) At earliest sign of mental disorder (–––, males, N = 117; ..., females, N = 131. (b) At first psychotic symptom (–––, males, N = 125; ..., females, N = 139) From Häfner et al. (1993)

196

Estrogens and Perinatal Disorders

attempted to identify abnormalities in sex steroid levels in relation to postpartum mood have all yielded negative results. For example, Nott et al. (1976), and Harris et al. (1989) found no difference in estradiol levels between women who developed postpartum depression and those who did not, a finding replicated by O’Hara et al. (1989), who also measured free estriol. As neither antenatal levels nor changes in levels predict depression, the remaining possibility is that the normal physiological changes lead to changes in mental state only in women rendered vulnerable by some other factor. Bloch et al. (2000) provide evidence for such a process in a small but elegant study. Eight women with a history of postnatal depression and eight women with no history of depression were given a gonadotrophin-releasing hormone (Gn-RH) agonist (leuprolide acetate) to induce hypogonadism, and supraphysiological doses of estradiol and progesterone were then administered for 8 weeks to simulate late pregnancy levels. Following double-blind withdrawal of the estradiol and progesterone, five of the eight women with a history of postpartum depression developed significant mood symptoms as opposed to none of the women without such history. Other, more naturalistic examples of observed changes in mental state associated with withdrawal of estrogens have been described. For example, a psychotic episode X days after the removal of a hydatidiform mole (Hopker and Brockington 1991) and “puerperal psychosis” following estrogen withdrawal in a transsexual man who had been taking estradiol before gender reassignment surgery (Mallett et al. 1989). Further such cases are reviewed by Mahé and Dumaine (2001). These cases among men suggest that sensitivity to estrogen withdrawal is not exclusive to women, although obviously more rarely seen.

Therapeutic Effects of Estrogen Treatment The first published randomized controlled trial of estrogen treatment for a psychiatric disorder was that of Klaiber et al. (1979), who randomly allocated a sample of 40 women with treatment-resistant depression to augmentation with high-dose estrogens (Premarin 15–25 mg) or placebo. Treatment-resistant depression was defined as “at least 2 years of unsuccessful treatment by a variety of conventional therapies.” They found a significant positive effect of estradiol

197

Alain Gregoire

over placebo but, surprisingly, the study was never replicated. Another randomized controlled trial examined the effects of the tricyclic antidepressant Imipramine combined with either Premarin or placebo in 11 women with resistant depression of at least 6 months’ duration (Shapira et al. 1985; Zohar et al. 1985). Although no overall statistically significant difference in symptomatic improvement emerged between the two groups, they noted that one subject in the estrogen group got dramatically better within 2 weeks and one became manic. This could be purely artifactual, but would also be compatible with the hypothesis that some women have a particular sensitivity to the antidepressant effects of estrogen. Oppenheim (1984) also described a case of rapid cycling which appears to have been induced by the addition of conjugated estrogen to antidepressant treatment. A number of studies have also provided evidence that estrogens are effective in reducing symptoms in premenstrual syndrome, the perimenopause (Soares et al. 2001), and may augment the effect of antidepressants in postmenopausal women (Schneider et al. 1997; 2001). Reports are also emerging of the potentially therapeutic effects of estrogens in schizophrenia. Hoff et al. (2001) found a strong positive correlation between estradiol levels and cognitive function, though not other symptoms, in women with schizophrenia. In a double-blind randomized controlled trial, Kulkarni et al. (2001; see chap. 5) found that women with schizophrenia receiving 50 µg/day or 100 µg/day of estradiol in addition to standard antipsychotic medication showed a significantly better response than women receiving placebo estradiol plus antipsychotics. Furthermore, Thompson et al. (2000) demonstrated that higher levels of estrogen are associated with lower extrapyramidal side effects. The apparent contradiction of enhanced antipsychotic effects and reduced extrapyramidal symptoms is consistent with the hypothesis advanced by Van Hartesvelt and Joyce (1986) of the differential effects of estrogen in mesolimbic and mesostriatal pathways. These findings obviously suggest an exciting therapeutic possibility which merits further exploration. Similarly encouraging results emerge from studies of the impact of estrogens on cognitive function in Alzheimer’s disease. For example, Asthana et al. (2001) conducted a randomized controlled trial of high-dose estradiol, which produced significant benefits in cognitive function in postmenopausal women with Alzheimer’s disease.

198

Estrogens and Perinatal Disorders

Estrogen Treatment for Postpartum Disorders The only double-blind randomized controlled trial of estrogen treatment for postpartum psychiatric disorder was conducted by Gregoire et al. (1996), who studied the effects of six months’ treatment with estradiol skin patches (200 mg/day) on 63 women with moderate to severe depressive illness with postpartum onset [means for treatment and control groups, respectively: Edinburgh Postnatal Depression Scores (EPDS) 21.8 and 21.3; duration 31 and 36 weeks]. The estradiol was given without progesterone for the first 3 months after which cyclical dydrogesterone (10 mg for 12 days) was given. Twenty-six of the women were also taking antidepressants at baseline but only women who had shown no improvement over the 6 weeks prior to recruitment and who had had no change in the antidepressant type or dosage during that time were included. Only women who had EPDS of 14 or more at the time of recruitment (scores indicative of moderate to severe depression) and 4 weeks later at baseline were included in the study. Considerable efforts were made to maintain blindness in scoring depression: stage of treatment, compliance, side effects, and physical health checks were recorded by a different researcher to the one scoring depressive symptoms, and participants were asked not to discuss these aspects of the trial with the psychiatric interviewer, a request which they observed well. Psychopathology and functioning was rated by interview using the Schedule for Affective Disorders and Schizophrenia (SADS) and Global Assessment of Functioning (GAF), and participants completed the EPDS. Of the 63 women, two did not complete the trial: one woman was found to have hyperthyroidism and was excluded in the first month of the study and another participant committed suicide in the first month of treatment. She had been withdrawn from her trial medication by her local psychiatrist without consultation with the research team and had been started on progesterone. She was admitted to a psychiatric hospital and drowned herself while on leave. Of the remaining sample, 27 women were in the placebo group and 34 in the active treatment group. As would be expected in any placebo-controlled trial of treatment for depression, there was a significant improvement over time in both the active and the placebo group. However, there was a significantly greater improvement in the active group compared to the placebo

199

Alain Gregoire

group at 1 month and this difference was maintained over the subsequent 5-month treatment period and during the 3 months of follow-up (ANOVA means vs. means of recruitment minus 1 month and baseline scores, f = 12.36; p < 0.001). These findings, based on analysis of the EPDS results, also emerged when SADS depression and GAF scores were analyzed. The results of all three measurements were in fact very highly correlated (r = 0.61–0.87; p < 0.001). The treatment effect was also analyzed using a best-fit repeated measures analysis which yielded a treatment effect of p < 0.001. The mean difference in EPDS score between the active and placebo group was 4.38 (95% CI = 1.89–6.87). This suggests a clinically, as well as statistically, significant difference between the groups, which is further demonstrated by the finding that 69% of the women in the placebo group still had scores above 14 by 3 months compared to 20% of the women in the active group. The above study has not yet been replicated using a randomized control design, but a number of small open-label trials and cases have been reported. Ahokas et al. (1998; 1999) reported four cases of women with moderate to severe postnatal depression lasting 2–5 months, all of whom were documented as having low estradiol levels. One woman had a past history of postnatal depression and one was on treatment with Moclobemide but had shown no response to this drug. All four were treated with oral micronized 17β-estradiol 4 mg/day, and all four were described as improving within 1 week and being almost free of symptoms by 2 weeks. The same group (Ahokas et al. 1999; 2000; 2001) has also reported the results of estradiol treatment for puerperal psychosis in a sample of ten women as well as case reports of one and two women, although it appears that these samples may overlap. All of these patients had low estradiol levels (13–90 pmol/l); all were better in 2 weeks. The authors also describe the relapse of florid psychotic symptoms within 1 week in one patient who discontinued the estradiol after 5 weeks.

Estrogen Prophylaxis for Postpartum Disorders The possibility that the rapid drop in estrogen levels at delivery can precipitate puerperal psychoses suggests that attenuation of this drop by administration of estrogens after delivery might have a prophylactic effect. Hamilton and Sichel (1992) describe a series of 50 women with a history of postpartum psychosis treated in this way.

200

Estrogens and Perinatal Disorders

The women were given a high dose of estrogen (10 mg by injection) immediately after delivery followed by oral Premarin for 14 days. None of the women relapsed (as opposed to the expected rate of 30%–70% described above). Sichel et al. (1995) describe 12 cases of estrogen used prophylactically immediately after delivery. Four of the women in their study had a history of postpartum depression and eight had a history of puerperal psychosis. A starting dose of 5 mg Premarin twice daily was administered from day 1 postpartum, gradually decreasing over a period of 28 days. Progress was evaluated daily for the first 5 days, then at 4 weeks and 3, 6, and 12 months. None of the four women with a history of postpartum depression showed any sign of relapse, and all but one of the women with a history of psychosis remained well. The one woman who did relapse was believed to have been noncompliant with the estrogen regime. Kumar et al. (2002) carried out an open trial of three gradually reducing regimes of estradiol over 12 days in 29 women with a history of bipolar or schizoaffective disorder. Three different starting doses, administered within 48 h of delivery, were used: transdermal estradiol 200 mg, 400 mg, and 800 mg for 24 h. The dose was then halved every 4 days over 12 days. The study program began with the 200 µg starting dose, but as relapse rates appeared not to be reduced, this was increased to 400 µg and then 800 µg. No difference was found between the relapse rates in this study sample, or in subsamples, and the expected rates. However, among the women who did relapse, those who had received the higher-dose regime (starting at 800 µg/day) had significantly shorter hospital stays and possibly smaller (though not significantly) doses of antipsychotic medication. No differences in estradiol levels were found between women who did and did not relapse. This is intriguing as it appears to contradict the findings of Ahokas et al., all of whose subjects had low estradiol levels. At least two factors may explain this apparent discrepancy. Firstly, estradiol levels remain elevated soon after childbirth, 2–3 times the normal luteal phase levels. Secondly, it may be that there is a subgroup of women with these disorders who have low levels and who respond to estradiol treatment. It is not clear from the reports by Ahokas et al. whether their samples might have been selected on the basis of low levels and therefore constitute such a subgroup. The apparent lack of a prophylactic effect of estrogen found by

201

Alain Gregoire

Kumar et al. (2002), which is in stark contrast to the findings described by Hamilton and Sichel (1992) and those of Sichel et al. (1995), also demands explanation. The authors do not attempt to address this issue. At present, the paucity of evidence in this area allows multiple hypotheses: (a) type-I or type-II errors in the studies (which are small and uncontrolled); (b) differences in dosages, preparations, and regimes; and (c) possible sampling differences exposing important heterogeneity in the population.

Implications for Clinical Practice Although the current state of knowledge is clearly still in its infancy, some tentative conclusions can be drawn that may be of relevance in the clinical setting. Firstly, the accumulated evidence and the findings of one randomized controlled trial support the efficacy of estradiol in the treatment of postpartum depression. Small, uncontrolled case series also suggest that it may be effective for the treatment of puerperal psychosis and prevention of depression and psychosis. Trials so far have concentrated on moderate to severe and resistant cases of depression, and we therefore do not have evidence on effectiveness in mild to moderate depression. It seems that a response is usually seen within the first 4 weeks of treatment and, in some women, within a few days. It also appears that benefit from treatment is maintained following discontinuation of treatment after 6 months. There is no evidence of a disadvantage in combined treatment with antidepressants; however, it is not clear whether this confers any additional benefit, and so far there have been no comparative trials to tell us whether treatment with either one of these substances alone is superior to the other. Given the very limited evidence at present, clinical use of estradiol should most likely be restricted to moderate to severe and resistant cases of postpartum depression, probably including those with psychotic features. Estradiol treatment can also be appropriate for women who have a particularly strong history of postpartum depressive illness and those who tend to be well during pregnancy, or have a history of premenstrual mood disorder or other history suggestive of responsivity to estrogens. It may also be appropriate to offer this treatment to women who refuse or will not engage with any other form of treatment but are prepared to consider a hormonal approach. Although one research group has provided data suggestive

202

Estrogens and Perinatal Disorders

of an association between response to treatment and low baseline estrogen levels, this may not be consistent with findings from other studies.

Research Implications Research in this area is still in its infancy, but so far it does seem to suggest tantalizing opportunities for a better understanding of one of the elements of the complex processes underlying our mental state and may even offer effective new intervention strategies in a range of mental disorders. A few specific findings on neurotransmitter effects in animal studies need extending and of course, as much as possible, assumptions about generalization to humans need testing. At present, the very best evidence of clinical applications comes from single, randomized, controlled studies with relatively small numbers that are unable to provide anything more than evidence on overall efficacy. We now need further preliminary randomized trials of this type across the range of disorders, but more importantly, larger trials, not only for replication but also to yield more detailed and clinically useful data. The questions which clinicians need answers to, in addition to efficacy in particular disorders, include: • Are estrogens effective in prophylaxis? • Which individual factors predict response? • What are the long-term outcomes for mother and child? • What are the dose–response relationships? • What are the minimum and optimum durations of treatment? • What are the undesirable effects in various patient populations? • Are there short- or long-term effects on infants if used during breastfeeding? Answers to some of these questions, hopefully available in the next few years, could reveal not only new insights into etiological processes but also open up a potentially exciting new area of treatment and prevention.

References Ahokas AJ, Turtiainen S, Aito M (1998) Sublingual oestrogen treatment of postnatal depression. Lancet 351: 109 Ahokas A, Aito M, Rimon R (2000) Positive treatment effect of estradiol in postpartum psychosis: A pilot study. J Clin Psychiatry 61: 166-169

203

Alain Gregoire Ahokas A, Aito M, Turtiainen S (2000) Association between oestradiol and puerperal psychosis. Acta Psychiatr Scand 01: 167-170 Ahokas A, Aito M (1999) Role of estradiol in puerperal psychosis. Psychopharmacol 147: 108-110 Ahokas A, Kaukoranta J, Aito M (1999) Effect of oestradiol on postpartum depression. Psychopharmacol 146: 108-110 Ahokas A, Kaukoranta J, Wahlbeck K, Aito M (2001) Estrogen deficiency in severe postpartum depression: Successful treatment with sublingual physiologic 17βestradiol: A preliminary study. J Clin Psychiatry 62: 332-336 Asthana S, Baker LL, Craft S, Stanczyk FZ, Veith RC, Raskind MA, Playmate SR (2001) High-dose estradiol improves cognition for women with AD – results of a randomised study. Neurology 57: 605-612 Backström T, Carstenson H, Sodergaard R (1976) Concentration of estradiol, testosterone and progesterone in cerebrospinal fluid compared to plasma unbound and total concentrations. J Steroid Biochem 7: 469-472 Biegon A, McEwen BS (1982) Modulation by estradiol of serotonin, receptors in brain. J Neurosci 2: 199-205 Bloch M, Schmidt PJ, Danaceau M, Murphy J, Nieman L, Rubinow DR (2000) Effects of gonadal steroids in women with a history of postpartum depression. Am J Psychiatry 157: 924-930 Brockington IF, Kelly A, Hall P, Deakin W (1988) Premenstrual relapse of puerperal psychosis. J Affect Disord 14: 287-292 Cooper PJ, Campbell EA, Day A, Kennerley H, Bond A (1988) Non-psychotic psychiatric disorder after childbirth. A prospective study of prevalence, incidence, course and nature. Br J Psychiatry 152: 799-806 Cooper PJ, Murray L, Hooper R, West A (1996) The development and validation of a predictive index for postpartum depression. Psychol Med 26: 627-634 Cooper PJ, Murray L (1995) Course and recurrence of postnatal depression – evidence for the specifity of the diagnostic concept. Br J Psychiatry 166: 191-195 Cox JL, Murray D, Chapman G (1993) A controlled study of the onset, duration and prevalence of postnatal depression. Br J Psychiatry 163: 27-31 Craddock B, Jones I (1999) Genetics of bipolar disorder. J Med Genet 36: 585-594 Daves A, McIvor RJ, Kumar C (1995) Impact of childbirth on a series of schizophrenic mothers: a comment on the possible influence of oestrogen on schizophrenia. Schizophr Res 16: 25-31 Dean C, Kendell RE (1981) The symptomatology of puerperal illness. Br J Psychiatry 139: 128-133 Dean C, Kendell RE (1981) The symptomatology of puerperal illnesses. Br J Psychiatry 139: 128-133 Fink G and Sumner B (1996) Oestrogen and mental state. Nature, Sci Correspond 383: 306 Gregoire AJP, Kumar R, Everitt B, Henderson AF, Studd JW (1996) Transdermal oestrogen for treatment of severe postnatal depression. Lancet 347: 930-933 Gregoire A (2000) Mentally ill parents. In: Gregoire A (ed) Adult Severe Mental Illness. GMM, London Häfner H, Riecher-Rössler A, an der Heiden W, Maurer K, Fatkenheuer B, Löffler W

204

Estrogens and Perinatal Disorders (1993) Generating and testing a causal explanation of the gender difference in age at first onset of schizophrenia. Psychol Med 23: 925-940 Hamilton JA, Sichel DA (1992) Prophylactic measures. In: Hamilton JA, Harberger PN (eds) Postpartum Psychiatric Illness. University of Pennsylvania Press, Philadelphia, 219-224 Harris B, Johns S, Fung H, Thomas R, Walker R, Read G, Riad-Fahmy D (1989) The hormonal environment of postnatal depression. Br J Psychiatry 154: 660-667 Hoff AL, Kremen WS, Wieneke MH, Lauriello J, Blankfeld HM, Faustman WO, Csermansky JG, Nordahl TE (2001) Association of estrogen levels with neuropsychological performance in women with schizophrenia. Am J Psychiatry 158: 1134-1139I Hopker SW, Brockington IF (1991) Psychosis following hydatidiform mole in a patient with recurrent puerperal psychosis. Br J Psychiatry 158: 122-123 Kendell RE, Chalmers JC, Platz C (1987) Epidemiology of puerperal psychosis. Br J Psychiatry 150: 662-673 Kennerley H, Gath D (1989) Maternity blues: detection and measurement by questionnaire. Br J Psychiatry 155: 356-362 Kessler RC, McGonagle KA, Swartz M, Blazer DG, Nelson CB (1993) Sex and depression in the National Comorbidity Survey I: Lifetime prevalence, chronicity and recurrence. J Affect Disord 29: 85-96 Klaiber EL, Broverman DM, Vogel W, Kobayashi Y (1979) Estrogen therapy for severe persistent depressions in women. Arch Gen Psychiatry 36: 550-554 Kulkarni J, Riedel A, de Castella AR, Fitzgerald PB, Rolfe TJ, Taffe J, Burger H (2001) Estrogen – a potential treatment for schizophrenia. Schizophr Res 48: 137-144 Kumar C, McIvor RJ, Davies T, Brown N, Papadopoulos A, Wieck A, Checkley SA, Campbell IC, Marks MN (2003) Estrogen administration does not reduce the rate of recurrence of affective psychosis after childbirth. J Clin Psychiatry 64:112-118 Lovestone S (1992) Periodic psychosis associated with the menstrual cycle ad increased blink rate. Br J Psychiatry 161: 402-404 Mahé V, Dumaine A (2001) Oestrogen withdrawal associated psychoses. Acta Psychiatr Scand 104: 323-331 Mallett P, Marshall EJ, Balcker CVR (1989) Puerperal psychosis following male-tofemale sex reassignment? Br J Psychiatry 155: 257-259 McEwen BS, Alves SE, Bulloch K, Weiland NG (1997) Ovarian steroids and the brain: Implications for cognition and ageing. Neurology 48: S8-S15 Nott PM, Franklin M, Armitage C, Gelder MG (1976) Hormonal changes and mood in the puerperium. Br J Psychiatry 128: 379-383 O’Hara MW, Zekoski EM, Philipps LH, Wright EJ (1990) Controlled prospective study of postpartum mood disorders: comparison of childbearing and nonchildbearing women. J Abnorm Psychology 99: 3-15 Oppenheim G (1984) A case of rapid mood cycling with oestrogen: implications for therapy. J Clin Psychiatry 45: 34-35 Schneider LS, Small GW, Clary CM (2001) Estrogen replacement therapy and antidepressant response to sertraline in older depressed women. Am J Geriatr Psychiatry 9: 393-399 Schneider LS, Small GW, Hamilton SH, Bystritsky A, Nemeroff CB, Meyers BS

205

Estrogens and Perinatal Disorders (1997) Estrogen replacement and response to fluoxetine in a multicentre geriatric depression trial. Am J Geriatr Psychiatry 1997: 97-106 Shapira B, Oppenheim G, Zohar J, Segal M, Malach D, Belmaker RH (1985) Lack of efficacy of estrogen supplementation to imipramine in resistant female depressives. Biol Psychiatry 20: 576-579 Sichel DA, Cohen LS, Robertson LM, Ruttenberg A, Rosenbaum JF (1995) Prophylactic oestrogen in recurrent postpartum affective disorder. Biol Psychiatry 38: 814-818 Soares CN, Almeida OP, Joffe H, Cohen LS (2001) Efficacy of estradiol for the treatment of depressive disorders in perimenopausal women: a double-blind, randomized, placebo-controlled trial. Arch Gen Psychiatry 58: 529-534 Stevens JR (2002) Schizophrenia: reproductive hormones and the brain. Am J Psychiatry 159: 713-719 Sumner B, Fink G (1995) Estrogen increases the density of 5-Hydroxytryptamine2A receptors in cerebral cortex and nucleus accumbens in the female rat. J Steroid Biochem Mol Biol 54: 15-20 Thompson KN, Kulkani J, Sergejew AA (2000) Extrapyramidal symptoms and oestrogen. Acta Psychiatr Scand 101: 130-134 Van Hartesvelt C, Joyce J (1986) Effects of oestrogen on the basal ganglia. Neurosci Biobehav Rev 10: 1-14 Weissman MM, Leaf PJ, Tischler GL, Blazer DG, Karno M, Bruce ML, Florio LP (1988) Affective disorders in five United States communities. Psychol Med 18: 141-153 Wisner KL, Peindl K, Hanusa BH (1994) Symptomatology of affective and psychotic illnesses related to childbearing. J Affect Disord 30: 77-87

206

10 Estrogen Therapy in Perimenopausal Affective Disorders Gabriela Stoppe and Martina Dören

Perimenopause and Affective Disorders A role of estrogen in the regulation of mood has been postulated since extracts of animal ovarian tissue were administered to oophorectomized women at the end of the 1990s to alleviate psychological symptoms thought to be related to the removal of the ovaries (Notelovitz 1999). Today, however, the exact role of endogenous estrogen and its deficiency in depressive disorders and mood changes still needs to be defined. In this review, menopause is defined as the last episode of menstrual bleeding generated by endogenous ovarian function. This is a definition which only applies to women with an intact uterus and can only be established retrospectively. Postmenopause is the subsequent period in a woman’s remaining life span. The term “perimenopause” is less well defined; this interval includes the time of change of ovarian function associated with estrogen deficiency symptoms such as bleeding irregularities and vasomotor symptoms. By definition, the term perimenopause includes both the immediate end of the reproductive potential at the time before menopause and the very beginning of the time after menopause. This is a somewhat arbitrary definition; however, it serves the purpose of highlighting a period which, for many women, is marked by a health-related impairment of quality of life due to symptoms that are thought to be predominantly related to estrogen deficiency, such as vasomotor complaints. This chapter focuses on depression, because this disorder, and not mania or anxiety, has been discussed in the context of hormonal changes and therapy. Gynecologists and psychiatrists usually see other types of patients and are confronted with presumably other types and patterns of complaints. Therefore, it seems important to

207

Gabriela Stoppe and Martina Dören

distinguish between mood changes and complaints (e.g., depressive symptoms) and depressive disorders using operationalized criteria, for example, those in DSM-IV or ICD-10 (American Psychiatric Association 1994; World Health Organization 1991). According to these criteria, „major depression” is defined as a reduction of mood, performance, and drive along with hopelessness usually accompanied by feelings of reduced self-esteem, guilt, suicidal tendency and changes in appetite, weight, libido, psychomotor activity, and sleep, generally lasting at least 2 weeks. To define an episode as “depression”, these symptoms must represent a significant change compared to the patient’s previous state. There is still ongoing debate on whether only major or also minor and other unspecified forms of depression are of clinical relevance, because the physical and socioeconomic impacts of the different forms of depression do not vary substantially (Simon et al. 1995). Numerous epidemiological studies revealed an approximately twofold prevalence rate of depression and dysthymia in women – with less difference in childhood and advanced old age and a female:male ratio of 3–4:1 during reproductive years (Weissman et al. 1988; Wells et al. 1989; Parry 1989; Leon et al. 1993). This difference has been attributed to biological reasons (Wilhelm and Parker 1989; Gater et al. 1989; Harris et al. 1991; Hamilton and Halbreich 1993) as well as more extraindividual parameters such as the social role of women (Aro 1994; Silverstein and Perlick 1991; Jorm 1987; Lalive d’Epinay 1985). However, there were also discussions regarding the relevance of comorbidity (Breslau et al. 1995) and epidemiological methods (e.g., the symptom threshold, definition of case, selection of instruments), factors which may contribute toward higher rates of depression in women than in men (Wilhelm and Parker 1989; Young et al. 1990; Stommel et al. 1993; Angst and Dobler-Mikola 1984). Finally, there are reports about gender-type patient behavior, with women reporting more and remembering better any depressive symptoms (Angst and DoblerMikola 1984; Wilhelm and Parker 1994). It has been shown that female patients complain more often about depressed mood, seek their physicians’ help more frequently, and are more likely to receive adequate treatment than men (Williams et al. 1995). According to our own investigations, identical complaining would lead to a significantly higher recognition rate in women, which might be the result of probabilistic behavior and/or of role stereotypes on the part of the physicians (Stoppe et al. 1999).

208

Estrogen Therapy in Perimenopausal Affective Disorders

Community-based studies providing data on self-reported depressive symptoms suggest that most peri- and postmenopausal women do not experience symptoms of major depression (Kaufert et al. 1992; Avis et al. 1994; Matthews et al. 1994; Matthews 1992), but rather symptoms of mild depression (Avis et al. 1994; Hunter 1990; Hay et al. 1994). However, many studies conducted to date simply asked about mental and emotional states without sufficiently investing in the selection or development of appropriate questionnaires. Furthermore, by announcing to undertake “menopause research,” many studies introduced a bias toward attribution of symptoms to this female life event. There appear to be substantial differences regarding a variety of symptoms reported by perimenopausal women from various cultures including symptoms of depression. European and North American women seem to report similar frequencies of menopausal symptoms, the prevalence of which appears to be lower in Asian countries (Makhlouf Obermeyer 2000). Social factors commonly related to mental health also vary between these areas: In Europe and the USA, menopausal symptoms are more frequent in lower socioeconomic groups, whereas the opposite is true in Asia (Dennerstein 1996; Neri et al. 1997; Polit and LaRocco 1982; Kuh et al. 1997). Regarding the range of symptoms, irritability, suggested to constitute a female-specific mood disorder by some authors (Born and Steiner 1999), seems to be more common in women with hormonal changes including the perimenopause and was, for example, reported as often as hot flushes in a representative cross-sectional nationwide survey of 1,038 German women aged 50–70 years (Schultz-Zehden 1998). Some authors found that the extent of previous premenstrual symptoms – such as irritability, anxiety and panic, fatigue, sleep disturbances, depression and cognitive dysfunction – is related to the occurrence of similar symptoms during perimenopause (including postpartum blues and depression as well as oral contraceptive dysphoria), implying a common physiological denominator (Abraham et al. 1994; Stewart and Boydell 1993; Arpels 1996; Holte and Mikkelsen 1982; Hunter 1990; Avis and McKinlay 1991; Greene and Visser 1992; Dennerstein 1996). The menopause itself has not been identified as a major cause of depressive symptoms (Nicol-Smith 1996); signs of mild depression/ mood changes were observed in perimenopausal women in some prospective population-based studies (Bungay et al. 1980; Ballinger

209

Gabriela Stoppe and Martina Dören

1975; Hunter 1992) including women with surgical menopause (McKinlay et al. 1987), but did not increase in other longitudinal investigations that included women with surgical menopause, too (Neugarten and Kraines 1965; Hallström et al. 1985; Busch et al. 1994). According to recent, large, population-based longitudinal studies, more than 50% of the variance for the development of depression in the menopausal transition is explained by previous depression as well as cognitive and social factors such as poor social support and/or poor marital relationship (Hunter 1990; Avis et al. 1994; Greene and Visser 1992; Kaufert et al. 1992; Pearlstein et al. 1997). Cross-sectional studies failed to show a higher prevalence of depression in the perimenopausal period. If serum hormone levels were obtained concomitantly, there was no direct correlation to the severity of depression (Coope 1981; Ballinger et al. 1987; Hunter 1990; Wilbur et al. 1995; Holte 1992). Only four studies, largely with considerable methodological problems, reported associations between well-being (not further specified) or depression and serum levels of dehydroepiandrosterone (DHEA), dehydroepiandrosterone sulfate (DHEA-S) (Morales et al. 1994; Cawood and Bancroft 1996; Barrett-Connor et al. 1999), or follicle-stimulating hormone FSH (Huerta et al. 1995). However, among women attending (gynecology) outpatient clinics, the prevalence of depression including major depression may be considerably higher (Anderson et al. 1987; Montgomery et al. 1987; Stewart et al. 1992; Dennerstein et al. 1993; Hay et al. 1994). Women with previous use of gynecological services and resources, e.g., hysterectomy and/or ovariectomy and use of contraceptives, are much more prevalent in a (gynecological outpatient) clinic. It is the same group of women, however, who complain significantly more often about menopausal symptoms including depression and also show a much higher rate of hormone (replacement) therapy (Topo et al. 1995; Schultz-Zehden 1998; Barrett-Connor et al. 1999; Li et al. 2000). It is noteworthy that perimenopausal women utilizing health care services differ from women who do not. The former may embrace medical treatments including hormone therapy (HT) in the hope of solving an existing, not necessarily medical, problem (Morse et al. 1994). This could be one reason that clinicians may develop opinions regarding the symptomatology of the menopause which might not be applicable to more representative population samples including women not seeking medical attention.

210

Estrogen Therapy in Perimenopausal Affective Disorders

The end of the reproductive life span does not cause problems in a majority of women, which contradicts an often-expressed opinion that menopausal women suffer due to the loss of their social role. About 75% of women in the postmenopause feel well (again), sometimes even happier and healthier than in previous life spans with a trend to an increasingly “positive” and accepting attitude toward the menopause (von Sydow and Reimer 1995; Wilbur et al. 1995; Dennerstein et al. 1994; Kaufert et al. 1998; Schultz-Zehden 1998). This opinion is consistent with a Dutch investigation that did not demonstrate any gender difference for a variety of psychological and emotional complaints in women and men in their midlife (van Hall et al. 1994). A German study on a representative sample of men older than 40 years revealed mild to moderate somato-vegetative symptoms in about 50% (Heinemann et al. 1999). Whether depression can trigger an early onset of menopause or whether the specific pharmacological treatment is responsible for the premature termination of cyclic ovarian function was only recently identified as an area for future research (Harlow and Signorello 2000). An earlier onset of menopause was found in women at risk for dementia in later life (van Duijn 1997).

Impact of Hormone (Replacement) Therapy on (Signs and Symptoms of) Depression When using the term “hormone replacement therapy”, one should consider that “replacement” implies the correction of “ovarian failure” and a deficiency state. Since all women experience sinking hormonal levels during the menopausal transition, this could lead to the assumption that all women are somewhat “defective” (i.e., ill) after the end of the reproductive cycle. The term “hormone treatment” could be more neutral and therefore better for this discussion (Stoppe et al. 2000). Today, it is still difficult to describe precisely the specific effects of various compounds in common clinical use. The findings of many early studies are difficult to interpret because of their methodological shortcomings regarding design and patient selection. Several, but not all, prospective, randomized, double-blind, placebo-controlled,

211

Gabriela Stoppe and Martina Dören

short-term clinical trials suggest a role for various natural oral and parenteral estrogens in enhancing depressive mood (Dören 1999). Only a few of the published studies fulfill basic methodological quality requirements (see above; von Sydow and Reimer 1995; Greendale et al. 1999). A review of 111 studies revealed no consistent effect of HT on depression in patients with natural menopause and some effects in women with surgical menopause (Pearce et al. 1995). The Cochrane Library – based on a meta-analysis of Zweifel and O’Brien (1997; 2000) of no more than 26 studies with sample sizes of 10–110 which used at least one method to quantify depression – concluded that overall, there seems to be a moderate effect of estrogens (mostly conjugated equine estrogens 0.625 mg or 1.25 mg). However, the authors rated the quality of data as too poor to allow for general recommendations. Only 18 studies were randomized, a third of all studies included no control group. Three studies included women without any menopausal symptoms, seven other studies did not comment on this topic. Most patients were recruited from special menopause clinics with the above-mentioned high proportion of women with surgical menopause. In addition, most women included in the studies were not depressed or only mildly depressed! There seems to be no proven efficacy of estrogens in moderate and severe depression. However, estrogens appear to exert positive effects on well-being, which has been described as a “tonic” effect by Utian (1972). Recent studies found an increase of vigilance measured with electroencephalography (Saletu et al. 1995). A corresponding survey finding could be that women who depend on their cognitive skills show a more positive attitude to HT (Collins and Landgren 1997; for review, Stoppe et al. 2000). This view is now supported by the results of the Women’s Health Initiative (WHI) study, which assigned 16.608 postmenopausal women (50–79 years) with an intact uterus to HT (0.625 mg conjugated equine estrogen plus 2.5 mg medroxyprogesterone acetate). There were no significant effects on general health, vitality, mental health, depressive symptoms, or sexual satisfaction. After one, but not after three years HT resulted in statistically significant however small and clinically not meaningful benefit in terms of sleep disturbances and physical functioning (Hays et al. 2003). The authors discussed that women with a “positive” view on HT might have been reluctant to accept randomization, pointing again to the selection problem discussed above.

212

Estrogen Therapy in Perimenopausal Affective Disorders

The efficacy of estrogens seems to be demonstrated in surgically menopausal women. These women can be treated with unopposed estrogens, without coadministration of a progestogen, which may be a reason for this finding. In the WHI study the estrogen-only treatment is still ongoing. Results are expected by the year 2005. However, some authors stress the importance of the acuteness of hormonal changes for the occurrence of symptoms and therapy (Rubinow and Schmidt 1995). In addition, the relative depletion of androgens in ovarectomized women should be discussed (Khastgir and Studd 1998). According to representative studies, the percentage of surgical menopausal women varies between 3% (Japan), 14% (Great Britain), 19.7% (Sweden), 26% (Germany) and 34% (USA), and about every fifth hysterectomized woman is also ovarectomized (Stoppe et al. 2000; Khastgir and Studd 1998). Progestins appear to reduce the effects on the central nervous system when combined with estrogens (Archer 1999; Zweifel and O’Brien 1997). Data are insufficient to distinguish between the two major types of progestins: compounds derived from hydroxyprogesterone and nortestosterone. Androgens may be at least of similar effectiveness as estrogens. Most data are available for longer-acting parenteral testosterone preparations such as implants and injectables. However, due to a paucity of studies that directly compare various treatment modalities – estrogen versus estrogen plus progestin, estrogen versus estrogen plus androgen – great caution is warranted in qualifying any form of hormone therapy according to its ability to influence depressed mood. In addition, the schedule/ timing of use – continuous versus sequential – and the dosage should be considered (Halbreich 1997). A completely unresolved question is the issue of any potential, optimal combination of (the type of) antidepressants and hormones, because there are at least some studies suggesting a role for estrogens as adjunct in the treatment of depression. However, this is a controversial debate (Sherwin 1991; Schneider et al. 1997; Stahl 1998; Snow et al. 2000). New data add to the complexity of this topic showing that the selective serotonin reupteka inhibitor and anti-depressant paroxetine is effective in the treatment of vasomotor symptoms (Stearns et al. 2003). DHEA, thought to be a neurosteroid because it is possibly synthesized in the central nervous system of humans (Lacroix et al. 1987), has been suggested for treatment of major depression on the basis of

213

Gabriela Stoppe and Martina Dören

a few small-scale preliminary studies (for review, see Wolf et al. 1999). Recent studies looking at the sense of well-being including mood in pre- and postmenopausal women yielded different findings. Placebo-controlled trials suggested a small positive effect (Morales et al. 1994) or no effect at all (Wolf et al. 1997; Barnhart et al. 1999). Whether DHEA may act synergistically with HT (Stomati et al. 1999) has not been adequately studied.

Pathophysiological Mechanisms Unfortunately, the above mentioned meta-analysis does not appear to address the so-called domino effect, which denotes the observation that the alleviation of vasomotor symptoms and sleep disturbances may lead to an improvement of depressed mood (Campbell and Whitehead 1977). The majority of studies included for the metaanalysis investigated women with vasomotor symptoms; thus, it is quite possible that the alleviation of these most common symptoms of the menopause contributed to the observed impact on symptoms of depressed mood. The selection of studies using validated instruments to assess depressive symptoms does not exclude this possibility. According to representative surveys, the number of women reporting sleep complaints increases during the menopausal transition (Porter et al. 1996; Kuh et al. 1997; Ledésert et al. 1995; Hunter 1990). Many authors underlined the importance of a body mass index between 20 and 30 for better sleep (Adam 1987; Owens and Matthews 1998; Polo-Kantola et al. 1999). As a consistent finding, women taking HT reported favorable effects on sleep quality (Asplund and Aberg 1995; Wiklund et al. 1992; Polo-Kantola et al. 1998). Estrogens are involved in the regulation of sleep, particular rapid eye movement (REM) sleep (Thomson and Oswald 1977; Schiff et al. 1979; Purdie et al. 1995). Progesterone affects primarily non-REM sleep. It may have a sedative effect at high doses and may change sleep architecture, similar to benzodiazepines, including shortening the sleep onset and reducing wakefulness after sleep onset (Halbreich 1997). While most studies have limited value for methodological reasons (only self report, small number etc.), one recent study (n = 62) tested transdermal HT versus placebo in a crossover design. While subjective sleep quality increased, the number of nocturnal arousals decreased (Polo-Kantola et al. 1999). Another recent study suggests that trans-

214

Estrogen Therapy in Perimenopausal Affective Disorders

dermal estradiol restores the normal sleep electroencephalogram pattern (Antonijevic et al. 2000). Thus, improvements of sleep quality may positively influence mood and substantially contribute to the “tonic” effect by enhancement of daytime vigilance and everyday performance (see above). As reported in other chapters in this book, estrogens are thought to modulate neurotransmitter activities in several ways, e.g., including the synthesis of serotonin, serotonin uptake, serotonin receptor transcription and receptor density, and the response to serotonin stimulation (Gallo et al. 1999; Archer 1999). Regarding menopause, earlier animal studies showed that the receptor dysfunction increases with the duration of hormone deficiency (Clark et al. 1981). However, the question of whether, and after which time span, these dysfunctions are reversible by exogenous hormones is still unresolved, both in animals and in humans, and may be significant in menopausal research (Klaiber et al. 1997; Barrett-Connor et al. 1999).

Conclusion The occurrence of depressive symptoms in the perimenopause is associated with a variety of factors. A previous history of either depression and/or premenstrual syndrome as well as cognitive factors (e.g., attitude to menopause) explain most of the variance. There are no consistent findings of a correlation between any serum hormone level and severity or presence of mood symptoms. Neurobiological studies show promising effects of estradiol – with regard to an antidepressant effect – on serotonergic, noradrenergic, cholinergic, dopaminergic, and GABAergic functions. Progestogens seem to oppose some of these effects. The role of androgenic hormones and DHEA(-S) is less clear. Clinical trials showed, in general, a modest effect on symptoms of depression. However, the predominantly poor methodological quality does not allow generalization and recommendations. A “tonic” effect on well-being in non- or mildly depressed women should not be regarded as a true antidepressant effect. Results from studies of surgically menopausal women may not be applicable to women with natural menopause. There is a great potential for exploring various types, doses, and routes of administration of both antidepressants and sex hormones. With regard to the domino

215

Gabriela Stoppe and Martina Dören

theory, future studies should also focus on the mediation of treatment effects through alleviation of vasomotor symptoms or sleep disturbances. These conclusions are supported by the large Women’s Health Initiative (WHI) study, which revealed no benefit of combined estrogen-progestin treatment in terms of mental health and depressive symptoms, and only small positive effects on sleep disturbances.

References Abraham S, Llewellyn-Jones D, Perz J (1994) Changes in Australian women’s perception of the menopause and menopausal symptoms before and after the climacteric. Maturitas 20: 121-128 Adam K (1987) Total and percentage REM sleep correlate with body weight in 36 middle-aged people. Sleep 10: 69-77 American Psychiatric Association (1994) Diagnostic and statistical manual of mental disorders, 4th ed. (DSM-IV). American Psychiatric Association, Washington DC Anderson E, Hamburger S, Liu JH, Rebar RW (1987) Characteristics of menopausal women seeking assistance. Am J Obstet Gynecol 156: 428-433 Angst J, Dobler-Mikola A (1984) Do the diagnostic criteria determine the sex ratio in depression? J Affect Disord 7: 189-198 Antonijevic IA, Stalla GK, Steiger A (2000) Modulation of the sleep electroencephalogram by estrogen replacement in postmenopausal women. Am J Obstet Gynecol 182: 277-282 Archer JSM (1999) Relationship between estrogen, serotonin, and depression. Menopause 6: 71-78 Aro H (1994) Risk and protective factors in depression: a developmental perspective. Acta Psychiatr Scand (Suppl 377): 59-64 Arpels JC (1996) The female brain hypoestrogenic continuum from the premenstrual syndrome to menopause. A hypothesis and review of supporting data. J Reprod Med 41: 633-639 Asplund R, Åberg HE (1995) Body mass index and sleep in women aged 40 to 64 years. Maturitas 22: 1-8 Avis NE, Brambilla D, McKinlay SM, Vass K (1994) A longitudinal analysis if the association between menopause and depression: results from the Massachusetts Women’s Health Study. Annu Epidemiol 4: 214-220 Avis NE, McKinlay SM (1991) A longitudinal analysis of women’s attitudes toward the menopause: results from the Massachusetts Women’s Health Study. Maturitas 13: 65-79 Ballinger CB (1975) Psychiatric morbidity and the menopause: screening of a general population sample. BMJ 3: 344-346 Ballinger CB, Browning MCK, Smith AHW (1987) Hormone profiles and psychological symptoms in perimenopausal women. Maturitas 9: 235-251

216

Estrogen Therapy in Perimenopausal Affective Disorders Barnhart KT, Freeman E, Grisso JA, Rader DJ, Sammel M, Kapoor S, Nestler JE (1999) The effect of dehydroepiandrosterone supplementation to symptomatic perimenopausal women on serum endocrine profiles, lipid parameters, and healthrelated quality of life. J Clin Endocrinol Metab 84: 3896-3902 Barrett-Connor E, Von Muhlen D, Laughlin GA, Kripke A (1999) Endogenous levels of dehydroepiandrosterone sulfate, but not other sex hormones, are associated with depressed mood in older women: The Rancho Bernardo Study. J Am Geriatr Soc 47: 685-691 Born L, Steiner M (1999) Irritability: the forgotten dimension of female-specific mood disorders. Arch Women Ment Health 2: 153-167 Breslau N, Schultz L, Peterson E (1995) Sex differences in depression: a role for preexisting anxiety. Psychiatry Res 58: 1-12 Bungay GT, Vessey MP, McPherson MK (1980) Study of symptoms in middle-life with special reference to the menopause. BMJ 2: 181-183 Busch CM, Zonderman AB, Costa PT Jr (1994) Menopausal transition and psychological distress in a nationally representative sample: is menopause associated with psychological distress ? J Aging Health 6: 209-228 Campbell S, Whitehead M (1977) Estrogen therapy and the menopausal syndrome. Clin Obstet Gynecol 4: 31-47 Cawood EHH, Bancroft J (1996) Steroid hormones, the menopause, sexuality and well-being of women. Psychol Med 26: 925-936 Collins A, Landgren BM (1997) Psychosocial factors associated with the use of hormonal replacement therapy in a longitudinal follow-up of Swedish women. Maturitas 28: 1-9 Coope J (1981) Is oestrogen therapy effective in the treatment of menopausal depression? J R Coll Gen Pract 31: 134-140 Dennerstein L (1996) Well-being, symptoms and the menopausal transition. Maturitas 23: 147-157 Dennerstein L, Smith AM, Morse C, Burger H, Green A, Hopper J, Ryan M (1993) Menopausal symptoms in Australian women. BMJ 159: 232-236 Dennerstein L, Smith AMA, Morse C (1994) Psychological well-being, mid-life and the menopause. Maturitas 20: 1-11 Dören M (1999) Effects of SERMs on the uterus and menopausal symptoms. J Endocrinol Invest 22: 625-635 Gater RA, Dean C, Morris J (1989) The contribution of childbearing to the sex difference in first admission rates for affective psychosis. Psychol Med 19: 719-24 Greendale GA, Lee NP, Arriola ER (1999) The menopause. Lancet 353: 571-580 Halbreich U (1997) The role of estrogen in postmenopausal depression. Neurology 48 (Suppl 7): S16-S20 Hallström T, Samuelsson S (1985) Mental health in the climacteric: the longitudinal study of women in Gothenburg. Acta Obstet Gynecol Scand 130: 13-18 Hamilton J, Halbreich U (1993) Special aspects of neuropsychiatric illness in women: Focus on depression. Ann Rev Med 44: 355-364 Harlow BL, Signorello LB (2000) Factors associated with early menopause. Maturitas 35: 3-9

217

Gabriela Stoppe and Martina Dören Harris T, Surtees P, Bancroft J (1991) Is sex necessarily a risk factor for depression? Br J Psychiatry 158: 708-712 Hay AG, Bancroft J, Johnstone EC (1994) Affective symptoms on women attending a menopause clinic. Br J Psychiatr 164: 513-516 Hays J, Ockene JK, Brunner RL, Kotchen JM, Manson JE, Patterson RE, Aragaki AK, Shumaker SA, Brzyski RG, La Croix AZ, Granek IA, Valanis BG, for the Women’s Health Initiative Investigators (2003) Effects of estrogen plus progestin on healthrelated qualitiy of life. N Engl J Med 348: 1839-1854 Heinemann LAJ, Zimmermann T, Vermeulen A, Thiel C, Hummel W (1999) A new ‘Aging Males’ Symptom Rating Scale. The Aging Male 2: 105-114 Holte A (1992) Influences of natural menopause on health complaints: a prospective study of healthy Norwegian women. Maturitas 14: 127-141 Holte A, Mikkelsen A (1982) Menstrual coping style, social background and climacteric symptoms. Psychiatr Soc Sci 2: 41-45 Huerta R, Mena A, Malacara JM, Diaz de Leon J (1995) Symptoms at perimenopausal period: its association with attitudes toward sexuality, life-style, family function, and FSH levels. Psychoneuroendocrinol 20: 135-148 Hunter M (1990) Somatic experience of the menopause: a prospective study. Psychosom Med 52: 357-367 Hunter M (1992) The South-East England longitudinal study of the climacteric and postmenopause. Maturitas 14: 117-126 Huppert FA, van Niekerk JK, Herbert J (2000) Dehydroepiandrosterone (DHEA) supplementation for cognition and well-being (cochrane review). In: The Cochrane Library, issue 1, Oxford: update software CD 000304 Jorm AF (1987) Sex and age differences in depression: a quantitative synthesis of published research. Austral N Zeal J Psychiatry 21: 46-53 Kaufert P, Boggs PP, Ettinger B, Fugate Woods N, Utian WH (1998) Women and menopause: beliefs, attitudes, and behaviors. The North American Menopause Society 1997 Menopause Survey. Menopause 5: 197-202 Kaufert PA, Gilbert P, Tate R (1992) The Manitoba project: a re-examination of the link between menopause and depression. Maturitas 14: 143-155 Kessler R, McGonagle R, Swartz M, Blazer DG, Nelson CB (1993) Sex and depression in the National Comorbidity Survey: lifetime prevalence, chronicity, and recurrence. J Affect Disord 29: 85-96 Khastgir G, Studd J (1998) Hysterectomy, ovarian failure, and depression. Menopause 5: 113-122 Klaiber EL, Broverman DM, Vogel W, Peterson LG, Snyder MB (1997) Relationships of serum estradiol levels, menopausal duration, and mood during hormonal replacement therapy. Psychneuroendocrinol 22: 549-558 Kuh DL, Wadsworth M, Hardy R (1997) Women’s health in midlife: the influence of the menopause, social factors and health in earlier life. Br J Obstet Gynaecol 104: 923-933 Lacroix C, Fiet J, Benais JP, Gueux B, Bonete R, Villette JM, Gourmel B, Dreux C (1987) Simultaneous radioimmunoassays of progesterone, androst-4-enedione, pregnenolone, dehydroepiandrosterone and 17-hydroxyprogesterone in specific regions of human brain. J Steroid Biochem 28: 317-325

218

Estrogen Therapy in Perimenopausal Affective Disorders Lalive d’Epinay CJ (1985) Depressed elderly women in Switzerland: an example of testing and of generating theories. Gerontologist 25: 597-604 Ledésert B, Ringa V, Bréat G (1995) Menopause and perceived health status among the women of the French GAZEL cohort. Maturitas 20: 113-120 Leon AC, Klerman GL, Wickramaratne P (1993) Continuing female predominance in depressive illnes. Am J Public Health 83: 754-757 Li C, Samsioe G, Lidfeldt J, Nerbrand C, Agardh CD (2000) Important factors for use of hormone replacement therapy. A population-based study of Swedish women. The Women’s Health in Lund Area (WHILA) Study. Menopause 7: 273-281 Makhlouf Obermeyer C (2000) Menopause across cultures: a review of the evidence. Menopause 7: 184-192 Manber R, Armitage R (1999) Sex, steroids, and sleep: a review. Sleep 22: 540-555 Matthews KA (1992) Myths and realities of the menopause. Psychosom Med 54: 1-9 Matthews KA, Kuller LH, Wing RR, Meilahn EN (1994) Biobehavioral aspects of menopause: lessons from the healthy women study. Exp Gerontol 29: 337-342 McKinlay JB, McKinlay SM, Brambilla D (1987) The relative contributions of endocrine changes and social circumstances to depression in mid-aged women. J Health Soc Behav 28: 345-363 Montgomery JC, Appleby L, Brincat M, Versi E, Tapp A, Fenwick PB, Studd JW (1987) Effect of oestrogen and testosterone implants on psychological disorders on the climacteric. Lancet: 297-299 Morales A, Nolan J, Nelson J, Yen S (1994) Effects of replacement dose of dehydroepiandrosterone in men and women of advancing age. J Clin Endcorinol Metab 78: 1360-1367 Morse CA, Smith A, Dennerstein L, Green A, Hopper J, Burger H (1994) The treatment-seeking woman at menopause. Maturitas 18: 161-173 Neri I, Demyttenaere K, Facchinetti F (1997) Coping style and climacteric symptoms in a clinical sample of postmenopausal women. J Psychosom Obstet Gynecol 18: 229-233 Neugarten BL, Kraines RJ (1965) Menopausal symptoms of women of various ages. Psychosom Med 27: 266-273 Nicol-Smith L (1996) Causality, menopause, and depression: a critical review of the literature. Br Med J 313: 1229-1232 Notelovitz M (1999) Discovery and identification of estrogen. Menopause 6: 2-3 Owens JF, Matthews KA (1998) Sleep disturbances in healthy middle-aged women. Maturitas 30: 41-50 Parry B (1989) Reproductive factors affecting the course of affective illness in women. Psychiatr Clin North Am 12: 207-220 Pearce J, Hawton K, Blake F (1995) Psychological and sexual symptoms associated with the menopause and the effects of hormone replacement therapy. Br J Psychiatry 167: 163-173 Pearlstein T, Rosen K, Stone AB (1997) Mood disorders and menopause. Endocrinol Metab Clin North Am 26: 279-294 Polit DF, LaRocco SA (1982) Social and psychological correlates of menopausal symptoms. Psychosom Med 42: 335-345

219

Gabriela Stoppe and Martina Dören Polo-Kantola P, Erkkola R, Helenius H, Irjala K, Polo O (1998) When does estrogen replacement therapy improve sleep quality? Am J Obstet Gynecol 178: 1002-1009 Polo-Kantola P, Erkkola R, Irjala K, Pillinen S, Virtanen I, Polo O (1999) Effect of short-term transdermal estrogen replacement therapy on sleep: a randomised, double-blind crossover trial in postmenopausal women. Fertil Steril 71: 873-880 Porter M, Penney GC, Russell D, Russell E, Templeton A (1996) A population based survey of women’s experience of menopause. Br J Obstet Gynaecol 103: 1025-1028 Purdie DW, Empson JA, Crichton C, Macdonald L (1995) Hormone replacement therapy, sleep quality and psychological wellbeing. Br J Obstet Gynaecol 102: 735-739 Rubinow DR, Schmidt PJ (1995) The neuroendocrinology of menstrual cycle mood disorders. Ann N Y Acad Sci 771: 648-659 Saletu B, Brandstaetter N, Metka M, Stamenkovic M, Anderer P, Semlitsch HV, Heytmanek G, Huber J, Grunberger J, Linzmayer L (1995) Double-blind, placebocontrolled, hormonal, syndromal and EEG mapping studies with transdermal oestradiol therapy in menopausal depression. Psychopharmacol 122: 321-329 Schiff I, Regestein Q, Tuchinsky D, Ryan KJ (1979) Effects of estrogens on sleep and psychological state of hypogonadal women. J Am Med Assoc 245: 1741-1744 Schneider L, Small G, Hamilton S, Bystritsky A, Nemeroff C, Myers B (1997) Estrogen replacement and response to fluoxetine in a multicenter geriatric depression trial: fluoxetine collaborative study group. Am J Psychiatry 5: 97-106 Schultz-Zehden B (1998) Frauengesundheit in und nach den Wechseljahren. Die 1000-Frauen-Studie. Kempkes, Gladenbach Sherwin B (1991) Estrogen and refractory depression. In: Amsterdam JD (ed) Advances in neuropsychiatry and psychopharmacology, vol 2: Refractory depression. Raven Press, New York, 209-218 Silverstein B, Perlick G (1991) Gender differences in depression: historical changes. Acta Psychiatr Scand 84: 327-331 Simon G, Ormel J, von Korff M, Barlow M (1995) Health Care costs associated with depressive and anxiety disorders in primary care. Am J Psychiatry 152: 352-357 Snow V, Lascher S, Mottur-Pilson C (2000) Pharmacologic treatment of acute major depression and dysthymia. American College of Physicians – American Society of Internal Medicine. Ann Intern Med 132: 738-742 Stahl SM (1998) Basic neuropharmacology of antidepressants, part 2: estrogen as an adjunct to antidepressant treatment. J Clin Psychiatry 59 (Suppl 4): 15-24 Stearns V, Beebe KL, Iyenger M, Dube E (2003) Paroxetine controlled release in the treatment of menopausal lot flush. A randomized controlled trial. J Am Med Assoc 289: 2827-2834 Stewart DE, Boydell K (1993) Psychologic distress during menopause. Associations across the reproductive life cycle. Int J Psychiatr Med 23: 157-162 Stewart DE, Boydell K, Derzko C, Marshall V (1992) Psychologic distress during the menopausal years in women attending a menopause clinic. Int J Psychiatr Med 22: 213-220 Stomati M, Rubino S, Spinetti A, Parrini D, Luisi S, Casarosa E, Petraglia F, Genazzani AR (1999) Endocrine, neuroendocrine and behavioural effects of oral

220

Estrogen Therapy in Perimenopausal Affective Disorders dehydroepiandrosterone sulfate supplementation in postmenopausal women. Gynecol Endocrinol 13: 15-25 Stommel M, Given BA, Given CW, Kalaian HA, Schulz R, McCorkle R (1993) Gender bias in the measurement properties of the Center for Epidemiologic Studies Depression Scale (CES-D). Psychiatry Res 49: 239-250 Stoppe G, Sandholzer H, Huppertz C, Duwe H, Staedt J (1999) Gender differences in the recognition of depression in old age. Maturitas 32: 205-212 Stoppe G, von Sydow K, Krasney N (2000) Die Psyche in der Peri- und Postmenopause. Reproduktionsmedizin 16: 253-260 Thomson J, Oswald I (1977) Effect of estrogen on sleep, mood and anxiety of menopausal women. Br Med J 2: 1317-1319 Topo P, Koster A., Holte A, Collins A, Landgren BM, Hemminmi E et al. (1995) Trends in the use of climacteric and postclimacteric hormones in nordic countries. Maturitas 22: 89-95 Utian WH (1972) The mental tonic effect of oestrogens administered to oophorectomized females. S Afr Med J 46: 1979-1082 van Duijn CM (1997) Menopause and the brain. J Psychosom Obstet Gynaecol 18: 121-125 van Hall EV, Verdel M, van der Velden J (1994) “Perimenopausal” complaints in women and men: a comparative study. J Women’s Health 3: 45-55 von Sydow K, Reimer C (1995) Psychosomatik der Menopause: Literaturübersicht 19882–19992. Psychother Psychosom Med Psychol 45: 225-236 Weissman M, Leaf PJ, Tischler GL, Blazer DG, Kerno M, Bruce ML, Florio LP (1988) Affective disorders in five United States communities. Psychol Med 18:141-153 Wells JE, Bushnell JA, Hornblow AR, Joyce PR, Oakley-Browne MA (1989) Christchurch psychiatric epidemiology study. Part I. Methodology and lifetime prevalence for specific psychiatric disorders. Austral N Zeal J Psychiatry 23: 315-326 WHO – World Health Organization (1991) Mental and behavioural disorders (including disorders of psychological development). Clinical descriptions and diagnostic guidelines. 10th revision of the International Classification of diseases, Chap V (F), Geneva Wiklund I, Berg G, Hammar M, Karlberg-J, Lindgren R, Sandin K (1992) Long-term effect of transdermal hormonal therapy on aspects of quality of life in postmenopausal women. Maturitas 14: 225-236 Wilbur JE, Dan A, Hedricks C, Holm K (1990) The relationship among menopausal status, menopausal symptoms, and physical activity in midlife women. Fam Community Health 13: 67-78 Wilhelm K, Parker G (1989) Is sex necessarily a risk factor for depression? Psychol Med 19: 401-413 Wilhelm K, Parker G (1994) Sex differences in lifetime depression rates: fact or artefact? Psychol Med 24: 97-111 Williams JB, Spitzer RL, Linzer M et al. (1995) Gender differences in depression in primary care. Am J Obstet Gynecol 173: 654-659 Wolf OT, Kirschbaum C (1999) Actions of dehydroepiandrosterone and its sulfate in the central nervous system: effects on cognition and emotion in animals and humans. Brain Res Rev 30: 264-288

221

Estrogen Therapy in Perimenopausal Affective Disorders Wolf OT, Neumann O, Hellhammer DH, Geiben AC, Strasburger CJ, Dressendörfer RA, Pirke K-M, Kirschbaum C (1997) Effects of a two-week physiological dehydroepiandrosterone substitution on cognitive performance and well-being in healthy elderly man and women. J Clin Endocrinol Metab 82: 2363-2367 Young MA, Fogg LF, Scheftner WA, Keller MB, Fawcett JA (1990) Sex differences in the lifetime prevalence of depression: does varying the diagnostic criteria reduce the female/male ratio? J Affect Disord 18: 187-192 Zweifel JE, O´Brien WH (1997) A meta-analysis of the effect of hormone replacement therapy upon depressed mood. Psychoneuroendocrinology 22: 189-212 (see also: The Cochrane Library 2000; 1: 1–5)

222

11 The Effects of Estrogens on Cognition and Alzheimer’s Dementia Tony Edwin and Uriel Halbreich

Introduction Several lines of evidence suggest that mood and behavior effects of estrogen reflect the direct action of this hormone on neurons and other components of the central nervous system (CNS). Elements within the CNS that are responsive to estrogens encompass structural neuronal systems, blood flow, energy delivery and utilization, dendrite activity, neurotransmitters, and neuromodulators as well as intraneuronal and other processes. Estrogens readily cross the blood-brain barrier where they interact with nuclear estrogen receptors present in neuronal populations from different brain regions and with membrane-bound receptors (Sherwin 1997; Kawata 1995). Estrogen modulates growth proteins specifically associated with axonal elongation (Shugrue and Dorsa 1993), enhances the outgrowth or nerve processes in cultured neurons (ToranAllerand 1984), and promotes the formation of dendrite spines and synapses (Chung et al. 1988). The viability of in vitro cultures of differentiated amygdala (Arimatsu and Hatamaka 1986) or hypothalamic (Faivre-Bauman et al. 1981) neurons is prolonged by the addition of estrogen. Neurite outgrowth in the developing brain is stimulated by estrogens (Toran-Allerand 1993). Dendritic growth is stimulated by estrogen and is responsive to the hormonal fluctuations along the estrous cycle (McEwen et al. 1997). The physiological relevance of some of these estrogen effects is suggested by enhanced long-term potentiation (Warren et al. 1995), in parallel with increased synapse formation in the CA1 region of the hippocampus (Woolley and McEwen 1993) during the proestrus (high-estrogen) phase of the rat estrus cycle. Estrogen influences several neurotransmitter systems, including those using acetylcholine (Singh et al. 1994), norepinephrine (Sar and Stumpf 1981), serotonin (Kendall et al. 1981), dopamine (Toran-Allerand 1984), and other neurotransmitters.

223

Tony Edwin and Uriel Halbreich

Cholinergic and noradrenergic neurons from the basal forebrain, noradrenergic neurons from the brain stem locus caeruleus, and serotonergic neurons from the midbrain regions are all substantially affected by estrogens (Coyle et al. 1983). Estrogen’s interactions with the cholinergic systems are especially noteworthy in the context of this chapter. Cholinergic mechanisms are critically involved in attentional processes, learning, and memory: cognitive functions that are critically affected by Alzheimer’s disease (AD) (Bartus et al. 1981). Basal forebrain cholinergic neurons possess nuclear receptors for estrogen and low-affinity receptors for nerve growth factor (Torran-Allerand et al. 1992). Nerve growth factor prevents atrophy of cholinergic neurons after experimental injury (Hefti et al. 1993), and estrogens may regulate or modulate neurotrophins (Sorabji et al. 1994). Even though the hippocampus contains few estrogen and progestin receptors, this structure displays a robust response to exposure to exogenous estrogen and progestin treatment and to endogenous ovarian steroids during the natural estrous cycle (Toran-Allerand 1984). This first became apparent with the finding of cyclic variations in the threshold of the dorsal hippocampus to elicitation of seizures, with the greatest sensitivity occurring during proestrus (Terasawa and Timiras 1968). Morphologic studies indicate that estrogen induces dendritic spines and new synapses in the venteromedial hypothalamus of the female rat but also increases density of dendritic spines on pyramidal neurons in the hippocampus (McEwen and Woolley 1994). Spine density also changes cyclically during the estrus cycle of the female rat. These findings indicate that synapses are formed and broken down rapidly during the natural reproductive cycle. It is of significance that several gender differences in brain development are regulated by estrogen. In the CAl region of males, estrogen treatment fails to induce as great a number of spine synapses as in females, but blockade at birth of the aromatization of testosterone to estradiol in male neonates increases the number of spine synapses induced by estrogen treatment in adulthood (Wong and Moss 1992). This suggests that the responsiveness of the hippocampus to estrogenic regulation of synapse formation is defeminized in males by the neonatal actions of testosterone. Ovarian steroids regulate the midbrain serotonergic system by mechanisms as yet undefined. Estrogen increases serotonergic post-

224

The Effects of Estrogens on Cognition and Alzheimer’s Dementia

synaptic responsivity (Rosencrans 1970) and increases both the number of serotonergic receptors and neurotransmitter uptake (Rosencrans 1970). Estrogen also increases 5-hydroxytryptamine (5-HT) synthesis and 5-HTlA levels; it upregulates 5-HT1 receptors and downregulates 5-HT2 receptors. It decreases monoamine oxidase activity (Chakravorthy and Halbreich 1997) and increases the responsitivity of postsynaptic receptors to stimulation with the serotonergic agonist mCPP (Halbreich et al. 1995. The cumulative effect of estrogen on serotonergic function is as a 5-HT agonist (Kahn and Halbreich 1999).

Estrogen and Human Cognition Sex differences in brain structure and brain function including cognition are believed to result from differential exposure of men and women to sex hormones during fetal life (Chung et al. 1988). On average, men excel in spatial and quantitative abilities and in gross motor strength, whereas women excel in verbal abilities, in perceptual speed and accuracy, and in fine motor skills (Jarvik 1975). It is important to note, however, that the magnitude of the sex differences in test performance is modest and ranges between 0.25 and 1.0 standard deviations (SD) from normative scores (Chung et al. 1988). Sex differences in specific cognitive abilities in healthy men and women are probably due to prenatal influences on brain organization. This information can be extrapolated from studies of individuals affected by genetic syndromes that altered, in some manner, the concentration of gonadal hormones to which the fetuses were exposed, such as in congenital adrenal hyperplasia (CAH) (Chung et al. 1988). Fetuses with CAH are exposed to high levels of adrenal androgen production because of their adrenocorticotropic hormone (ACTH) levels. Predictably, women with CAH have a lower verbal IQ score (Perlman 1973), increased spatial abilities (Resnik et al. 1986), and an increased frequency of specific learning disabilities (dyscalculia) (Nass and Baker 1991) compared to their unaffected sisters. It may be inferred from this finding that the prenatal hormonal environment influences gender differences in certain cognitive functions. Studies of short-term verbal memory in humans indicate that estrogens have beneficial effects in women whose

225

Tony Edwin and Uriel Halbreich

estrogen levels have been reduced by various means (e.g., physiological postmenopausal status or “medical menopause” by administration of gonadotropin-releasing hormone [GnRH] analogues] (Sherwin 1994; Sherwin and Tulandi 1996). An increased estrogen level in women has also been associated with better performance on tests of fine motor skills and somewhat poorer performance on tests of spatial recognition (Hampson and Kimura 1988). The selective deterioration of some cognitive functions in menopausal women might suggest an association with low levels of estrogen and not age per se. Several cognitive functions such as delayed recall, visual reproduction, digit span, as well as spatial ability and attention span or visual memory are probably not improved with estrogen replacement therapy (ERT). Since the early 1950s, it has been suggested that estrogen’s influence might be diversified and may depend on the cognitive construct studied (Caldwell and Watson 1952). In most of the cases, when various tests of memory were studied, a positive effect of estrogen was reported (Hackman and Galbraith 1976; Campbell and Whitehead 1977; Fedor-Freybergh 1977; Phillips and Sherwin 1992). However, this improvement is not generalized. For example, some women who had improved immediate recall and association learning in response to estradiol (Phillips and Sherwin 1992) did not show improvement of delayed recall, visual reproduction, or digit span. This was demonstrated by Kampen and Sherwin (1994), who found that verbal memory improved, whereas spatial ability and attention span did not. In a classic study, Sherwin and Phillips (1990) tested estrogen’s effect on cognition in premenopausal women who underwent total abdominal hysterectomy (TAH) and bilateral salpingo-oophorectomy (BSO) for benign disease. They administered the Paragraph Recall Test (for verbal memory) to women having a normal estradiol level before undergoing surgery. This assessment was repeated after the surgical procedure, when women were randomly assigned to receive intramuscular androgen, combined estrogen – androgen, estrogen, or placebo for 3 months. Sherwin and Phillips found that women receiving any of the single or combination hormone treatments following the ovariectomy had at least maintained their baseline paragraph recall scores, whereas those women receiving placebo after removal of their uterus and ovaries had a statistically significant decrease in scores. From these results, Sherwin and

226

The Effects of Estrogens on Cognition and Alzheimer’s Dementia

Phillips then undertook a similar comparison with a more comprehensive battery of neuropsychologic tests. Their findings indicate a significant decrease in paired association scores, tests for memory, and capacity to form new associations in women on placebo who maintained baseline levels in the paragraph recall test. However, women who received estrogen scored higher than at baseline. Two studies failed to find any effects of estrogen on memory in postmenopausal women. Ditkoff and colleagues (Ditkoff et al. 1991) administered either 0.625 mg conjugated equine estrogen (CEE), 1.25 mg CEE, or placebo to postmenopausal women who had a previous TAH and who were not experiencing vasomotor symptoms. After 3 months of treatment, there were no within- or between-group differences in scores on the digit symbol or the digit span subtests of the Wechsler Adult Intelligence Scale, which were the only cognitive tests administered. Barret-Connor and Kritz-Silverstein (1993) administered a comprehensive battery of neuropsychologic tests to 800 women between the ages of 65 and 95 years who were in the Rancho Bernado cohort assembled in 1972–1974 to study heart disease risk factors. Almost half of this upper middle class cohort had used estrogen at some point after the menopause and one third were current users. Women who had used estrogen for at least 20 years had higher scores on the category fluency test, whereas those who were past users had significantly higher scores on the mini-mental state examination. However, no differences were found between the performance of past users, current users, or those who had never used estrogen in other tests of verbal memory or tests of visual memory. In a more recent study to determine whether endogenous hormone levels predict cognitive function in older women, BarrettConnor and Goodman-Gruen (1999) evaluated cognitive function in 393 community-dwelling women aged 55–89 years who were not using replacement estrogen. It was found that in these older women, higher endogenous estrogen levels were not associated with significantly better performance on any cognitive function test administered. Our own studies (Halbreich et al. 1993; Halbreich 1995; 1997) with an extensive battery of cognitive tests showed improvement in complex integrative cognitive tasks that require integration of several cognitive constructs, such as recognition, interpretation, decision making, eye-hand coordination, and reaction. We also found positive effects of estrogen on several short-term memory functions,

227

Tony Edwin and Uriel Halbreich

but not on tests of manual dexterity, visuospatial and verbal abilities, indicating a selective task-dependent effect of estrogen on cognition (Halbreich 1995).

Estrogen and Dementia of the Alzheimer’s Type Dementia of the Alzheimer’s type (DAT) is precipitated by a combination of genetic vulnerability and environmental and biological processes. Many of these mechanisms are influenced by estrogen. A genetic disposition is widely accepted. Aside from advanced age, the most consistently identified risk factor is a family history of dementia. An individual’s risk for developing AD is more than doubled when a parent or sibling is demented, and the risk estimate is even greater for two or more affected first-degree relatives (Pericak-Vance and Haines 1995). With respect to early-onset AD, it now appears that the illness can be attributed to point mutations in the chromosome 14 gene encoding a cell-membrane-spanning protein of uncertain function, referred to as presenilin-1. Less common are point mutations in the chromosome 1 gene for a homologous protein (presenilin-2) or the chromosome 21 gene for the amyloid precursor protein (Henderson 1997). These autosomal-dominant defects show virtually complete age-dependent penetrance and share similar clinical and pathologic phenotypes. In contrast, late-onset AD is infrequently associated with recognized mutations, but the risk for dementia in this older age group is strongly influenced by the polymorphism of apolipoprotein E (ApoE4), a plasma lipoprotein constituent encoded by chromosome 19. The observation that many identical twins are discordant for AD (Toran-Allerand 1984; Nee et al. 1987; Small et al. 1993) implies the existence of environmental factors or a multifactorial model leading to late-onset AD. Accumulating evidence suggests that for women, one such exogenous risk factor is postmenopausal estrogen deprivation. The insidious onset and progressive decline in mental and cognitive functions characterize DAT. It is more prevalent in women than in men (up to three times) (Jorm et al. 1987), even when adjusted for age. Although there is a genetic predisposition to DAT, environmental factors play an important role in the actual manifestation of the disorder, and some other risk factors have been identified (Jorm

228

The Effects of Estrogens on Cognition and Alzheimer’s Dementia

et al. 1987; Birge 1997). Some of these risk factors have been shown to be more prevalent among women and to be influenced by estrogen. They include female gender itself, hysterectomy, hip fracture, hypertension, myocardial infarction, diabetes, hypothyroidism, and increased hematocrit (Birge 1997) as well as depression. Therefore, it appears logical that the lack of estrogen would be implicated in increasing the risk for DAT and that ERT would be suggested as a preventive and treatment modality. So far, most of the evidence for a positive influence of ERT on DAT is derived from epidemiological studies showing a lower incidence of DAT in women who received ERT (Birge 1994; Henderson et al. 1994; Mortel and Meyer 1995; Lerner et al. 1995) as compared to those who did not. Thus, the use of ERT has been shown to reduce the relative risk (odds ratio) of developing DAT by about 0.5 (Paganini-Hill and Henderson 1996; Brenner et al. 1994; Morrison et al. 1996; Caldwell 1954) in most, but not all studies. The demonstration of dose and duration effects of ERT on the rate of DAT reinforces the credibility of the epidemiological data. Women who received higher doses of estrogen for longer times had lower risks for developing DAT (Paganini-Hill et al. 1996). Actual prospective treatment trials with estrogen in patients with dementia and specifically with DAT are still scarce, even though they date back to the early 1950s (Kantor et al. 1973). Nonetheless, their positive results are quite consistent (Birge 1997; Fillit et al. 1986; Honjo et al. 1989; Okhura et al. 1995; Schneider et al. 1996; Halbreich 1997). Available epidemiological studies have examined AD risk when information on postmenopausal estrogen use was collected mostly retrospectively before the presumptive onset of dementia symptoms. The Leisure World retirement community cohort in southern California was established by a postal survey in 1981 (Paganini-Hill and Henderson 1994). Detailed information on hormone use in this upper middle-class population was collected from each female participant at the time of enrollment, and death certificates have subsequently been obtained for virtually all deceased cohort members. From the records of 2,529 female cohort members who died before 1993, Paganini-Hill and Henderson (1994) identified diagnoses of AD and other diagnoses believed likely to represent AD. Estrogen users in this nested case-control study had about a 30% lower risk of developing AD or related diseases (odds ratio, 0.69; 95% CI, 0.49–1.03).

229

Tony Edwin and Uriel Halbreich

Morrison and coworkers (Morrison et al. 1996) identified 38 incident cases of AD from among 472 older women participating in the Baltimore Longitudinal Study of Aging. The use of oral and transdermal estrogen therapy was documented prospectively and, after adjusting for education, the relative risk of developing AD among “ever-users” of postmenopausal estrogen as compared to “neverusers” was reduced by over one half (odds ratio, 0.44; 95% CI, 0.20–0.97). Tang and colleagues (Tang et al. 1996) examined information on oral estrogen use for 1,124 cognitively intact older women. In this community-based northern Manhattan cohort, 167 incident cases of AD were identified over a follow-up period of 1–5 years. Postmenopausal use of oral estrogens was reported by 6% of women with AD and 16% of other cohort members. Among estrogen users, the risk of AD was reduced by 60% (odds ratio, 0.40; 95% CI, 0.22-0.85), and the age at onset of AD was significantly higher among estrogen users who became demented than never-users who became demented. However, in another case-control study derived from a large health maintenance organization population in the Puget Sound area of Washington state, Brenner and colleagues (Brenner et al. 1994) compared estrogen exposure between 107 incident cases of AD and 120 control subjects, who were frequency matched for age. They failed to confirm an association with postmenopausal ERT (everversus never-use as documented by computerized pharmacy records; odds ratio, 1.10; CI, 0.6–1.8). In a retrospective longitudinal study, Costa and Reus (1999) assessed cognitive functioning in female estrogen users and non-estrogen users (n = 3,128). It was found that at baseline, estrogen users had significantly lower rates of DAT diagnoses than non-estrogen users. ERT was significantly associated with higher cognitive functioning at baseline and at 1-year follow-up. Asthana and Craft (1999) evaluated the cognitive effects of transdermal estrogen in a small sample of postmenopausal women with DAT during an 8-week treatment period. The salutary effects of estrogen on cognition were observed after the first week of treatment, and started to diminish when treatment was terminated. Enhancement of verbal memory was positively correlated with plasma levels of estradiol. In addition, the authors suggest that several markers of neuroendocrine activity may serve to index the magnitude of estrogen-induced facilitation on cognition.

230

The Effects of Estrogens on Cognition and Alzheimer’s Dementia

The advantages are further supported by Henderson’s finding that the effects of HRT were dose and duration dependent: Women on a higher dose (1.25 mg of conjugated estrogen) had less relative risk for DAT than women who had not received ERT. Women who received HRT for more than 7 years had less relative risk for DAT than women who were treated for under 1 year. The hormone’s biological effects support the epidemiological and clinical reports on the efficacy of ERT for prevention and treatment of DAT. It has already been mentioned here that estrogen, which crosses the blood–brain barrier, has genomic intracellular signal transduction effects. It also has neurostructural effects on dendrites and axons as well as synapse level effects on several related neurotransmitters, especially acetylcholine and 5-HAT as well as cation channels. Therefore, the future of ERT for prevention of the deterioration of cognitive functioning and dementia is suggested to be quite certainly positive (Van Duijn and Hofman 1992). It is still unknown, however, whether estrogen would be effective as a treatment once DAT has developed. An important mechanism of estrogen’s improvement of cognitive function in women with DAT is the effect of the hormone on blood flow. The role of vascular dysfunction in the pathogenesis of dementia is now being revisited as a consequence of our greater ability to assess cerebral blood flow through advances in neuroimaging. Increasing evidence suggests that estrogen replacement could not only prevent the development of vascular disorder but also improve blood flow in women with existing vascular disorders (Mendelsohn and Karas 1999). These effects of estrogen are believed to be mediated through its direct effects on the endothelium and vasomotor function, specifically, the inhibition of the vasoconstrictor endothelin (Polderman et al. 1993) and the stimulation of the vasodilator endothelium-derived relaxing factor (van Buren et al. 1992). In postmenopausal women, estrogens administration increases cardiac output and systemic arterial blood flow, including internal carotid artery blood flow and cerebral blood flow (Ohkura et al. 1994). Because of the probable role of vascular disease in the pathogenesis of DAT and estrogen’s effect on cerebral blood flow, estrogen replacement is an attractive adjuvant for the treatment and prevention of DAT. Addition of estrogen to in vitro cultures of neurons results in stimulation of neurite outgrowth and promotion of neuron viability

231

Tony Edwin and Uriel Halbreich

(Arimatsu and Hatamak 1986). These effects are replicated in live experimental animal models. Estrogen may attenuate neuronal injury related to its effects on the metabolism of the amyloid protein (APP). APP is expressed during neuronal injury (Regland and Gottfries 1992). Deposition of β-amyloid in brain parenchyma is a distinctive feature of the neuropathology of AD but also occurs to a lesser extent in normal aging (Armstrong 1995). In an estrogenreceptor-containing cell culture system, 17β-estradiol at physiological concentrations increases the secretory metabolism of the soluble fragment of APP without increasing intracellular levels of APP. Therefore, estrogen may favorably modify the processing of APP, thereby reducing the accumulation of the neurotoxic β-amyloid fragment (Birge 1997).

Conclusions Although the hypothesis that postmenopausal ERT favorably affects a women’s risk for developing AD is attractive, the issue is far from being settled. With regard to biological credibility, much remains to be done in terms of understanding how potentially relevant estrogen actions might affect the pathogenesis or progression of AD. Evidence favoring the estrogen hypothesis has been strengthened by recent experimental and epidemiological findings. However, even in positive epidemiological studies, the effects are moderate and the possibility that undetected selection or observation biases account for the reported associations cannot be excluded. Because a number of demographic features and lifestyles choices distinguish estrogen users from non-users (Hemminki et al. 1993; Derby et al. 1995), confounding of the estrogen–AD association by other unidentified variables must also be considered. Hogervorst and colleagues (Hogervorst et al. 1999) recently reported that the design and type of memory tests could explain the controversial results of studies on the effect of HRT on cognitive function. Furthermore, the authors suggested that HRT has a global activating effect, rather than a specific direct effect on cognitive function. Consistency of findings in future cohorts would strengthen an inferred causal link between estrogen replacement and a lowered risk for AD. More convincing evidence would come from adequately powered, randomized, placebo-controlled, primary intervention

232

The Effects of Estrogens on Cognition and Alzheimer’s Dementia

trials, such as the one through the Women’s Health Initiative of the National Institutes of Health (Henderson 1997). However, given the distressing high prevalence of AD among older women and the exorbitant social and economic costs associated with this disorder, a true risk reduction in the order of one third to one half, as suggested by several recent analytical studies, would be of tremendous public significance. Even a delay of 5–10 years in onset of DAT will have a high impact on prevalence and public health consequences. Preliminary evidence from the use of Selective Estrogen Receptor Modulators (SERMs) such as Raloxifene, which has estrogen-like properties on hippocampal choline acetyltransferase (ChAT) in vivo, suggests that they exert a beneficial effect on cholinergic transmission without producing peripheral stimulation of breast and uterine tissue (Wu and Glinn 1999). The significance of this effect on cognitive function needs to be further explored.

References Arimatsu Y, Hatamaka H (1986) Estrogen treatment enhances survival of cultured amygdala neurons in a defined medium. Dev Brain Res 26: 151-159 Armstrong RA (1995) Beta-amyloid deposition in the medial temporal lobe in elderIy non-demented brains and in Alzheimer’s disease. Dementia 6: 121-125 Asthana S, Craft S et al. (1999) Cognitive and neuroendocrine response to transdermal estrogen in postmenopausal women with Alzheimer’s disease: results of a placebo controlled, double-blind, pilot study. Psychoneuroendocrinolgy 24: 657-677 Barret-Connor, Kritz-Silverstein D (1993) Estrogen replacement therapy and cognitive function in older women. JAMA 260: 2637-2641 Bartus RT, Dean RL, Beer B, Lippa AD (1981) The cholinergic hypothesis of geriatric memory dysfunction. Science 217: 208-217 Birge SJ (1994) The role of estrogen deficiency in the aging of the central nervous system. In: Lobo RA (ed) Treatment of the postmenopausal woman: basic clinical aspects. Raven Press, New York, 153-157 Brenner DE, Kukull WA, Stergachis A, van Belle G, Bowen JD, McCormick WC, Teri L, Larson EB (1994) Postmenopausal estrogen replacement therapy and the risk of Alzheimer’s disease: a population-based case-control study. Am J Epidemiol 140: 262-267 Caldwell BM (1954) An evaluation of psychological effects of sex hormone administration in aged women. J Gerontol 9: 168-174 Caldwell BM, Watson RI (1952) An evaluation of psychologic effects of sex hormone administration in aged women. Results of therapy after 6 months. J Gerontol 7: 228-244

233

Tony Edwin and Uriel Halbreich Campbell S, Whitehead M (1977) Oestrogen therapy and the menopausal syndrome. Clin Obstet Gynecol 1: 31-47 Chakravorthy, Halbreich U (1997) Estrogen and platelet activity in postmenopausal women. Psychopharm Bull 33: 229-233 Chung SK, Pfaff DW, Cohen RS (1988) Estrogen-induced alterations in synaptic morphology in the midbrain central gray. Exp Brain Res 69: 522-530 Costa MM, Reus VI (1999) Estrogen replacement therapy and cognitive decline in memory impaired post-menopausal women. Biol Psychiatry 46: 182-188 Coyle JT, Price DL, DeLong MR (1983) Alzheimer’s disease: a disorder of cortical cholinergic innervation. Science 219: 1184-1190 Derby CA, Hume AL, McPhillips JB, Rarbour MM, Carleton RA (1995) Prior and current health characteristics of postmenopausal estrogen replacement therapy users compared with nonusers. Am J Obstet Gynecol 173: 544-550 Ditkoff EC, Crary WG, Cristo M, Lobo RA (1991) Estrogen improves psychological function in asymptomatic postmenopausal women. Obstet Gynecol 78: 991-995 Faivre-Bauman A, Rosenbaum E, Puymirate J, Grouselle D, Tixier-Vidal A (1981) Differentiation of fetal mouse hypothalamic cells in a serum-free medium. Dev Neurosci 4: 118-129 Fedor-Freybergh P (1977) The Influence of oestrogens on the wellbeing and mental performance in climacteric and postmenopausal women. Acta Obstet Gynecol Scand 3: 64-69 Fillit H, Weinreb H, Cholst I (1986) Observations in a preliminary open trial of estradiol therapy for senile dementia (Alzheimer’s type). Psychoneuroendocrinology 11: 337-345 Hackman R., Galbraith D (1976) Replacement therapy with piperazine oestrone sulfate and its effect on memory. Curr Med Res Opin 4: 303-306 Halbreich U (1995) Estrogen as a potential psychotropic medication. Presented at the 26th Congress of the International Society of Psychoneuroendocrinology, September, Munich, Germany Halbreich U (1995) Menstrually related disorders: What we do know, what we only believe we know, and what we know that we do not know. Crit Rev Neurobiol 2/3: 163-175 Halbreich U (1997) Role of estrogen in postmenopausal depression. Neurology 48 (Suppl 7): 16-20 Halbreich U, Piletz JE, Carson S, Halaris A, and Rojansky N (1993) Increased imidazoline and alpha-2 adrenergic binding in platelets of women with dysphoric premenstrual syndromes. Biol Psychiatry 34: 676-686 Halbreich U, Rojansky N, Palter S, Tworek H, Hissin P, Wang K (1995) Estrogen augments serotonergic activity in postmenopausal women. Biol Psychiatry 37: 434-441 Hampson E, Kimura D (1988) Reciprocal effects of hormonal fluctuations on human motor and perceptual-spatial skills. Behav Neurosci 102: 456-459 Hefti F, Knusel B, Lapchak PA (1993) Protective effects of nerve growth factor and brain derived neurotrophic factor on basal forebrain cholinergic neurons in adult rats with partial fimbrial transections. Progr Brain Res 98: 257-263 Hemminki E, Malin M, Topo P (1993) Selection to postmenopausal therapy by women’s characteristics. J Clin Epiderniol 46: 211-219

234

The Effects of Estrogens on Cognition and Alzheimer’s Dementia Henderson VW (1997) The epiderniology of estrogen replacement therapy and Alzheimers disease. Neurology 48 (Suppl 7): 27-35 Henderson YW, Paganini-Hill A, Emmanuel CK, Dunn ME, Buckwalter JG (1994) Estrogen replacement therapy in older women. Comparisons between Alzheimers disease cases and non-demented control subjects. Arch Neurol 51: 896-900 Hogervorst E, Boshuisen M, Riedel W, Willeken C, Jolles J (1999) The effect of hormone replacement therapy on cognitive function in elderIy women. Psychoneuroendocrinology 24: 43-68 Honjo H, Ogino Y, Naitoh K (1989) In vivo effects by estrone sulfate on the central nervous system: Senile dementia (Alzheimer’s type). J Steroid Biochem Mol Biol 34: 521-525 Jarvik LF (1975) Human intelligence: sex differences. Acta Genet Med Gemellol (Rome) 24: 189-211 Jorm AF, Korten AE, Henderson AS (1987) The prevalence of dementia: a quantitative integration of the literature. Acta Psychiatr Scand 76: 465-479 Kahn L, Halbreich U (1999) Menopause and psychopharmacology: signs, symptoms, and treatment. In: Crosignani PG, Kenemans K, Wenger NK, Jackson AS (eds) Women’s Health and Menopause: Risk Reduction Strategies – Improved Quality of Health. Kluwer Academic Publishers, Boston, 131-136 Kampen D, Sherwin B (1994) Estrogen use and verbal memory in healthy postmenopausal women. Obstet Gynecol 83: 979-83 Kantor HI, Michael CM, Shore H (1973) Estrogen for older women. Am J Obstet Gynaecol 116: 115-118 Kawata M (1995) Roles of steroid hormones and their receptors in structural organization in the central nervous system. Neurosci Res 24: 1-46 Kendall DA, Stancel GM, Enna SJ (1981) Imipramine: effect of ovarian steroids on modifications in serotonin receptor binding. Science 211: 1183-1185 Lerner A, Cole R, Debanne S (1995) Immunological and endocrine conditions in an Alzheimer’s disease case-control study (abstract). Neuroepidemiology 14: 307 McEwen BS, Alves SE, Bulloch K, Weiland NG (1997) Ovarian steroids and the brain: implications for cognition and aging. Neurology 48 (Suppl 7): S8-S15 McEwen BS, Woolley CS (1994) Estradiol and progesterone regulate neuronal structure and synaptic connectivity in adult as well as developing brain. Exp Gerontol 29: 431-436 Mendelsohn ME, Karas RH (1999) The protective effects of estrogen on the cardiovascular system. N Engl J Med 340: 1801-1811 Morrison A, Resnik S, Corrada M, Zonderman A, Kawas C (1996) A prospective study of estrogen replacement therapy and the risk of developing Alzheimer’s disease in the Baltimore longitudinal study of aging (abstract). Neurology 46 (Suppl 2): 435-436 Mortel KF, Meyer JS (1995) Lack of postmenopausal estrogen replacement therapy and risk of dementia. J Neuropsychiatr Clin Neurosci 7: 334-337 Nass R, Raker S (1991) Androgen effects on cognition: congenital adrenal hyperplasia. Psychoendocrinology 16: 189-201 Nee LE, Eldrige R, Sunderland T, Thomas CB, Katz D, Thompson KE, Weingartner H, Weiss H, Julian C, Cohen R (1987) Dementia of the Alzheimer type: clinical and family study of 22 twin pairs. Neurology 37: 359-363

235

Tony Edwin and Uriel Halbreich Ohkura T, Teshima Y, Isse K, et al (1994) Estrogen increases cerebral and cerebellar blood flows in postmenopausal. Menopause 2: 13-18 Okhura T, Isse K, Akazawa K (1995) Long-term estrogen replacement therapy in female patients with dementia of the Alzheimer’s type: 7 case reports. Dementia 6: 99-107 Paganini-Hill A, Henderson VW (1994) Estrogen deficiency and risk of Alzheimer’s disease in women. Am J Epidemiol 140: 256-261 Paganini-Hill A, Henderson YW (1996) Estrogen replacement therapy and risk of Alzheimer’s disease. Arch Intern Med 156: 2213-2217 Pericak-Vance MA, Haines JL (1995) Genetic susceptibility to Alzheimer’s disease. Trends Genet 11: 504-508 Perlman S (1973) Cognitive abilities of children with hormone abnormalities: screening by psychoeducational tests. J Learn Disabil 6: 24-34 Phillips SM, Sherwin ER (1992) Effects of estrogen on memory function in surgically menopausal women. Psychoneuroendocrinology 17: 485-495 Polderman KR, Sthouwer CD, van Kamp GJ, Schalkwijk CG, Gooren LJ (1993) Influence of sex hormones on plasma endothelin levels. Ann Intern Med 118: 429-432 Rarrett-Connor, E and Goodman-Gruen (1999) Cognitive function and endogenous sex hormones in older women. J Am Geriatr Soc 11: 1289-1293 Regland B, Gottfries CG (1992) The role of amyloid beta-protein in Alzheimer’s disease. Lancet 340: 467-469 Resnik S, Berenbaum S, Gottesman T, Bouchard T (1986) Early hormonal influences on cognitive functioning in congenital adrenal hypreplasia. Dev Psychol 22: 191-198 Rirge SJ (1997) The role of estrogen in the treatment of Alzheimer’s disease. Neurology 48 (Suppl 7): 36-41 Rosencrans JA (1970) Differences in brain area 5-hydroxytryptamine turnover and rearing behavior in rats and mice of both sexes. Eur J Pharmacol 9: 379-382 Sar M, Stumpf WE (1981) Central noradrenergic neurones concentrate 3H-oestradiol. Nature 289: 500-502 Schneider LS, Farlow MR, Henderson VW, Pogoda JM (1996) Effects of estrogen replacement therapy on response to tacrine in patients with Alzheimer’s disease. Neurology 46: 1580-1584 Shaywitz SE, Shaywitz BA, Pugh KR et al. (1999) Effect of estrogen on brain activation patterns in postmenopausal women during working memory tasks. JAMA 281: 1197-1202 Sherwin BB (1997) Estrogen effects on cognition in menopausal women. Neurology 48 (Suppl 7): 21-26 Sherwin BB, Phillips S (1990) Estrogen and cognitive functioning in surgically menopausal women. Ann NY Acad Sci 592: 474-475 Sherwin BR, Tulandi T (1996) “Add-back” estrogen reverses cognitive deficits induced by a gonadotropin-releasing hormone agonist in women with leiomyomata uteri. J Clin Endocrinol Metab 81: 2545-2549 Sherwin ER (1994) Estrogenic effects on memory in women. Ann NY Acad Sci 743: 213-231

236

The Effects of Estrogens on Cognition and Alzheimer’s Dementia Shugrue PJ, Dorsa DM (1993) Estrogen modulates the growth-associated protein GAP-43 (neuromodulin) mRNA in the rat preoptic area and basal hypothalamus. Neuroendocrinology 57: 439-447 Singh M, Meyer EM, Millard WJ, Sirnpkins JW (1994) Ovarian steroid deprivation results in a reversible learning impairment and compromised cholinergic function in female Sprague-Dawley rats. Brain Res 644: 305-312 Small GW, Leuchter AF, Mandelkern MA, La Rue A, Okonek A, Lufkin RB, Jarvik LF, Matsuyama SS, Bondareff W (1993) Clinical, neuroimaging, and environmental risk differences in monozygotic female twins appearing discordant for dementia of the Alzheimer type. Arch of Neurol 50: 209-219 Sorabji F, Miranda RC, Toran-Allerand CD (1994) Estrogen differentially regulates estrogen and nerve growth factor receptor mRNAs in adult sensory neurons. J Neurosci 14: 459-471 Tang MX, Jacobs D, Stern Y, Marder K, Schofield P, Gurland B, Andrews H, Mayenx R (1996) Effect of estrogen during menopause on risk and age at onset of Alzheime’s disease. Lancet 348: 429-432 Terasawa E, Timiras P (1968) Electrical activity during the estrous cycle of the rat: cyclic changes in limbic structures. Endocrinology 83: 207-216 Toran-Allerand CD (1984) On the genesis of sexual differentiation of the central nervous system: morphogenic consequences of steroidal exposure and possible role of a fetoprotein. Progr Brain Res 61: 63-98 Toran-Allerand CD (1993) Orgnotypic culture of the developing cerebral cortex and hypothalamus: relevance to sexual differentiation. Psychoneuroendocrinology 57: 439-447 Toran-Allerand CD, Miranda RC, Bentharn WD, Sohrabji F, Brown TJ, Hochberg RB, MacLusky NJ (1992) Estrogen receptors colocalize with low-affinity nerve growth factor receptors in cholinergic neurons of the basal forebrain. Proc Natl Acad Sci USA 89: 4668-4672 Van Buren G, Yang D, Clark KE (1992) Estrogen-induced uterine vasodilation isantogonized by L-nitroarginine methyl ester, an inhibitor of nitric oxide synthesis. Am J Obstet Gynecol 167: 828-833 Van Duijn CM, Hofman A (1992) Risk factors for Alzheimer’s disease: the EURODEM collaborative re-analysis of case-control studies. Neuroepidemiology 11 (Suppl 1): 106-113 Warren SG, Humphreys AG, Juraska JM, Greenough WT (1995) LTP varies across the estrous cycle: enhanced synaptic plasticity in proestrus rats. Brain Res 703: 26-30 Wong M, Moss RL (1992) Long-term and short-term electrophysiological effects of estrogen on the synaptic properties hippocampal CA1 neurons. J Neurosci 12: 3217-3225 Woolley CS, McEwen BS (1993) Roles of estradiol and progesterone in regulation of hippocampal dendritic spine density during the estrous cycle in the rat. J Comp Neurol 57: 935-939 Wu X, Glinn MA, Ostrowski NL, Su Y, Ni B, Cole HW, Bryant HN, Paul SM (1999) Raloxifene and estradiol benzoate both fully restore hippocampal choline acetyltransferase activity in ovarectomized rats. Brain Res 847: 98-104

237

12 Effects of Hormone Therapy on Patterns of Brain Activation during Cognitive Activity: A Review of Neuroimaging Studies Pauline M. Maki and Susan M. Resnick

Epidemiological studies suggest that hormone therapy might lower the incidence of Alzheimer’s disease and delay its onset (Tang et al. 1996; Kawas et al. 1997). Behavioral studies suggest that hormone therapy might also protect against age-related declines in memory and other cognitive abilities in individuals who are free of dementia (Resnick et al. 1997; Kampen and Sherwin 1994; Robinson et al. 1994), in part by enhancing encoding while learning new information (Maki et al. 2001). Biological support for these findings comes from animal studies showing beneficial effects of estrogen on neuronal survival and connectivity in regions of the brain subserving memory, in particular, the hippocampus (McEwen et al. 1997). Evidence that estrogen influences brain function in humans comes from neuroimaging studies. In this chapter, we review studies of the effects of hormone therapy on cerebral blood flow, cerebral glucose metabolism, and patterns of brain activity during performance of cognitive tasks. A greater understanding of the effects of hormone therapy on brain function is valuable, because neurophysiological studies offer unique insights into the biological plausibility of the epidemiological and behavioral findings. Studies of the effects of hormone therapy on resting cerebral blood flow and metabolism preceded neuroimaging studies using cognitive challenge paradigms. In the mid-1980s, Namba and Sokoloff (1984) demonstrated that acute administration of high doses of estrogen resulted in significant increases in glucose metabolism in ovariectomized rats. Subsequent studies of cerebral blood flow during the resting state in middle-aged postmenopausal women demonstrated estrogen-related increases at the level of the convexity corresponding to the cerebral vasculature (Greene 2000), in cerebellar and whole brain blood flow (Ohkura et al. 1995), and more specifically across

239

Pauline M. Maki and Susan M. Resnick

right lower frontal regions, bilateral middle and upper frontal regions, bilateral medial and upper temporal regions, bilateral caudate and thalamus, bilateral parietal regions, bilateral lower occipital regions, left upper occipital regions, and bilateral precentral regions (Ohkura et al. 1996). Estrogen-related increases in global blood flow have been reported in older women with a history of cerebrovascular disease (Funk et al. 1991), and in relative blood flow in healthy older women in the right middle/superior temporal gyrus, the right inferior temporal gyrus, and the left middle temporal gyrus (Maki and Resnick 2000). Additional evidence of hormone effects on cerebral activity comes from demonstrations of changes in regional glucose metabolism in female rats across the estrous cycle (Nehlig et al. 1985) and in premenopausal women across the menstrual cycle (Reimann et al. 1996). These studies suggest that hormone therapy increases resting cerebral blood flow and glucose metabolism. Recent advances in functional neuroimaging studies allow for direct assessments of the effects of hormone therapy on regional patterns of brain activity during cognitive test performance. Two primary neuroimaging techniques have been used. One, positron emission tomography (PET), involves a radiolabeled tracer or radioisotope. One of the more commonly used radioisotopes is H215O, or radiolabeled water, which is used to measure cerebral blood flow. H215O has a short radioactive half-life (2 min.), enabling researchers to perform successive neuroimaging scans under different cognitive instructions. Another commonly used radioisotope is 18-F-Fluorodeoxyglucose (18-F-FDG) radiolabeled glucose, which is used to measure glucose metabolism. 18-F-FDG is used less commonly in cognitive studies than H215O, because 18-F-FDG has a longer half-life (110 min) and is more difficult to use for sequential assessments of different cognitive states. In cognitive studies, the radioisotope is typically injected into an arm vein very shortly after the participant begins to perform a particular cognitive task. The isotope then distributes throughout the brain depending upon where blood flow or energy metabolism is greatest and emits positrons as it decays. Under typical conditions, regional cerebral blood flow and metabolism are thought to be closely coupled and reflect the underlying neural activity. The second neuroimaging approach, functional magnetic resonance imaging (fMRI), uses MRI technology to measure functional processes in the brain. This technique measures particular magnetic

240

Effects of Hormone Therapy on Patterns of Brain Activation

signal changes called blood oxygen level dependent (BOLD) signal changes that occur with neural activity. BOLD changes reflect two factors – first, that there are local changes in the amount of oxygen in active neural tissue, and second, that deoxygenated hemoglobin is magnetic whereas oxygenated hemoglobin is not (Cohen 1996). To detect the change in the magnetic properties of brain regions, the fMRI technique involves serial imaging of brain tissue during active and rest states. A brain region with more oxygenated blood will show up more intense on certain MRI images compared to when there is less oxygenated blood. fMRI has better spatial and temporal resolution than PET, but may have limited resolution in inferior temporal and frontal areas due to airway (i.e., sinus) artifacts. To date, of the functional neuroimaging studies on hormone therapy and brain activation, one used fMRI (Shaywitz 1999), three used PET and H215O (Maki and Resnick 2000; Berman et al. 1997; Resnick et al. 1998), and one used PET and 18-F-FDG (Eberling et al. 2000). Both PET and fMRI techniques have been sensitive to the effects of hormone therapy on patterns of brain activation during performance of cognitive tests. Berman and colleagues (1997) were the first to investigate the effects of ovarian steroid hormones on regional cerebral blood flow during performance of a cognitive task. In an add-back design, they used PET and H215O in 11 young premenopausal women as they performed a modified version of a common test of executive functioning and a control test during ovarian hormone suppression with Lupron (i.e., lowered estrogen and progesterone). In a randomized manner, they then repeated the tests during treatment with Lupron plus transdermal estradiol to isolate the effects of estradiol alone, and during treatment with Lupron plus progesterone to isolate the effects of progesterone alone. With Lupron, the young women showed suppression of the activation pattern typically obtained during the test, which is activation of the prefrontal cortex, inferior parietal lobule, and other cortical regions. With add-back estrogen and add-back progesterone, they showed the normal pattern of activation and showed increased activity in parietal and temporal regions. There was greater hippocampal activation in the estrogen add-back condition than in the progesterone add-back condition. This study provides evidence that estrogen and progesterone influence patterns of brain activation during performance of an executive task in premenopausal women. The add-back design in

241

Pauline M. Maki and Susan M. Resnick

premenopausal women has been proposed as a model for estrogen replacement and progesterone replacement in postmenopausal women, although potential differences between acute and chronic hormone depletions warrant caution. Five studies of the effects of hormone therapy on regional brain activation patterns in postmenopausal women followed Berman’s seminal study in premenopausal women. Three of those were observational studies (also called naturalistic studies) (Maki and Resnick 2000; Resnick et al. 1998; Eberling et al. 2000), which compared women who chose to receive hormone therapy to those who did not choose to receive such therapy. Observational studies of hormone therapy can be limited by the healthy user bias, the tendency for women who choose to receive hormone therapy to be younger, healthier, and better educated than women who do not choose to receive hormone therapy (Matthews et al. 1996). One approach to minimizing these confounds is to match the study groups for age and education (Maki and Resnick 2000; Resnick et al. 1998). Another approach is to covary for group differences in confounding factors (Eberling et al. 2000); however, the validity of such techniques, particularly in small samples, is questionable. In the first of the observational studies, we conducted a crosssectional investigation in postmenopausal women participating in the Longitudinal Neuroimaging Study of the Baltimore Longitudinal Study of Aging (BLSA) (Resnick et al. 2000). Our goal was to build on our findings that postmenopausal BLSA participants receiving hormone therapy performed better on tests of verbal (Maki et al. 2001) and figural (Resnick et al. 1997) memory than did participants who were not receiving hormones. To this end, we used PET and H215O to measure patterns of brain activity during performance of delayed verbal and figural recognition memory tasks (Golski et al. 1998) and also during rest in two groups of older women differing in hormone use. The study involved 15 women who were receiving hormone therapy (typically, Premarin 0.625 mg/day) and 17 who were receiving no therapy. The two groups were matched in age (mean age 66 years), education, and general verbal knowledge (Resnick et al. 1998). Women receiving hormone therapy and those receiving no hormone therapy showed different patterns of brain activation during the tasks (Fig. 1). Comparisons between the verbal memory and resting conditions revealed group differences in the right parahippocampal gyrus,

242

Effects of Hormone Therapy on Patterns of Brain Activation

Fig. 1. Hormone therapy is associated with changes in activity in the right parahippocampal gyrus and inferior frontal regions during performance of a delayed verbal memory task (Source: Resnick et al. 1998)

precuneus, inferior frontal cortex, and dorsal frontal gyrus. Comparisons between the figural memory and resting conditions revealed group differences in the right inferior parietal region, right parahippocampal gyrus, left visual association area, left anterior thalamus, and a region proximal to the right mammilary body. These group differences do not indicate that one group shows consistently greater activation of specific brain regions, but rather they reveal greater relative activation by one or the other group for some regions, whereas in other regions the pattern is in the opposite direction. We focused on regional patterns rather than directions of differences, because greater regional cerebral blood flow response to a task has been associated with both better performance (Nyberg et al. 1996) and less efficient processing (Grady et al. 1995). Although the group receiving hormone therapy did not show a clear advantage on the activation task, they showed superior memory on standardized neuropsychological tests of verbal and figural memory, suggesting that the group differences in regional cerebral blood flow during performance of memory tasks were indicative of a beneficial effect of estrogen on brain function. This study points to the parahippocampal gyrus, frontal cortex, and inferior parietal lobule as regions that may modulate the beneficial effects of hormone therapy on verbal and figural memory. In a follow-up study, we examined longitudinal changes in regional cerebral blood flow to determine whether women who received hormone therapy and those who received no such therapy showed different patterns of brain aging over time (Maki and Resnick

243

Pauline M. Maki and Susan M. Resnick

2000). We compared patterns of change over a 2-year interval in 28 of the women who participated in the previous cross-sectional analysis and maintained the same group assignment during that interval. In this study, we focused primarily on the change over time across the three conditions combined (i.e., rest and figural and verbal memory) to maximize statistical power. Results indicated significant differences in the patterns of blood flow changes over time, with most changes indicating areas of increased blood flow in women receiving hormone therapy. Notably, the largest of these differences were in the right hippocampus, entorhinal cortex, middle temporal gyrus, and parahippocampal gyrus (Fig. 2). These are the same regions that show hypoperfusion in individuals at increased risk for Alzheimer’s disease (Kennedy et al. 1995; Reiman et al. 1996; Johnson et al. 1998). This demonstration that estrogen modulates hippocampal functioning in humans lends support to the biological plausibility that estrogen may protect against age-associated declines in memory and may lower the risk of Alzheimer’s disease.

Sagittal

Coronal

Axial 6

Zvalue

4 3 2 1 0

R

L

Axial Slice superimposed on structural MRI

Fig. 2. Women who receive hormone replacement therapy show greater longitudinal increases in blood flow in the hippocampus over a 2-year interval compared with women who receive no hormone replacement (Source: Maki and Resnick 2000)

244

Effects of Hormone Therapy on Patterns of Brain Activation

A recent observational study used PET and 18-F-FDG to measure regional glucose metabolism during performance of a continuous recognition memory task for words (Eberling et al. 2000). Compared with women with Alzheimer’s disease (n = 13), nondemented, postmenopausal women receiving hormone therapy (n = 8) showed greater relative metabolism, adjusted for whole brain values, in dorsolateral frontal, middle temporal, and inferior parietal cortex. In contrast, nondemented women receiving no hormone therapy (n = 5) did not differ from women with Alzheimer’s disease in glucose metabolism. The difference in metabolism between nondemented women receiving hormone therapy and those not receiving hormone therapy was evident only indirectly, in reference to patients with Alzheimer’s disease. The failure to find hormone-related differences in glucose metabolism among the neurologically normal women may have been due to insufficient statistical power. To date, there has been one published randomized, placebo-controlled intervention trial of the effects of hormone therapy on brain activation patterns (Shaywitz et al. 1999). In intervention studies, women are treated with hormone therapy for the purpose of the study. Intervention studies have a notable advantage over observational trials – they are not susceptible to the healthy user bias, because treatment is decided by investigators or by random assignment. The use of a placebo control group in intervention studies allows for estimations of placebo effects and of the effects of repeated assessments on outcomes. Shaywitz and colleagues conducted a double-blind, placebo-controlled, crossover intervention study of short-term estrogen therapy (i.e., 1.25 mg conjugated equine estrogen/day for 21 days) on activation patterns during figural and verbal working memory tests in 46 middle-aged, postmenopausal women (mean age 51 years). Working memory involves the temporary maintenance of information in memory. To examine the effects of hormones on different aspects of working memory, functional MRI imaging scans were acquired under three conditions – encoding (i.e., study), storage (i.e., a 20-s delay), and retrieval (i.e., forced choice recognition) – for each of the two memory tasks (i.e., figural and verbal). There was no change in performance on the tests with treatment, but there were changes in the patterns of brain activation during task performance. Across the two tests combined, estrogen therapy was associated with an exaggeration of the typical pattern observed during encoding and retrieval, namely, increased

245

Pauline M. Maki and Susan M. Resnick

activation in the left superior frontal gyrus during encoding and right superior frontal gyrus during retrieval. During storage of verbal material there were estrogen-related increases in anterior frontal regions and the inferior parietal lobule bilaterally, and decreases in the inferior parietal lobule (at a level lower than the area of increased inferior parietal activation), left central sulcus, and right superior temporal gyrus. During storage of nonverbal material, there was decreased activation in the inferior parietal lobule at the same level at which activations increased during storage of verbal material. This study demonstrates that hormones modulate brain activity during working memory tasks in middle-aged, postmenopausal women. In summary, observational, intervention, and add-back studies indicate that hormone therapy modulates brain activity during performance of a variety of cognitive tasks (Maki and Resnick 2000; Shaywitz et al. 1999; Berman et al. 1997; Resnick et al. 1998). Studies of hormone effects on brain activity during memory tests are of particular interest, given that hormone therapy appears to exert the most consistent beneficial effects on memory, particularly verbal memory. Although not all studies have shown a beneficial effect of hormone therapy on verbal memory (Barrett-Connor and Kritz-Silverstein 1993; Szklo et al. 1996), we argue that the failure to find such effects reflects the particular type of memory tests selected for use in those studies. In a recent study, women receiving hormone therapy showed enhanced encoding of words (Maki et al. 2001), suggesting that the beneficial effects of hormone therapy on verbal recall may in part reflect enhanced encoding. In studies involving memory tests that maximize the encoding of words by directing rehearsal of unrecalled words (Barrett-Connor and Kritz-Silverstein 1993; Szklo et al. 1996), no beneficial effects of hormone therapy are seen. In contrast, where studies include memory tests involving no such rehearsal, a beneficial effect on memory is observed (Maki et al. 2001; Kampen and Sherwin 1994; Robinson et al. 1994; Resnick et al. 1998). If estrogen influences memory in part by enhancing encoding, then rehearsing items until they are well learned may obscure the beneficial effects of estrogen on memory. The fact that verbal memory has shown the most consistent improvement with hormone therapy (Maki et al. 2001; Kampen and Sherwin 1994; Robinson et al. 1994; Sherwin 1988; Sherwin and Phillips 1990; Phillips and Sherwin 1992) has led to the view that the

246

Effects of Hormone Therapy on Patterns of Brain Activation

beneficial effects of estrogen may be limited to verbal memory (Sherwin 1997). However, the beneficial effects of hormone therapy on figural memory are evident in some behavioral studies (Resnick et al. 1997; Duka et al. 2000) and supported by three neuroimaging studies (Maki and Resnick 2000; Shaywitz et al. 1999; Resnick et al. 1998). Recent observational data from our laboratory showed beneficial effects of hormone therapy on both verbal and figural memory performance, with only those for verbal memory reaching statistical significance in a moderate sized sample (Maki et al. 2001). Several lines of evidence demonstrate significant effects of estrogen on the hippocampus, an area critical for encoding and retrieval of both words and figures (Schacter and Wagner 1999; Schacter et al. 1999). Animal studies show beneficial effects of estrogen on dendritic spine density (Gould et al. 1990) and synaptic excitability (Wong and Moss 1992) in hippocampal neurons. Evidence of increased hippocampal blood flow over time in women receiving hormone therapy (Maki and Resnick 2000) provides support for hormone-related enhancements in hippocampal functioning in humans. Neuroimaging studies point to hormone-related changes in a number of extra-hippocampal brain areas that subserve distinct memory processes. A review of the evidence of hormone-related effects on encoding, storage, and retrieval may help to highlight the particular brain regions that are sensitive to hormone therapy. Greater left hemisphere activation during encoding and greater right hemisphere activation during retrieval is the typical pattern observed in imaging studies of memory processes (Tulving et al. 1994). Estrogen therapy influences activation in left frontal regions during encoding (Shaywitz et al. 1999) and right frontal areas during retrieval (Shaywitz et al. 1999; Resnick et al. 1998). During storage of verbal material, estrogen therapy is associated with increased activity in the anterior frontal and inferior parietal lobule (Shaywitz et al. 1999). Hormone therapy is associated with decreased activation in the inferior parietal lobule during storage of nonverbal material (Shaywitz et al. 1999) and during recognition of figures (Resnick et al. 1998). PET studies revealed hormone-associated increases in activity in the hippocampus and parahippocampal cortex in association with performance on delayed memory tasks (Maki and Resnick 2000; Resnick et al. 1998) and executive tasks (Berman et al. 1997). Due to airway artifacts that cause distortions in imaging some regions, some fMRI techniques may have limited sensitivity for detection of hor-

247

Pauline M. Maki and Susan M. Resnick

mone effects in particular inferior frontal and anterior temporal lobe regions. Hormone therapy influences patterns of brain activity on cognitive tests that do not typically show beneficial effects in behavioral outcomes, including tests of executive function (Berman et al. 1997) and working memory (Shaywitz et al. 1999). This suggests that neuroimaging techniques may be more sensitive than behavioral techniques in detecting beneficial effects of hormones on brain function associated with a number of cognitive activities. The argument that any estrogen-related effect on neural activity is reflective of a meaningful change in cognition is bolstered by concurrent demonstrations of beneficial effects on standardized neuropsychological tests. In summary, results of neurophysiological studies add to the biological plausibility of epidemiological and behavioral findings of a protective effect of hormone therapy on age-related changes in cognition and Alzheimer’s disease. Studies suggest that hormone therapy affects neural substrates of several cognitive functions in young, middle-aged, and older women. Randomized, double-blind, placebo-controlled studies of hormone effects on neuroimaging outcomes during cognitive tasks are underway to determine whether these effects can be generalized to older women in a randomized trial. Although we are hopeful that observational findings (Crawford 1996; Simpkins et al. 1997) will replicate in randomized trials, we must be cautious in attributing differences in brain functioning among women receiving hormone therapy to the physiological properties of the hormones. Findings from ongoing neuroimaging studies and large-scale, randomized trials of hormone therapy on cognitive outcomes will provide data necessary for determining whether hormone therapy is a useful preventative treatment for minimizing age-related declines in brain function in postmenopausal women.

References Barrett-Connor E, Kritz-Silverstein D (1993) Estrogen replacement therapy and cognitive function in older women. JAMA 269: 2637-2641 Berman KF, Schmidt PJ, Rubinow DR, Danaceau MA, van Horn JD, Esposito G, Ostrem JL, Weinberger DR (1997) Modulation of cognition-specific cortical activity by gonadal steroids: a positron-emission tomography study in women. Proc Natl Acad Sci 94: 8836-8841

248

Effects of Hormone Therapy on Patterns of Brain Activation Cohen MS (1996) Methods. In: Toga AW, Mazziotta JC (eds) Brain mapping. Academic Press, San Diego, 223-255 Crawford JG (1996) Alzheimer’s disease risk factors as related to cerebral blood flow. Med Hypotheses 46: 367-377 Duka T, Tasker R, McGowan JF (2000) The effects of 3-week estrogen hormone replacement on cognition in elderly healthy females. Psychopharmacology 149: 129-139 Eberling JL, Reed BR, Coleman JE, Jagust WJ (2000) Effect of estrogen on cerebral glucose metabolism in postmenopausal women. Neurology 55: 875-877 Funk J, Mortel K, Meyer J (1991) Effects of estrogen replacement therapy on cerebral perfusion and cognition among postmenopausal women. Dementia 2: 268-272 Golski S, Resnick SM, Malamut BL, Zonderman AB (1998) Verbal and figural recognition memory in relation to age and neuropsychological measures. Exp Aging Res 4: 359-385 Gould E, Woolley CS, Frankfurt M, McEwen BS (1990) Gonadal steroids regulate dendritic spine density in hippocampal pyramidal cells in adulthood. J Neurosci 10: 1286-1291 Grady CL, McIntosh AR, Horwitz B (1995) Age-related reductions in human recognition memory due to impaired encoding. Science 269: 218-221 Greene RA (2000) Estrogen and cerebral blood flow: a mechanism to explain the impact of estrogen on the incidence and treatment of Alzheimer’s disease. Int J Fertil Menopausal Stud 45: 253-257 Johnson KA, Jones K, Holman BL, Becher JA, Spiers PA, Satlin A, Albert MS (1998) Preclinical prediction of Alzheimer’s disease using SPECT. Neurology 50: 1563-1571 Kampen DL, Sherwin BB (1994) Estrogen use and verbal memory in healthy postmenopausal women. Obstet Gynecol 83: 979-983 Kawas C, Resnick S, Morrison A et al. (1997) A prospective study of estrogen replacement therapy and the risk of developing Alzheimer’s disease: the Baltimore Longitudinal Study of Aging. Neurology 48: 1517-1521 Kennedy AM, Frackowiak RS, Newman SK et al. (1995) Deficits in cerebral glucose metabolism demonstrated by positron emission tomography in individuals at risk of familial Alzheimer’s disease. Neurosci Lett 186: 17-20 Maki PM, Resnick SM (2000) Longitudinal effects of estrogen replacement therapy on PET cerebral blood flow and cognition. Neurobiol Aging 21: 373-383 Maki PM, Zonderman AB, Resnick SM (2001) Enhanced verbal memory in nondemented elderly estrogen users. Am J Psychiatry 158: 227-233 Matthews KA, Kuller LH, Wing RR, Meilahn EN, Plantinga P (1996) Prior to use of estrogen replacement therapy, are users healthier than nonusers? Am J Epidemiol 143: 971-978 McEwen BS, Alves SE, Bulloch K, Weiland NG (1997) Ovarian steroids and the brain: implications for cognition and aging. Neurology 48: S8-S15 Namba H, Sokoloff L (1984) Acute administration of high doses of estrogen increases glucose utilization throughout brain. Brain Res 291: 391-394 Nehlig A, Porrino LJ, Crane AM, Sokoloff L (1985) Local cerebral glucose utilization in normal female rats: variations during the estrous cycle and comparison with males. J Cereb Blood Flow Metab 5: 393-400

249

Pauline M. Maki and Susan M. Resnick Nyberg L, Cabeza R, Tulving E (1996) PET studies of encoding and retrieval: the HERA model. Psychon Bull Rev 3: 135-148 Ohkura T et al. (1996) Effect of estrogen on regional cerebral blood flow in postmenopausal women. J Japan Menopause Soc 4: 254-261 Ohkura T et al. (1995) Estrogen increases cerebral and cerebellar blood flows in postmenopausal women. Menopause 2: 13-18 Phillips SM, Sherwin BB (1992) Effects of estrogen on memory function in surgically menopausal women. Psychoneuroendocrinology 17: 485-495 Reiman EM et al. (1996) Preclinical evidence of Alzheimer’s disease in persons homozygous for the epsilon 4 allele for apolipoprotein E. N Engl J Med 334: 752-758 Reiman EM, Armstrong SM, Matt KS, Mattox JH (1996) The application of positron emission tomography to the study of the normal menstrual cycle. Hum Reprod 11: 2799-2805 Resnick SM et al. (2000) One-year age changes in MRI brain volumes in older adults. Cereb Cortex 10: 464-472 Resnick SM, Maki PM, Golski S, Kraut MA, Zonderman AB (1998) Estrogen effects on PET cerebral blood flow and neuropsychological performance. Hormones Behav 34: 171-184 Resnick SM, Metter EJ, Zonderman AB (1997) Estrogen replacement therapy and longitudinal decline in visual memory: a possible protective effect? Neurology 49: 1491-1497 Robinson D, Friedman L, Marcus R, Tinklenberg J, Yesavage J (1994) Estrogen replacement therapy and memory in older women. J Am Geriatr Soc 42: 919-922 Schacter DL, Wagner AD (1999) Medial temporal lobe activations in fMRI and PET studies of episodic encoding and retrieval. Hippocampus 9: 7-24 Schacter DL et al. (1999) Medial temporal lobe activation during episodic encoding and retrieval: a PET study. Hippocampus 9: 575-581 Shaywitz SE et al. (1999) Effect of estrogen on brain activation patterns in postmenopausal women during working memory tasks. JAMA 281: 1197-1202 Sherwin BB, Phillips S (1990) Estrogen and cognitive functioning in surgically menopausal women. Ann N Y Acad Sci 592: 474-475 Sherwin BB (1988) Estrogen and/or androgen replacement therapy and cognitive functioning in surgically menopausal women. Psychoneuroendocrinology 13: 345-357 Sherwin BB (1997) Estrogen effects on cognition in menopausal women. Neurology 48: S21-S26 Simpkins JW et al. (1997) Estrogens may reduce mortality and ischemic damage caused by middle cerebral artery occlusion in the female rat. J Neurosurg 87: 724-30 Szklo M et al. (1996) Estrogen replacement therapy and cognitive functioning in the Atherosclerosis Risk in Communities (ARIC) study. Am J Epidemiol 144: 1048-1057 Tang MX, Jacobs D, Stern Y et al. (1996) Effect of oestrogen during menopause on risk and age at onset of Alzheimer’s disease. Lancet 348: 429-432 Tulving E, Kapur S, Craik FI, Moscovitch M, Houle S (1994) Hemispheric encoding/ retrieval asymmetry in episodic memory: positron emission tomography findings. Proc Natl Acad Sci 91: 2016-2020

250

Effects of Hormone Therapy on Patterns of Brain Activation Wong M, Moss R (1992) Long-term and short-term electrophysiological effects of estrogen on the synaptic properties of hippocampal CA1 neurons. J Neurosci 12: 3217-3225

251

13 Estrogens in Alzheimer’s Disease: A Clinical and Neurobiological Perspective Peter Schönknecht, Johannes Pantel, Aoife Hunt, Marcus Henze, Thomas Strowitzki and Johannes Schröder

Introduction: Cerebral and Molecular Changes in Alzheimer’s Disease Dementia is one of the most common diseases of the elderly with prevalence rates of 5% at age 74 and more than 10% among those older than 80 years. The majority (50%–60%) of dementia cases can be characterized as Alzheimer’s disease (AD). AD is a progressive neurodegenerative disorder with gradual onset and deterioration of cognitive functions such as memory, language, and visuospatial skills. The clinical diagnosis of AD necessitates exclusion of other forms of dementia such as vascular dementia or dementia due to cerebral neoplasm, inflammation, or other pathologies. Histopathologically, AD is associated with neurofibrillary tangle formation and deposition of amyloid plaques in susceptible brain regions. Tau protein, a microtubule-associated protein, seems to be released into the cerebrospinal fluid (CSF) during neurofibrillary tangle formation and has been found to be increased in patients with manifest AD (Schönknecht et al. 2002a) as well as in mild cognitive impairment or incipient AD (Buerger et al. 2002; Schönknecht et al. 2002b). A major feature of AD is the presence of amyloid plaques in the brain. Amyloid plaques are extracellular deposits of fibrillar aggregates mostly composed of a 4-kDa peptide, β-amyloid 1-42 (Aβ42), which is derived along with the peptide β-amyloid 1-40 (Aβ40) from the larger amyloid precursor protein (APP) by proteolytic cleavage (Haass et al. 1993; Jensen et al. 1999). Several studies have indicated that the deposition of Aβ42 constitutes an important process in the etiology of neuronal degeneration (for review, see

253

Peter Schönknecht et al.

Beyreuther et al. 1996). Significantly, CSF Aβ42 concentrations have been shown to be a sensitive marker of AD pathology (for review, see Ida et al. 1996). Neuroimaging studies demonstrate that cerebral changes characteristic of AD primarily strike the temporal lobe with a particular focus on medial temporal substructures. These cerebral changes can be reliably assessed using quantitative magnetic resonance imaging (MRI) (Pantel et al. 1997a, b). Therefore, one might assume CSF Aβ42 levels to be associated with measures of temporal lobe rather than global cerebral atrophy. Among AD patients, CSF Aβ42 levels were found to be significantly correlated with the volume of the temporal lobes but not with other volumetric measures (Schröder et al. 1997). Similar findings obtained in a larger patient sample indicate that changes in cerebral Aβ42 levels are strongly associated with temporal lobe but not general brain atrophy and thus emphasize the significance of Aβ in the etiology of AD (Schröder and Pantel 1999). Several studies indicate that both genetically determined disturbances in APP metabolism with consecutive overexpression of the protein and a pathological enzymatic processing of APP to Aβ are essentially involved in amyloid plaque production. The Aβ release, however, has been supposed to be modified by several cofactors such as metals (copper, zinc), apolipoprotein E (apo E), cholesterol, and estrogens.

Estrogen Effects on Central Nervous System and β-Amyloid Metabolism Estrogens can readily pass the blood-brain barrier and have several effects on the central nervous system (CNS) (for review, see Henderson 1997a). They enhance the outgrowth of neurites and promote the formation of dendrite spines and synapses (Woolley et al 1994; Lustig 1994). In the brain, estrogens increase cerebral blood flow and glucose metabolism and modulate acetylcholine metabolism (Ohkura et al. 1995; Bishop and Simpkins 1992; Luine 1985). Furthermore, the beneficial effect of estrogens on the CNS could be mediated by cholesterol and apo E metabolism, since estrogens have been found to decrease serum cholesterol (Bhatena et al. 1998) and apo E levels (Urabe et al. 1996) in postmenopausal women. In addition, estrogens have been characterized as having an im-

254

Estrogens in Alzheimer’s Disease: A Clinical and Neurobiological Perspective

portant impact on Aβ formation in the CNS, a hallmark of AD. Experiments in cell cultures indicated that physiological concentrations of 17β-estradiol increase production of soluble APP at the expense of Aβ40 and Aβ42 production. Jaffe et al. (1994) demonstrated in vitro that physiological concentrations of 17β-estradiol can modulate the APP metabolism by increasing the cellular release of the soluble, non-amyloidogenic components. Xu et al. (1998) investigated the effect of physiological concentrations of 17β-estradiol on the APP metabolism of neurites and found not only an increase of soluble APP components, but also significantly reduced Aβ42 and Aβ40 concentrations. Moreover, animal studies indicate that prolonged ovariectomy results in uterine atrophy and decreased serum 17β-estradiol levels, and is associated with increased cerebral Aβ levels (Petanceska et al. 2000). In the same study, 17β-estradiol treatment significantly reversed the ovariectomy-induced increase in brain Ab levels. The effect of estrogens on Aβ metabolism may be mediated by a morphological modification of the intracellular trans-Golgi compartments where maturated APP is processed to Aβ. Furthermore, estrogens support neuronal APP transport and act against the apparent dysregulation which leads to increased Aβ production (Xu et al. 1998). The protective effects of estrogens may also be mediated by interactions not directly related to APP processing such as modulation of neural functioning or prevention of oxidative toxicity due to glutamate, free radicals, and Aβ40 or Aβ42 (Mooradian et al. 1993). Behl et al. (1997) showed that 17β-estradiol can prevent intracellular peroxide accumulation and, ultimately, the degeneration of hippocampal neurons. To date, little is known about potential differential effects of the various estrogens on the CNS. The majority of studies focused on 17β-estradiol, which is generally considered to be the most potent estrogen. While some studies also addressed estrone, other estrogens such as estriol or progestagens such as progesterone were not addressed.

Estrogen Replacement Therapy and AD Most evidence for a beneficial effect of estrogen replacement therapy (ERT) on AD is derived from epidemiological studies demonstrating an approximately 50% lower incidence of AD in women receiving

255

Peter Schönknecht et al.

ERT (Henderson 1997b; Lamberts et al. 1997). This finding was recently confirmed in prospective studies (Tang et al. 1996; Kawas et al. 1997) and led to the hypothesis of a therapeutic effect of estrogens on the course and severity of AD in postmenopausal women. Retrospective or uncontrolled studies investigating ERT for women with AD support a facilitative effect of the hormone on memory (Henderson et al. 1997b). While these studies had methodological limitations, such as open-label design, small sample size, or short treatment duration, beneficial effects on memory and attention were also found in prospective, placebo-controlled, doubleblind studies in women with AD (Honjo et al. 1993; Fillit et al. 1994). Recently, randomized, double-blind, placebo-controlled, parallelgroup trials have been undertaken to define a therapeutic role for ERT in AD patients (Mulnard et al. 2000; Henderson et al. 2000; Wang et al. 2000; Asthana et al. 2001). In the majority of these studies (Mulnard et al. 2000; Henderson et al. 2000; Wang et al. 2000), no differences in cognitive function between estrogen- and placebo-treated groups could be found. These studies used conjugated estrogen, either 0.625 or 1.25 mg/day, for a duration of treatment between 12 and 52 weeks. Sample sizes ranged from 42 to 120 AD patients and control subjects. Hence, the major methodological limitations involved a small patient population (Henderson et al. 2000) and short treatment course (Wang et al. 2000). Asthana et al. (1999 and 2001) indicated that conjugated equine estrogen given orally may not be as effective in the CNS as pure 17β-estradiol, the most potent endogenous human estrogen. Using a transdermal application of 17β-estradiol at doses of 0.05 and 0.10 mg/day for 8 weeks, Asthana et al. (2001) demonstrated a potential therapeutic role for this estrogen in a double-blind, placebo-controlled study of 20 postmenopausal AD patients. Significant effects of 17β-estradiol estrogen treatment were observed on attention, verbal memory, and visual memory. However, evaluation of the importance of different estrogen compounds is rather difficult, since oral estradiol treatment leads to elevated estradiol levels in serum and supraphysiological levels of estrone as well. Taken together, data from studies investigating the potential role of estrogen as a beneficial agent for the course and progression of AD in women appear rather conflicting. Since basic studies of the mechanisms of neurodegenerative disorders suggest at least two phases in their pathogenesis – namely, initiation and propagation – it was

256

Estrogens in Alzheimer’s Disease: A Clinical and Neurobiological Perspective

hypothesized that estrogens may interact differentially in each phase (Johnson et al. 1998). Dosage form, dose required, and specific type of estrogen may influence ERT effects on initiation and propagation of AD. Nevertheless, the impact of age, early onset and severity of dementia, and longer-term treatment on the potential beneficial effects of ERT still has to be addressed.

Endogenous Estrogen Levels in Female AD Patients Since a potential beneficial effect of ERT on the development and course of AD has been suggested in the studies reviewed above, several investigators compared endogenous serum estrogen levels in demented and nondemented individuals (Table 1). At this point, an important methodological problem must be addressed: in a considerable proportion of women, postmenopausal estrogen levels are below the detection limits of the assays available. Data from these probands would be excluded in a simple comparison of estrogen means across patients and controls. This methodological problem applies particularly to larger studies comprising a considerable

Table 1. Serum estrogen levels in demented patients (Alzheimer’s disease; AD) and nondemented control subjects Reference

AD patients (g1)

Nondemented control subjects (g2)

Serum estrogen

Results

Honjo et al. 1989

n=7

n=7

Estronesulfate

g1 < g2 (p < 0.05)

Manly et al. 2000

n = 50

n = 93

Estrone Estradiol

g1 < g2* g1 < g2 (p < 0.05)

Cunningham et al. 2001

n = 52

n = 60

Estrone Estradiol

g1 > g2 (p < 0.05) g1 < g2*

Senanarong et al. 2002

n = 72 Demented (AD n = 37)

n = 63

Estradiol

g1 < g2*

Hogervorst et al. 2002

n = 66

n = 62

Estrone Estradiol

g1 > g2* g1>g2 (p < 0.05)

* Not significant.

257

Peter Schönknecht et al.

proportion of women with subthreshold levels. One way to overcome this difficulty is to compare frequency differences among quartiles of hormone levels between patients and controls for significant differences of distribution (Manly et al. 2000). A pilot study conducted by Honjo et al. (1989) described lower estrone-sulfate but not estradiol levels in seven female AD patients when compared with seven control subjects. However, in a study of 50 patients compared with 93 nondemented controls, Manly et al. (2000) found postmenopausal AD patients to have lower serum estradiol levels. Although this difference failed to reach statistical significance, AD patients were much more likely to have estradiol levels lower than 20 pg/ml than estradiol levels higher than 20 pg/ml. Concerning estrone, no significant group differences arose. In a study by Senanarong et al. (2002), elderly individuals with lower estradiol levels were found to have more impaired cognition as revealed by the TMSE, a modified version of the mini-mental status examination (MMSE; Folstein et al. 1975). In this study, the 72 patients with dementia (including 36 AD patients) were characterized by lower estradiol levels than the 63 nondemented control subjects. These two studies were contrasted by the recent findings of Hogervorst et al. (2002), who reported AD patients to have significantly higher serum 17β-estradiol levels than controls. In this study, lower cognitive performance measured on the MMSE was associated with a high ratio of 17β-estradiol to total estrogen, and lower serum folate concentration in AD patients. In a study of 52 postmenopausal women with AD and 60 nondepressed cognitively healthy controls, significantly higher serum levels for estrone but not estradiol were found in the AD group compared to the controls (Cunningham et al. 2001). Since another study (Yaffe et al. 1998) reported an association between higher serum estrone levels and lower scores on the digit symbol and Trail-Making B test in 532 women older than 65 years, an antagonistic effect of estrone and estradiol has been hypothesized. In fact, estrone has a C-2 hydroxylated subtype that acts as competitive inhibitor of estradiol (Lim et al. 1997; Vandewalle et al. 1989; Schneider et al. 1984). Furthermore, these conflicting results can be explained by the metabolic exchange rate between estrone and estradiol. Under in vivo conditions, estrone can be rapidly converted to estradiol and it acts as a reservoir for estradiol production.

258

Estrogens in Alzheimer’s Disease: A Clinical and Neurobiological Perspective

CSF Estrogen Levels and Clinical Characteristics of AD Although several studies investigating serum estrogen and estrone levels among demented and nondemented individuals have revealed conflicting results, to date, endogenous CSF estrogen status has not been investigated in demented patients. Regarding CSF estrogen levels in pre- and postmenopausal healthy women, one study found no significant differences between these two groups (Molnár et al. 1997). Since the CSF-blood barrier is supposed to protect the brain from the effects of peripheral estrogen deficiency, the impact of CSF estrogen status on Aβ42 metabolism needed to be further addressed. Concerning this issue, we investigated 30 women with probable AD (NINCDS-ADRDA criteria; McKhann et al. 1984) and 11 women with nondementing diseases such as major depression (DSM-IV) (Schönknecht et al. 2001). All patients were postmenopausal and had no history of ERT. Severity of dementia was rated on the MMSE. CSF 17β-estradiol levels were determined using an electro-chemiluminescence-immunoassay on the Roche Elecsys 2010 immunoassay analyzer. For CSF Aβ40 and Aβ42 concentrations, an enzyme-linked immunosorbent assay (ELISA) established previously by our group was employed (Schröder et al. 1997). In addition, we measured tau protein concentrations using the innotest (Innogenetics) htau antigen kit (Schönknecht et al. 2002a). In order to address the potential confounding effect of severity of dementia on the Aβ42 levels, MMSE scores were partialled out when the 17β-estradiol levels were correlated with Aβ concentrations. 17β-Estradiol levels were significantly (f = 5.5, p < 0.05) lower in the AD patients (16.0 ± 3.0 pg/ml) than in the patients with nondementing diseases (19.1 ± 5.2 pg/ml). While there were only minor, nonsignificant group differences with respect to age and body mass indices, MMSE scores were significantly lower in the AD patients than in the control subjects as expected (16.3 ± 6.6 vs. 26.7 ± 1.8, respectively; f = 21.4, p < 0.005). As expected, tau protein concentrations were significantly higher in the AD patients (591.8 ± 282.3 pg/ml) than in those with nondementing diseases (148.0 ± 30.2 pg/ml) (f = 26.6; p < 0.005). Within the AD group, 17β-estradiol and Aβ42 levels were inversely correlated when severity of illness (MMSE scores) was partialled out (r = – 0.36, p = 0.05) (Fig. 1). None of the other variables investigated including tau protein concentration were significantly corre-

259

Peter Schönknecht et al.

lated with 17β-estradiol levels. This study therefore provides two major findings: firstly, evidence that female AD patients have lower CSF 17β-estradiol levels than nondemented female patients; and secondly, an indication that this deficit may have a mediating effect on Aβ metabolism. This finding is consistent with results from cell culture experiments demonstrating that 17β-estradiol can significantly decrease Aβ40 and Aβ42 release (Xu et al. 1998). To our knowledge, this is the first clinical study which provides in vivo evidence supporting these experimental findings. We did not find any correlation between Aβ40 and 17β-estradiol levels; this, however, parallels the clinical finding of increased Aβ42 but not Aβ40 levels in patients with mild cognitive impairment and mild to moderate stages of AD (Jensen et al. 1999). Although no significant correlation between body mass indices or severity of dementia and 17β-estradiol levels was found, differences in the respective measures have to be discussed as potential confounding variables.

240,00

β-Amyloid 1-42 (pg/ml)

200,00

160,00

120,00

80,00

40,00 10,00

20,00 17β-Estradiol (pg/ml)

Fig. 1. CSF 17β-estradiol levels and CSF Ab42 concentrations in female Alzheimer’s disease patients

260

Estrogens in Alzheimer’s Disease: A Clinical and Neurobiological Perspective

Estrogen Levels and Potential Medication Effects In a further study, we investigated CSF 17β-estradiol concentrations with respect to the clinical characteristics of the disease as well as CSF tau protein concentration and potential medication effects in a larger sample of 59 postmenopausal patients with AD (Schönknecht et al. 2002c). The study confirmed our previous findings summarized above; no significant correlations between CSF 17β-estradiol levels and severity of dementia, age, age at onset of disease, or CSF tau protein levels were found. Subsequently, CSF 17β-estradiol levels were compared among those AD patients with and without neuroleptics. We found a trend towards slightly higher CSF 17β-estradiol levels in nine female AD patients on neuroleptics compared to 50 female AD patients not taking this medication. This difference, however, failed to reach statistical significance (p = 0.08). Based on the assumption that neuroleptic medication in AD might refer to potential subgroups of AD patients who require treatment because of behavioral or cognitive symptoms, we compared AD patients with and without neuroleptic medication with respect to clinical variables. However, age, age at onset of disease, severity of dementia, apo E genotype, and CSF tau protein concentration did not differ significantly between the groups.

17β-Estradiol and Cerebral Glucose Metabolism Recent studies have provided evidence of systematic activation effects of estrogens on cerebral activity and cognitive function (Maki et al. 2001). In animal studies, a beneficial effect of estrogen on memory dysfunction and disturbances in cerebral energy metabolism (Lannert et al. 1998), as well as an estrogen-induced enhancement of glucose transporter expression in cerebral cortical neurons of primates, have been shown (Cheng et al. 2001). Since neuroimaging studies give evidence of hormone-related changes in brain areas such as the hippocampus, an area typically affected by the disorder (Pantel et al. 1997a; Schröder et al. 2001), one could hypothesize an association of regional cerebral activity and CSF 17β-estradiol concentration in female patients with AD. We therefore investigated six female patients with probable AD with CSF 17β-estradiol measurements and positron emission tomography

261

Peter Schönknecht et al.

(PET). All patients were postmenopausal and had no history of ERT. Before injection of 225 MBq 18F-2-fluoro-2-deoxy-D-glucose (18F-FDG), blood glucose levels were determined and shown to be below 110 mg/dl in all patients. From 15 min before injection until 45 min after, patients rested in a quiet room with dimmed light. Emission scans over 20 min were acquired, followed by the transmission scans over 5 min using three 68Ge line sources. Measurements were obtained with a whole-body PET system, (ECAT EXACT HR+; CTI, Knoxville, TN, USA), covering 155 mm in the axial field of view (63 transversal slices; thickness of each slice, 2.4 mm). Data were acquired in the more sensitive three-dimensional (3D) mode without inter-slice tungsten septa, which was found to be equivalent to the 2D mode for quantification of radioactivities used in the clinical setting. The matrix size was 128 × 128 pixels. Basic image processing was done with MEDx 3.0 (Sensor Systems, Inc.), using SPM96 routines (Wellcome Department of Cognitive Neurology, London) on a Silicon Graphics station (Friston et al. 1995). All data were spatially normalized by affine 12-parameter transformation to a standard stereotactic space based on the atlas of Talairach and Tournoux (Talairach and Tournoux 1988). Normalized images were represented on a 78 × 76 × 85 matrix and smoothed by a Gaussian filter of 12 mm full width at half maximum (FWHM). Using a multisubject design, correlations were generated to assess the association of regional cerebral glucose metabolism and CSF 17β-estradiol concentration. The cerebral structures were identified by their coordinates according to the Talairach atlas. The age of the patients ranged from 68 to 78 years with a mean age of 70.3 (± 7.7) years. Mean age at onset of disease was 65.6 (± 8.9) years. Patients presented with a mean MMSE score of 20.5 (± 4.0); mean CSF concentration of 17β-estradiol was 12.86 pg/ml (± 4.0), ranging from 10 pg/ml to 20 pg/ml. SPM analyses revealed a selective significant (p < 0.001) correlation between CSF 17β-estradiol concentration and cerebral glucose metabolism in the left hippocampus (Schönknecht et al. 2002d) (Fig. 2). This effect was not confounded by age or severity of dementia as measured on the MMSE. None of the other variables investigated such as age, age at onset of disease, and MMSE score were significantly correlated with hippocampal glucose metabolism in the AD patients. This represents the first clinical study indicating an association between CSF 17β-estradiol concentration and hippocampal glucose

262

Estrogens in Alzheimer’s Disease: A Clinical and Neurobiological Perspective

Fig. 2. Spots indicating significant correlation of left hippocampal glucose metabolism and CSF 17β-estradiol levels

metabolism in postmenopausal women with AD. The findings confirm results from recent neuroimaging studies showing a significant effect of estrogen on the hippocampus, an area physiologically involved in encoding and retrieval and affected even in early AD (Schacter et al. 1999; Schacter and Wagner 1999). Recently, Maki et al. (2000) reported increased hippocampal blood flow over time in nondemented women receiving hormone therapy, which is consistent with our results. Interestingly, we found an association between CSF 17β-estradiol concentration and hippocampal glucose metabolism in the left hemisphere only. This finding is in accordance with a recent study demonstrating unilateral left hemisphere defects of regional cerebral blood flow as measured by single photon emission computed tomography (SPECT) in female AD patients (Ott et al. 2000). However, with respect to the small sample size in our study, an impact of CSF estrogen levels only on the left hemispheric hippocampal glucose metabolism cannot be generally assumed, but needs to be

263

Peter Schönknecht et al.

addressed in further studies. Similarly, in our study, the question remains unresolved whether the association between hippocampal glucose metabolism and CSF 17β-estradiol concentration represents a physiological impact of estrogen on hippocampal activity in postmenopausal women per se or refers to specific estrogen effects on mechanisms which are directly involved in the pathology of the disorder such as Aβ42 release.

Conclusion Several epidemiological studies have shown that estrogen replacement therapy may delay onset and progression of AD (Tang et al. 1996; Kawas et al. 1997). Yet, the question is unresolved whether this effect is mediated by a more global beneficial effect of estrogens on cerebral metabolism or refers to specific estrogen effects on the mechanisms involved in the pathogenesis of the disorder. Recent studies showed significantly lower CSF 17β-estradiol levels (Schönknecht et al. 2001) and lower serum 17β-estradiol concentrations (Manly et al. 2000), respectively, in female AD patients compared to nondemented controls. Moreover, in one of these studies, there was an indication that 17β-estradiol may have a mediating effect on cerebral β-amyloid metabolism (Schönknecht et al. 2001). Findings from this study support the hypothesis that 17β-estradiol may decrease Aβ42 levels in female patients with AD. Recent clinical trials, however, investigating ERT in postmenopausal AD patients revealed conflicting results (Mulnard et al. 2000; Henderson et al. 2000; Wang et al. 2000; Asthana et al. 2001). One explanation for this discrepancy may be that optimal levels of 17β-estradiol are needed to maintain brain function, but postmenopausal ERT might not restore estrogen homeostasis consistently in all women. In this respect, age of onset, duration, and severity of dementia have to be considered as potential confounds. Experiments in cell cultures indicate that specific types of estrogens have distinct effects on the CNS. Little is known about the effects of different estrogens on the brain; even more basic questions, such as the potential relationship between peripheral and central estrogen levels, are not yet clarified. Another methodological difficulty is the limited sensitivity of the assays available, which prevents the exact determination of rather low postmenopausal estrogen levels.

264

Estrogens in Alzheimer’s Disease: A Clinical and Neurobiological Perspective

Several clinical trials demonstrated the effects of ERT in AD for very specific patient groups concerning age, duration, and severity of dementia. In our study on CSF 17β-estradiol levels in AD patients and control subjects, we could exclude age, age at onset of disease, and severity of dementia as potential confounds. Consistently, we could not find any significant correlation between CSF 17β-estradiol and tau protein, a marker of neurofibrillary pathology in AD. However, in a larger sample of AD patients with and without neuroleptics, we found a trend towards higher CSF 17β-estradiol levels among patients who received this medication (Schönknecht et al. 2002c). This finding emphasizes that potential confounds of estrogen effects in AD still have to be considered. Significantly, in the AD group we found a rather weak (r = – 0.36, p = 0.05) correlation of CSF 17βestradiol and Aβ42levels, indicating further factors in the interaction of estrogens and AD pathophysiology. The finding of significantly correlated CSF 17β-estradiol levels and hippocampal glucose metabolism in postmenopausal woman with AD further supports the potential role of estrogens as mediators of cognitive functioning, even in individuals affected by the disease. Follow-up studies investigating estrogen effects on neuroimaging outcomes are necessary to broaden the evidence base regarding the hypothesis of beneficial estrogen effects in AD. In conclusion, results from the studies reviewed in this chapter suggest that estrogens may modify the onset and course of AD. While the majority of the epidemiological studies indicate that ERT might delay the onset of the disease, results from treatment studies appear to be more controversial. A modest although significant reduction of estrogen levels in serum and CSF of women with AD was described in several clinical studies; however, this effect appeared to strike subgroups of patients in particular. Moreover, 17β-estradiol levels in the CSF were found to be moderately correlated with Aβ42 levels and hippocampal glucose uptake in AD patients. These results have to be weighed against certain methodological problems as well as our limited knowledge on the central effects of different estrogens. Taken together, the available studies do, however, facilitate the hypothesis that estrogens may have a modifying effect on the onset and course of AD, at least in subgroups of patients.

265

Peter Schönknecht et al.

References Asthana A, Baker LD, Craft S, Stanczyk FZ, Veith RC, Raskind MA, Plymate SR (2001) High-dose estradiol improves cognition for women with AD. Neurology 57: 605-612 Asthana S, Craft S, Baker LD, Raskind MA, Birnbaum RS, Lofgreen CP, Veith RC, Plymate SR (1999) Cognitive and neuroendocrine response to transdermal estrogen in postmenopausal woman with Alzheimer’s disease: results of a placebocontrolled, double-blind, pilot study. Psychoneuroendocrino 24: 657-677 Behl C, Skutella T, Lezoualch F, Post A, Widmann M, Newton CJ, Holsboer F (1997) Neuroprotection against oxidative stress by estrogens: structure-activity relationship. Mol Pharmacol 51: 535-541 Beyreuther K, Multhaupt G, Masters CL (1996) Molecular neuropathology in the causation of Alzheimer’s disease. In: Stefanis C, Hippius H (eds) Neuropsychiatry in Old Age: An Update. Hogrefe and Huber Publishers, Seattle Toronto Bern Göttingen, 43-54 Bhathena RK, Anklesaria BS, Ganatra AM, Pinto R (1998) The influence of transdermal oestradiol replacement therapy and medroxyprogesterone acetate on serum lipids and lipoproteins. Br J Pharmacol 45: 170-172 Bishop J, Simpkins JW (1992) Role of estrogens in peripheral and cerebral glucose utilization. Rev Neurosci 3: 121-137 Buerger K, Zinkowski R, Teipel SJ, Tapiola T, Arai H, Blennow K, Andreasen N, Hofmann-Kiefer K, DeBernardis J, Kerkman D, McCulloch C, Kohnken R, Padberg F, Pirttilä T, Schapiro MB, Rapoport SI, Möller HJ, Davies P, Hampel H (2002) Differential diagnosis of Alzheimer disease with cerebrospinal fluid levels of tau protein phosphorylated at threonine 231. Arch Neurol 59: 1267-1272 Cheng CM, Cohen M, Wang J, Bondy CA (2001) Estrogen augments glucose transporter and IGF1 expression in primate cerebral cortex. FASEB J 15: 907-915 Cunningham CJ, Sinnott M, Denihan A, Rowan M, Walsh JB, O’Moore R, Coakley D, Coen RF, Lawler BA, O’Neill DD (2001) Endogenous sex hormone levels in postmenopausal women with Alzheimer’s disease. J Clin Endocrinol Metab 86: 1099-1103 Fillit H (1994) Estrogens in the pathogenesis and treatment of Alzheimer’s disease in postmenopausal women. Ann NY Acad Sci 743: 233-238 Folstein MF, Folstein SE, McHugh PR (1975) Mini-mental state: a practical method for grading the cognitive state of patients for the clinician. Psychiatr Res 12: 189-198 Friston KJ, Holmes AP, Worsley KJ, Poline JP, Frith CD, Frackowiak RSJ (1995) Statistical parametric maps in functional imaging: a general linear approach. Hum Brain Mapp 2: 189-210 Haass C, Selkoe DJ (1993) Cellular processing of ß-amyloid precursor protein and genesis of the amyloid-β peptide. Cell 75: 1039-1042 Henderson V (1997a) Estrogen replacement therapy for the prevention and treatment of Alzheimer’s disease. CNS Drugs 8: 343-351 Henderson VW (1997b) The epidemiology of estrogen replacement therapy and Alzheimer’s disease. Neurology 48: 27-35 Henderson VW, Paganini-Hill A, Miller BL, Elble RJ, Reyes PF, Shoupe D, McCleary

266

Estrogens in Alzheimer’s Disease: A Clinical and Neurobiological Perspective CA, Klein RA, Hake AM, Farlow MR (2000) Estrogen for Alzheimer’s disease in woman: randomized, double-blind, placebo-controlled trial. Neurology 54: 295-301 Hogervorst E, Smith D (2002) The interaction of serum folate and estradiol levels in Alzheimer’s disease. Neuroendocrinol Lett 23: 155-160 Honjo H, Ogino Y, Naitoh K, Urabe M, Kitawaki J, Yasuda J, Yamamoto T, Ishihara S, Okada H, Yonezawa T, Hayashi K, Nambara T (1989) In vivo effects by estrone sulfate on the central nervous system-senile dementia (Alzheimer’s type). J Steroid Biochem 34: 521-525 Honjo H, Ogino Y, Tanaka K et al. (1993) An effect of conjugated estrogen to cognitive impairment in women with senile dementia-Alzheimer’s type: a placebo controlled, double-blind study. J Japan Menopause Soc 1: 167-171 Ida N, Hartmann T, Pantel J, Schröder J, Zerfass R, Förstl H, Sandbrink R, Masters CL, Beyreuther K (1996) Analysis of heterogeneous βA4 peptides in cerebrospinal fluid and blood by a newly developed sensitive Western blot assay. J Biol Chem 271: 22908-22914 Jaffe AB, Toran-Allerand CD, Greengard P, Gandy SE (1994) Estrogen regulates metabolism of Alzheimer amyloid ß precursor protein. J Biol Chem 269: 13065-13068 Jensen M, Schröder J, Blomberg M, Engvall B, Pantel J, Basun H, Wahlund LO, Werle E, Jauss M, Beyreuther K, Lannfelt L, Hartmann T (1999) Cerebrospinal fluid A beta 42 is increased early in sporadic Alzheimer’s disease and declines with disease progression. Ann Neurol 45: 504-511 Johnson JK, McCleary R, Oshita MH, Cotman CW (1998) Initiation and propagation stages of beta-amyloid are associated with distinctive apolipoprotein E, age, and gender profiles. Brain Res 798: 18-24 Kawas C, Resnick S, Morrison A, Brookmeyer R, Corrada M, Zondermann A, Bacal C, Donnell Lingle D, Metter E (1997) A prospective study of estrogen replacement therapy and the risk of developing Alzheimer’s disease: The Baltimore Longitudinal Study of Aging. Neurology 48: 1517-1521 Lamberts SWJ, van den Beld AW, van der Lely AJ (1997) The endrocinology of aging. Science 278: 419-424 Lannert H, Wirtz P, Schuhmann V, Galmbacher R (1998) Effects of Estradiol (-17β) on learning, memory and cerebral energy metabolism in male rats after intracerebroventricular administration of streptozotocin. J Neuronal Transm 105: 1045-1063 Lim SK, Won YJ, Lee JH, Kwon AH, Lee EJ, Kim KR, Lee HC, Huh KB, Chung BC (1997) Altered hydroxylation of estrogen in patients with postmenopausal osteopenia. J Clin Endocrinol Metab 82: 1001-1006 Luine V (1985) Estradiol increases choline acetyltransferase activity in specific basal forebrain nuclei and projection areas of female rats. Exp Neurol 89: 484-490 Lustig RH (1994) Sex hormone modulation of nerual development in vitro. Hormon Behav 28: 383-395 Maki PM, Resnick SM (2000) Longitudinal effects of estrogen replacement therapy on PET cerebral blood flow and cognition. Neurobiol Aging 21: 373-383 Maki PM, Resnick SM (2001) Effects of estrogen on patterns of brain activity at rest

267

Peter Schönknecht et al. and during cognitive activity: a review of neuroimaging studies. Neuroimage 14: 789-801 Manly JJ, Merchant CA, Jacobs DM, Small SA, Bell K, Ferin M, Mayeux R (2000) Endogenous estrogen levels and Alzheimer’s disease among postmenopausal women. Neurology 54: 833-837 McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM (1984) Clinical diagnosis of Alzheimer’s disease: report of the NINCDS-ADRDA work group under the auspices of department of health an human services task force on Alzheimer’s disease. Neurology 34: 939-944 Molnár G, Kassai-Bazsa Z (1997) Gonadotropin, ACTH, prolactin, sexual steroid and cortisol levels in postmenopausal women’s cerebrospinal fluid (CSF). Arch Gerontol Geriatr 24: 269-280 Mooradian AD (1993) Antioxidant properties of steroids. J Steroid Biocgem Mol Biol 45: 509-511 Mulnard RA, Cotman CW, Kawas C, van Dyck CH, Sano M, Doody R, Koss E, Pfeiffer E, Jin S, Gamst A, Grundmann M, Thomas R, Thal LJ (2000) Estrogen replacement therapy for treatment of mild to moderate Alzheimer’s disease: a randomized controlled trial. JAMA 283: 1007-1015 Ohkura T, Isse K, Akazawa K, Hamamoto M, Yaoi Y, Hagino N (1995) Long-term estrogen replacement therapy in female patients with dementia of the Alzheimer type: 7 case reports. Dementia 6: 99-107 Ott BR, Heindl WC, Tan Z, Noto RB (2000) Lateralized cortical perfusion in woman with Alzheimer’s disease. J Gend Specif Med 3: 29-35 Pantel J, Schröder J, Schad LR, Friedlinger M, Knopp MV, Schmitt R, Geißler M, Bluml S, Essig M, Sauer H (1997a) Quantitative magnetic resonace imaging and neuropsychological functions in dementia of the Alzheimer type. Psychol Med 27: 221-229 Pantel J, Schröder J, Essig M, Popp D, Dech H, Knopp MV, Schad LR, Eysenbach K, Backenstraß M, Friedlinger, M (1997b) Quantitative magnetic resonance imaging in geriatric depression and primary degenerative dementia. J Affect Disord 42: 69-83 Petanceska S, Nagy V, Frail D, Gandy S (2000) Ovariectomy and 17β-estradiol modulate the levels of Alzheimer’s amyloid peptides in brain. Neurology 54: 2212-2217 Schacter DL, Wagner AD (1999) Medial temporal lobe activations in fMRT and PET studies of episodic encoding and retrieval. Hippocampus 9: 7-24 Schacter DL, Curran T, Reiman EM, Chen K, Bandy DJ, Frost JT (1999) Medial temporal lobe activation during episodic encoding and retrieval: a PET study. Hippocampus 9: 575-581 Schneider J, Huh MM, Bradlow HL, Fishman J (1984) Antiestrogen action of 2-hydroxyestrone on MCF-7 human breast cells. J Biol Chem 259: 4840-4845 Schönknecht P, Pantel J, Klinga K, Jensen M, Hartmann T, Salbach B, Schröder J (2001) Reduced cerebrospinal fluid estradiol levels are associated with increased β-amyloid levels in female patients with Alzheimer’s disease. Neurosci Lett 307: 122-124 Schönknecht P, Pantel J, Hartmann T, Werle E, Volkmann M, Essig M, Amann M, Zanabili N, Bardenheuer H, Hunt A, Schröder J (2002a) Cerebrospinal fluid

268

Estrogens in Alzheimer’s Disease: A Clinical and Neurobiological Perspective tau levels in Alzheimer’s disease are elevated when compared to vascular dementia but do not correlate with measures of cerebral atrophy. Psychiatr Res (in press) Schönknecht P, Pantel J, Volkmann M, Hunt A, Schroeder J (2002b) Total and phosphorylated cerebrospinal fluid tau protein in Alzheimer’s disease and vascular dementia. Neurobiol Aging 23: S380 Schönknecht P, Pantel J, Klinga K, Jensen M, Hartmann T, Beyreuther K, Schröder J (2002c) Cerebrospinal fluid estradiol and ß-amyloid levels in female patients with Alzheimer’s disease. Eur Psychiatry 17: 41s Schönknecht P, Henze M, Hunt A, Klinga K, Pantel J, Schröder J (2002d) Cerebrospinal fluid estrogen levels are associated with hippocampal glucose metabolism. Psychiatr Res: Neuroimaging (in press) Schröder J, Buchsbaum MS, Shihabuddin L, Tang C, Wei T, Spiegel-Cohen J, Hazlett EA, Abel L, Luu-Hsia C, Ciaravolo TM, Marin D, Davis KL (2001) Patterns of cortical activity and memory performance in Alzheimer’s disease. Biol Psychiatry 49: 426-436 Schröder J, Pantel J (1999) Morphologische und funktionelle Bildgebung. In: Förstl H (ed) Alzheimer Demenz. Springer, Berlin Heidelberg New York Schröder J, Pantel J, Ida N, Essig M, Hartmann T, Knopp MV, Schad LR, Masters CL, Beyreuther K (1997) Cerebral changes and cerebrospinal fluid-amyloid in Alzheimer’s disease: a study with quantitative magnetic resonance imaging. Mol Psychiatry 2: 505-507 Senanarong V, Vannasaeng S, Poungvarin N, Ploybutr S, Udompunthurak S, Jamjumras P, Fairbanks L, Cummings JL (2002) Endogenous estradiol in elderly individuals. Arch Neurol 59: 385-389 Talairach JA, Tournoux P (1988) Co-planar stereotaxic atlas of the human brain. Thieme, Paris Tang M, Jacobs D, Stern Y, Marder K, Schofield P, Gurland B, Andrews H, Mayeux R (1996) Effect of estrogen during menopause on risk and age at onset of Alzheimer’s disease. Lancet 348: 429-432 Urabe M, Yamamoto T, Kashiwagi T, Okubo T, Tsuchiya H, Iwasa K, Kikuchi N, Yokota K, Hosokawa K, Honjo H (1996) Effect of estrogen replacement therapy on hepatic triglyceride lipase, lipoprotein lipase and lipids including apolipoprotein E in climacteric and elderly women. Endocrinol J 43: 737-742 Vandewalle B, Lefebvre J (1989) Opposite effects of estrogen and catecholestrogen on hormone-sensitive breast cancer cell growth and differentiation. Mol Cell Endocrinol 61: 239-246 Wang PN, Liao SQ, Liu RS, Lin CY, Chao HT, Lu SR, Yu HY, Wang SJ, Lin HC (2000) Effects of estrogen on cognition, mood, and cerebral blood flow in AD. Neurology 54: 2061-2066 Woolley CS, McEwen BS (1994) Estradiol regulates hippocampal dendritic spine density via na N-methyl-D-aspartate receptor-dependent mechanism. J Neurosci 14: 7680-7687 Xu H, Gouras GK, Greenfield JP, Vincent B, Naslund J, Mazzarelli L, Fried G, Jovanovic JN, Seeger M, Relkin NR, Liao F, Checler F, Buxbaum JD, Chait BT, Thinakaran G, Sisodia SS, Wang R, Greengard P, Gandy S (1998) Estrogen reduces neuronal generation of Alzheimer ß-amyloid peptides. Nature Med 4: 447-451

269

Estrogens in Alzheimer’s Disease: A Clinical and Neurobiological Perspective Yaffe K, Grady D, Pressman A, Cummings S (1998) Serum estrone levels, cognitive performance, and risk of cognitive decline in older community women. J Am Geriatr Soc 46: 816-821

270

14 Estrogen Therapy: Interface Between Gynecology and Psychiatry Khaled M. K. Ismail, G. V. Sunanda and P. M. Shaughn O’Brien

Introduction One of the most widely documented findings in psychiatric epidemiology is that women have higher rates of major depressive episodes than men. This has been found all around the world using a variety of diagnostic schemes and interview methods (Bland et al. 1988; Cheng 1989; Weissman and Myers 1978; Nolen-Hoeksema 1987; Weissman and Klerman 1992). The prevalence of depression among women in these studies has been reported to be between one and a half and three times that of men. It can be argued that societal norms and sex role socialization experiences make it easier for women than for men to admit depression in epidemiological surveys (Phillips and Segal 1969; Young et al. 1990). However, a number of methodological studies have been carried out on this type of response bias in community surveys of nonspecific psychological distress using standard psychometric methods to assess potential biasing factors (Clancy and Gove 1972; Gove and Geerken 1977; Gove 1978). No evidence was found in any of these studies that the significantly higher levels of self-reported distress found among women than among men was due to these biasing factors. Female sex steroids profoundly influence the brain. Apart from the symptoms characterizing female-specific mood disorders such as irritability, dysphoria, affect lability, and changes in appetite, other aspects of brain function, such as sexual activity (Sanders et al. 1983), cognitive capabilities, sensorimotor function, and seizure susceptibility (Bäckström 1976) are also related to serum levels of estrogens, progesterone, and various progesterone metabolites. Moreover, sex steroids play a major role in the etiology and treatment of premenstrual syndrome (PMS), premenstrual dysphoric disorder (PMDD), postnatal depression (PND), and menopausal related complaints. These roles will be discussed in more details later in this chapter.

271

Khaled M. K. Ismail, G. V. Sunanda and P. M. Shaughn O’Brien

Estrogens in the Different Stages of the Life Cycle The major circulating estrogen in women of reproductive age is 17βestradiol. The other significant estrogen is estrone, which is produced from both direct ovarian secretion and peripheral aromatization of ovarian and adrenal-derived androstenedione in the adipose tissue. Estrogen circulates while bound to sex hormone-binding globulin (SHBG), which is produced by the liver; the quantity produced is under partial control of estrogen. The event of childbirth is a time of drastic endocrinological and neurobiological alterations in a woman’s body. During pregnancy, there is a gradual rise in estrogen with a peak occurring at the time of delivery (Williams et al. 1985). There is precipitous drop in estrogen levels just after delivery. Within 48 h of delivery, there is a 400–1000-fold drop in estrogen, bringing it down to follicular phase levels (Sichel et al. 1995). This has been called “estrogen withdrawal.” After the menopause, estradiol and estrone production by the ovaries dramatically declines. Estradiol levels decrease to less than 40 pg/ml, while testosterone production declines by only 25% at the time of the natural menopause. Although adrenal steroid production is gradually reduced with age, there is still substantial androstenedione production for peripheral conversion. Therefore, estrone is the major circulating estrogen in the menopause. Due to an increased androgen-to-estrogen ratio, SHBG declines (Yonkers et al. 2000).

Estrogen and Brain Function Behavioral studies in animals and humans have shown that the estrogen-dominated follicular phase is related to increased activity and wakefulness (Asso and Braier 1982). A higher rapidity of fine motor skills in both the hands and the legs has been noted during the follicular phase, when compared with the luteal phase. Animal experiments and clinical studies suggest that increased estradiol concentrations are related to increased two-points discrimination ability, touch sensitivity, visual function, hearing, and olfactory function (Hampson and Kimura 1988). In addition, an improvement in postural balance has been reported in postmenopausal women on estrogen replacement therapy when compared with women without treatment (Hammar et al. 1996).

272

Estrogen Therapy: Interface Between Gynecology and Psychiatry

Estrogen is important in neuronal development and health (Henderson 1997). Subsets of neurons within the brain possess intracellular estrogen receptors. The classic mechanism of action of a steroid is to act (in conjunction with a steroid receptor) as a genomic transcription factor, hence influencing protein synthesis; such effects are relatively slow (McEwen 1994; Joels 1997). Estrogens also affect brain function through interactions with membrane receptors that do not require genomic activation or ion channels (Joels 1997). In responsive neurons, estrogen promotes cell differentiation, the growth of nerve processes, and the formation of new synapses between nerve cells. Estrogen has clear effects on several neurotransmitter systems, including those using acetylcholine, noradrenaline, serotonin, and dopamine. This can explain many of the functional effects of estrogen. The epileptogenic effects of estrogens may be related to an effect on glutamate. The influence of sex steroids on symptoms such as irritability and depressed mode may be related to their effect on serotonergic transmission and the effect of estrogen on motor control could be dopamine-mediated. The relationship between estrogen and the serotonergic [5-hydroxytryptamine (5-HT)] system is of particular interest. In rat brain, estrogen treatment decreases 5-HT1 receptor density and increases 5-HT2 receptor density in the frontal cortex (Biegon et al. 1983; Summer and Fink 1995; Jarrett 1995; Fiscette et al. 1983) as well as the nucleus accumbens, cingulated cortex, and olfactory cortex (Sumner and Fink 1995; Jarrett 1995; Fink et al. 1996). In raphe cells, estrogen increases serotonin content via stimulation of tryptophan hydroxylase mRNA (Pecins-Thompson et al. 1996). In humans, estrogen increases platelet 5-HT2 binding and is associated with increased binding in the platelet serotonin transporter (Rojansky et al. 1991). Finally, both endogenous and exogenous estrogens increase the prolactin response to serotonin agonists (Halbreich et al. 1995; O’Keane et al. 1991). Estrogen also increases locomotor activity and enhances the activity of excitatory amino acids in the central nervous system of rodents (Smith 1994). It is suggested that the latter property is responsible for the cognition-enhancing effects, but may also relate to mood effects (Sherwin 1994; Phillips and Sherwin 1992; Kampen and Sherwin 1992). These various properties in humans and animals are similar to those found with antidepressant agents, and thus

273

Khaled M. K. Ismail, G. V. Sunanda and P. M. Shaughn O’Brien

strengthen the position that estrogens may have mood-elevating properties.

Premenstrual Syndrome and Premenstrual Dysphoric Disorder Premenstrual syndrome (PMS) is a psychological and somatic disorder of unknown etiology. However, hormonal and other, possibly neuroendocrine, factors probably contribute (O’Brien 1993; Rapkin et al. 1997). Most menstruating women exhibit some premenstrual symptomatology, but hormonal differences do not appear to account for the extreme severity of the symptoms seen in some women. There has been a reluctance, until relatively recently, to accept PMS as a serious condition. This has arisen because of a failure to distinguish true PMS from the milder physiological premenstrual symptoms occurring in the normal menstrual cycle of the majority of women. There is now a trend, especially with psychiatrically trained clinicians, to define PMS in terms of the American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders (DSM-IV; American Psychiatric Association 1994). PMS was defined

Table 1. DSM-IV PMDD diagnostic symptoms 1* Markedly depressed mood, feelings of hopelessness, or self-depreciating thoughts 2* Marked anxiety, tension, feelings of being ‘keyed up’, or ‘on edge’ 3* Marked affective lability 4* Persistent and marked anger, irritability, or increased interpersonal conflicts 5 Decreased interest in usual activities 6 Subjective sense of difficulty in concentrating 7 Lethargy, easy fatigability or lack of energy 8 Marked change in appetite 9 Hypersomnia or insomnia 10 Subjective sense of being overwhelmed or out of control 11 Other physical symptoms * Symptoms required for the diagnosis of premenstrual dysphoric disorder (PMDD) under the Diagnostic Statistical Manual IV; the symptoms marked with an asterisk are defining symptoms and at least one of these must be present for a diagnosis of PMDD

274

Estrogen Therapy: Interface Between Gynecology and Psychiatry

as late luteal phase dysphoric disorder (LLPDD) in DSM-III, and is now premenstrual dysphoric disorder (PMDD) under DSM-IV. Table 1 shows the criteria for a DSM-IV PMDD diagnosis. It is important to clarify what is meant by the terms PMDD and PMS, as current literature often uses them interchangeably. PMDD is the extreme, predominantly psychological end of the PMS spectrum. Some patients may additionally have an underlying psychological disorder that coexists with PMS; others self-diagnose PMS but actually have depression unrelated to their cycle. These patients are characterized by the fact that their symptoms fail to resolve after menstruation.

Definition A woman has PMS if she complains of recurrent psychological or somatic symptoms (often both), occurring specifically during the luteal phase of the menstrual cycle and resolving by the end of menstruation. These symptoms are so severe that they disrupt the patient’s normal functioning, quality of life and interpersonal relationships. Symptoms must have occurred in at least four of the previous six cycles. PMDD is defined according to DSM-IV as five or more of the diagnostic symptoms listed in Table 1 being present for most of the last week of the luteal phase and remitting within a few days after the onset of the follicular phase. At least one of the symptoms must be from the cluster of PMDD defining symptoms. The disturbance caused by the symptoms must interfere markedly with work, school, or usual social activities and relationships. The disturbance must not be an exacerbation of another psychiatric disorder such as major depression disorder, panic disorder, dysthymic disorder, or personality disorder. It has long been suggested that fluctuation in mood may be related to ovarian hormone imbalance (Dalton 1977). Research has produced data, which could support theories of estrogen excess, progesterone deficiency, estrogen/progesterone imbalance, and progesterone excess. None of these has been confirmed, and thus, factors other than differences in the levels of individual hormones must be important. Interactions with other endocrine or biochemical systems could be present or differences in receptor status may be relevant.

275

Khaled M. K. Ismail, G. V. Sunanda and P. M. Shaughn O’Brien

Investigations of the metabolites of progesterone have shown that women with PMS had lower levels of the progesterone metabolite allopregnanolone in the luteal phase (Rapkin et al. 1997). This provides a plausible theory because allopregnanolone has γ-aminobutyric acid (GABA)-ergic activity, which could be lost in allopregnanolone deficiency, thus giving rise to PMS. A link with ovarian hormone changes, particularly progesterone, seems likely however, since the temporal relationship between progesterone secretion and symptoms is so close. Ablation of the ovarian endocrine cycle by oophorectomy or more conveniently by the administration of analogs of gonadotropin-releasing hormone (Gn-RH) is associated with the parallel elimination of PMS symptoms (Hussain et al. 1992). Furthermore, in women whose ovarian cycles have ceased (due to the menopause or bilateral oophorectomy) and who subsequently receive hormone replacement therapy (HRT), a significant percentage redevelop PMS symptoms during the progesterone phase of therapy (Hammarbäck et al. 1985). In a pilot study, women with severe PMS who had undergone hysterectomy and bilateral salpingo-oophorectomy were recruited to assess the effects of hormone replacement on their PMS symptoms (Henshaw et al. 1993). During estrogen-only replacement therapy they remained asymptomatic; when progesterone was administered, PMS symptoms recurred, demonstrating fairly clearly that patients remained sensitive to the effects of progesterone. The knowledge of serotonin involvement in depression has been extended into PMS research. Low serotonin levels in red cells and platelets (Rapkin 1992) have been demonstrated in PMS patients. This serotonin deficiency has been proposed to enhance sensitivity to progesterone (Rapkin et al. 1997). Selective serotonin reuptake inhibitors (SSRIs), such as fluoxetine and sertraline, have been shown to be an extremely efficacious treatment for severe PMS/PMDD (Dimmock et al. 2000). This gives further indirect support to the involvement of serotonin in PMS etiology. Vitamin B6 (pyridoxine) is a cofactor in the final step of the synthesis of serotonin and dopamine from tryptophan. However, no data have yet demonstrated consistent abnormalities either of brain amine synthesis or deficiency of vitamin B6. Other neurotransmitters may have relevance to PMS, for example GABA, dopamine, and acetylcholine, although research data are less convincing for these in comparison to β-endorphin and serotonin.

276

Estrogen Therapy: Interface Between Gynecology and Psychiatry

If these theories are true, it would seem that PMS is not caused by an endocrine imbalance. However, it appears that there is increased sensitivity to normal circulating level of ovarian hormones, particularly progesterone, secondary to a neuroendocrine disturbance, probably serotonin deficiency. Accordingly, approaches to treatment fall into two broad strategies: (a) correction of the neuroendocrine anomaly and (b) suppression of ovulation.

Estrogen for the Treatment of PMS/PMDD There are studies to suggest that estrogenic ovarian suppression may eliminate PMS (Watson et al. 1989; de Lignieres 1986). Estradiol has been used in the form of patches (100–200 µg), subcutaneous implants (50–100 mg), or gel. The last was shown in one study to reduce premenstrual migraine (de Lignieres 1986). Progestogens are necessarily prescribed with estrogen in order to protect the endometrium from the untoward effects (hyperplasia and adenocarcinoma) of unopposed estrogen. However, progestogens may reintroduce PMS-like symptoms. To avoid this systemic effect, progestogen can be used locally [i.e., levonorgestrel intrauterine system (LNG-IUS), when systemic levels remain low and restimulation is not seen]. Research on the use of estrogen plus the LNG-IUS has not yet been published, although there is evidence to demonstrate that suppression of ovulation by estrogen treats PMS symptoms, and there is good evidence to demonstrate that the LNG-IUS successfully protects the endometrium from, or even reverses, hyperplasia. This combination has the potential to eliminate PMS, treat flushes, reduce heavy periods, give endometrial protection, and provide contraception. It is well known that these disorders frequently coexist in the same patient. Most gynecologists would reserve estradiol implants for severe cases and in those patients for whom the menopause is imminent. Testosterone implants have also been given empirically when diminished libido is a significant symptom. Therefore, although estrogen does not seem to be directly related to the etiology of PMS, it can be considered as one of the broadways of controlling the condition, through ovarian suppression or possibly its direct effect on the serotonergic system in the brain.

277

Khaled M. K. Ismail, G. V. Sunanda and P. M. Shaughn O’Brien

Postpartum Psychiatric Disorders Postpartum psychiatric illnesses have a long-lasting effect on the woman, the marital relationship, and the child’s emotional, social, and cognitive development (Weinberg and Tronick 1998). The likelihood of a woman experiencing a depressive episode in the postpartum period is higher in women with a family history and/or personal history of depressive disturbance. Obstetric difficulties appear to be irrelevant in both puerperal psychosis and postnatal depression. There is clear evidence of an increased susceptibility to depression postnatally in comparison with nonchildbearing women. There is a threefold increase in the relative risk of depression in the first 5 weeks after delivery (Cox 1993). Postpartum psychiatric illnesses can be classified into five groups (Sichel 2000) (Table 2). The first group is the maternity blues. This condition is relatively common, affecting 50%–80% of women postpartum. It begins on the second or third postpartum day and should start to remit by the second week (Kendell 1981). Symptoms are mild, transient, and require no treatment other than support and reassurance. The second group of patients are those suffering from “pure” postpartum depression. Postnatal depression is seen in 10%–12% of women, often within the first 4 weeks after delivery (Kumar and Robson 1984; O’Hara et al. 1984; Kendell et al. 1987). The onset of depression is early, rapid, and often severe enough to require therapy. Relapse rates have been estimated at 24%–40% (Davidson and Robertson 1985). The third group comprises postpartum psychiatric illnesses affecting women with a previous history of major or minor depression. These women often have a family history of depression and premenstrual dysphoria. Table 2. Incidence of postpartum psychiatric illnesses Group 1 2 3 4 5

Illness

Incidence

Maternity blues “Pure” postpartum depression Women with a previous history of major and minor depression Postpartum psychosis Comorbid emergence of postpartum depression

278

50–80% 10–12% – 0.05–0.1% –

Estrogen Therapy: Interface Between Gynecology and Psychiatry

The fourth group of postpartum psychiatric illness is postpartum psychosis. It occurs following 0.05%–0.1% of births (Brockington 1996). The psychosis presents acutely in the first 2–4 weeks postpartum with symptoms resembling mania with delusions, hallucinations, agitation, and confusion. It requires hospital admission and aggressive pharmacological management. Relapse rates greater than 90% are reported if the index episode of puerperal psychosis occurred during the 24 months prior to delivery (Kendell 1987). The last group comprises the comorbid emergence of postpartum depression. Panic attacks, generalized anxiety, and obsessive-compulsive disorder (OCD) are evident from a history in the pregravid state, but worsen in the postpartum period followed by depression.

Etiology of Postpartum Psychiatric Illnesses The etiology of postpartum psychiatric disorders is likely to be heterogeneous and the interaction between biological, psychological, and social factors is extremely complex. Consistent endocrine differences have not been found between women who develop postpartum depression and those who do not (Willcox 1985). However, failure to demonstrate systemic evidence of hormone deficiencies does not exclude sex hormones as etiological factors (O’Brien 1994). Peripheral hormone levels need not correspond with brain levels, nor are they necessarily an index of brain receptor numbers and affinity. It is possible that a susceptible subpopulation of women experiences mental disorders as a result of the “estrogen withdrawal” (Sichel 2000).

Role of Estrogen in Postpartum Psychiatric Disorders Gregoire et al (1996) conducted a randomized, controlled trial evaluating transdermal estrogen therapy for the treatment of severe postnatal depression. Sixty-one British women with major postpartum depression were recruited. They were assessed using the Research Diagnostic Criteria (RDC), the Schedule for Affective Disorders and Schizophrenia (SADS), and the Edinburgh Postnatal Depression Scale (EPDS); a score of more than 13 on the EPDS on two occasions was required for enrollment. Participants were randomized

279

Khaled M. K. Ismail, G. V. Sunanda and P. M. Shaughn O’Brien

to receive transdermal 17β-estradiol 200 µg/day or placebo for 6 months. After 3 months, dydrogesterone 10 mg/day was added for 12 days per month. More women in the treatment group required concurrent conventional antidepressant medication (16/34 vs. 10/27) and more women in the placebo group missed their follow-up visits (10/27 vs. 6/34). Estrogen therapy was associated with a greater improvement in depression scores than placebo, but the treatment and placebo groups may not have been sufficiently comparable (Lawrie et al. 2000). The results of Gregoire et al. (1996) suggest that high-dose estrogen therapy could be a useful adjunct after conventional antidepressant therapy alone has failed. However, because of the practical and theoretical disadvantages, which include deep vein thrombosis, endometrial hyperplasia, and inhibition of lactation, it is unlikely to be investigated further (Lawrie et al. 2000). High-dose estrogen was also used prophylactically in 11 postpartum women, seven of whom had a history of puerperal psychoses and four a history of puerperal major depression. (Sichel et al. 1995). Although 10 of these 11 women did not suffer from a postpartum psychiatric illness, the authors caution that this particular treatment regimen might be specific for a small subgroup of women who do not experience depressive disorder outside the context of their pregnancy. Moreover, high-dose estrogen could add to the risk of thromboembolism at a time when the chances of this complication are at their highest. Antidepressants and mood stabilizers are not only effective but are also safer. Although estrogen was shown to help some women with severe postnatal depression, its effectiveness has not been adequately shown and the side effects of this treatment have not been widely studied. Postnatal depression is treated by supportive therapy and conventional antidepressant medication. Finally, hormones will not be the only factors precipitating depression; rather it is a combination of fluctuating hormone levels and the impending social and psychological changes that go along with the birth of a child.

Menopause The World Health Organization (WHO) defines the menopause as “the permanent cessation of menstruation resulting from loss of ovarian follicular activity”. The average age in Western women is 51 years, range 48–58. The perimenopause is defined as the time period

280

Estrogen Therapy: Interface Between Gynecology and Psychiatry

immediately before menopause when endocrinological, biological, and clinical features of approaching menopause commence, and lasts until the first year after the menopause. The postmenopause is defined as dating from the time of menopause, although the menopause cannot be determined until after a period of 12 months of spontaneous amenorrhea has been observed (WHO 1981). Surgical menopause occurs as a result of bilateral oophorectomy, mostly, at the time of hysterectomy. It is estimated that approximately 35% of women will seek medical treatment for symptoms associated with the menopause (Bäckström 1995). The complaints most commonly noted by women include hot flushes, muscle and joint pain, headaches, weight gain, decreased libido, fatigue, insomnia, low mood, and irritability (Stuenkel 1989). Estrogen preparations are used to treat menopausal symptoms, including mood symptoms, which commonly accompany physical symptoms. While estrogen is effective in treatment of hot flushes, evidence for its therapeutic efficacy in ameliorating mood symptoms is less concrete. A possible antidepressant effect of estrogen has been evaluated in menopausal women suffering from depression. Several randomized controlled studies showed no antidepressant effect of estrogen in the doses used for HRT (Coope 1981; Campbell 1976; Thomson and Oswald 1977), whereas other studies suggested that estrogen may improve mood in mildly depressed, menopausal women and in high doses may even be useful for the treatment of severely depressed patients (Ditkoff et al. 1991; Klaiber et al. 1979). Investigation of the relationship between the timing of the menopause and depressed mood has been beset by methodological problems. Longitudinal studies have failed to demonstrate an increase in the frequency of major depression or severe mood changes in relation to the transition (Hällstrom and Samuelsson 1985; Kaufert et al. 1992; Avis et al. 1994). Some studies, however, suggest that milder mood changes, expressed as decreased well-being, are more common after the menopause than before; data in this field are not unanimous (Graham and Sherwin 1987; Kaufert et al. 1992; Hunter 1992; Holte 1992; Ballinger 1990; Dennerstein et al. 1993; McKinlay et al. 1992; Collins and Landgren 1994). Women attending menopause clinics may not be typical. High levels of depressive disorder have been reported in women attending gynecology clinics (Ballinger 1977; Ballinger et al. 1987). Pearce et al.

281

Khaled M. K. Ismail, G. V. Sunanda and P. M. Shaughn O’Brien

(1995) point out the importance of assessing and controlling for environmental factors and stresses. In a study of women with HRT given by estradiol implants, it was found that vasomotor symptoms were related to falling estradiol levels, but psychological symptoms were more closely related to minor stresses or hassles (Alder 1992). The contribution of estrogens to psychological health is controversial. A meta-analysis of the effect of HRT on depressed mood found that estrogen treatment of menopausal women had a moderate to large benefit on mood, compared to control or placebo conditions (Zweifel and O’Brien 1997). The review found a greater effect among perimenopausal than among postmenopausal women. According to this meta-analysis, women who were treated for longer than 8 months exhibited the greatest improvement. In addition, the effect was greater for naturally versus surgically induced menopause. However, in another qualitative review, the benefits of estrogen treatment appeared to be more consistent in surgically menopausal women (Yonkers et al. 2000). While studies did not generally compare one estrogen preparation to another, the trials, which were most reliably able to detect a drugplacebo difference, employed parenteral or transdermal estradiol as a treatment. In a qualitative review, only one of six studies using parenteral or transdermal estradiol was negative, whereas negative findings occurred in 9 of 12 studies using an estrogen compound that predominantly yielded estrone. 17β-Estradiol is the more biologically active estrogen, whereas estrone has one-tenth of the biological activity. This suggests that 17β-estradiol is possibly more effective at ameliorating mood symptoms associated with the menopause (Yonkers et al. 2000). There is a possibility that psychological benefits of HRT may occur because they are secondary to the relief of vasomotor symptoms and a reduction in vaginal dryness. In an American study of 426 pre- and perimenopausal American women, depression was significantly associated with hot flushes but not hormonal levels (Avis et al. 1994). In another study of hormones, sexuality and well-being in 140 women aged 40–60 years, it was found that tiredness was the only significant predictor of depression (Zuckerman et al. 1983). These findings suggest that mood enhancement in symptomatic women treated with HRT is more likely due to a “domino effect” rather than a direct psychotropic or placebo effect.

282

Estrogen Therapy: Interface Between Gynecology and Psychiatry

Role of the Obstetrician/Gynecologist and Psychiatrist Nearly all the psychological disorders associated with reproduction are amenable to treatment with either hormonal therapy or psychotropic drugs. Even though the patient might be seen by either specialist, it is apparent that psychiatrists will not usually prescribe hormones and obstetricians/gynecologists will rarely prescribe antidepressants even for the same patient. Such patients have historically fallen between the two specialists since neither has taken responsibility for the disorder. Most of these women present with a mixture of symptoms. For example, a patient presenting with PMS/PMDD might have associated gynecological disorders like heavy, painful periods, or she might be requesting contraception. Treatments (particularly hormonal) may have effects on the menstrual cycle and can cause irregular uterine bleeding. One cannot expect the psychiatrist to undertake the management of such side effects. Postnatal depression and PMS/PMDD may be associated with coexisting psychiatric problems. The psychiatrist will be the most appropriate specialist to identify and manage these cases. Appropriate training of obstetricians/gynecologists and psychiatrists might be the key factor in improving the service offered by either specialty to these patients. There is also a strong case for the development of combined clinics for treating women’s mental health when the expertise can be combined and shared. These clinics do exist in the USA and in continental Europe but are rarely seen in the UK.

References Alder E (1992) The effects of hormone replacement therapy on psychological symptoms. In: Wijma K, von Schoultz B (eds) Reproductive Life. The Parthenon Publishing Group, Carnforth, 359-364 Asso D, Braier JR (1982) Changes with the menstrual cycle in psychophysiological and self report measures of activation. Biol Psychol 15: 195-197 Avis NE, Brambilla D, McKinlay SM, Vass K (1994) A longitudinal analysis of the association between menopause and depression. Results from the Massachusetts Women’s Health Study. Ann Epidemiol 4: 214-220 Bäckstrom T (1976) Epileptic seizures in women related to plasma estrogen and progesterone during the menstrual cycle. Acta Neurol Scand 54: 321-347 Bäckström T (1995) Symptoms related to the menopause and sex steroid treatments. Ciba Found Symp 191: 171-186

283

Khaled M. K. Ismail, G. V. Sunanda and P. M. Shaughn O’Brien Ballinger CB (1977) Psychiatric morbidity and the menopause: survey of a gynaecological out-patient clinic. Br J Psychiatry 131: 83-89 Ballinger CB (1990) Psychiatric aspects of the menopause. Br J Psychiatry 156: 773-787 Ballinger CB, Browning MC, Smith AH (1987) Hormone profiles and psychological symptoms in perimenopausal women. Maturitus 9: 235-251 Biegon A, Reches A, Snyder L, McEwen BS (1983) Serotonergic and adrenergic receptors in the rat brain: modulation by chronic exposure to ovarian hormones. Life Science 32: 2015-2021 Bland RC, Newman SC, Orn H (1988) Lifetime prevalence of psychiatric disorders in Edmonton, Acta Psychiatr Scand 338: 24-32 Brockington I (1996) Motherhood and Mental Health. Oxford University Press, Oxford Campbell S (1976) Double blind psychometric studies on the effects of natural oestrogens on postmenopausal women. In: Campbell S (ed) Management of the Menopause and the Postmenopausal Years. MTP Press, Lancaster, 149-158 Cheng TA (1989) Sex difference in prevalence of minor psychiatric morbidity: a social epidemiological study in Taiwan. Acta Psychiatr Scand 80: 395-407 Clancy K, Gove W (1972) Sex differences in mental illness: an analysis of response bias in self reports. Am J Clin Hypnosis 14: 205-216 Collins A, Landgren BM (1994) Reproductive health, use of estrogen and experience of symptoms in perimenopausal women: a population-based study. Maturitas 20: 101-111 Coope J (1981) Is oestrogen therapy effective in the treatment of menopausal depression? JR Coll Gen Pract 31: 134-140 Cox JL, Murray D, Chapman G (1993) A controlled study of the onset, duration and prevalence of postnatal depression. Br J Psychiatry 163: 27-31 Davidson, Robertson E (1985) A follow-up study of postpartum illness. Acta Psychiatr Scand 71: 451-457 de Lignieres B, Vincens M, Mauvais-Jarvis P, Mas JL, Touboul PJ, Bousser MG (1986) Prevention of menstrual migraine by percutaneous oestradiol. BMJ 293: 1540 Dennerstein L, Smith AM, Morse C, Burger H, Green A, Hopper J, Ryan M (1993) Menopausal symptoms in Australian women. Med J Aust 159: 232-236 Dimmock PW, Wyatt KM, Jones PW, O’Brien PMS (2000) Efficacy of selective serotonin- reuptake inhibitors in premenstrual syndrome. A systematic. Lancet 356: 1131-1136 Ditkoff EC, Crary WG, Cristo M, Lobo RA (1991) Estrogen improves psychological function in asymptomatic postmenopausal women. Obstet Gynecol 78: 991-995 Fink G, Sumner BE, Rosie R, Grace O, Quinn JP (1996) Estrogen control of central neurotransmission: effect on mood, mental state, and memory. Cell Mol Neurobiol 16: 325-344 Fischette CT, Biegon A, McEwen BS (1983) Sex differences in serotonin 1 receptor binding in rat brain. Science 222: 333-335 Gove WR (1978) Sex differences in mental illness among adult men and women: an evaluation of four questions raised regarding the evidence on the higher rates of women. Soc Sci Med 12: 187-198

284

Estrogen Therapy: Interface Between Gynecology and Psychiatry Gove WR, Geerken MR (1977) Response bias in surveys of mental health: an empirical investigation. Am J Sociol 82: 1289-1317 Gregoire AJP, Kumar R, Everitt B, Henderson AF, Studd JWW (1996) Transdermal oestrogen for the treatment of severe postnatal depression. Lancet 347: 930-933 Halbreich U, Rojansky N, Palter S, Tworek H, Hissin P, Wang K (1995) Estrogen augments serotonergic activity in postmenopausal women. Biol Psychiatry 37: 434-441 Hällstrom T, Samuelsson S (1985) Mental health in the climacteric. The longitudinal study of women in Gothenburg. Acta Obstet Gynecol Scand 130 (Suppl): 13-18 Hammar ML, Lindgren R, Berg GE, Moller CG, Niklasson MK (1996) Effects of hormone replacement therapy on the postural balance among postmenopausal women. Obstet Gynecol 88: 955-960 Hammarbäck S, Backstrom T, Hoist J, von Schoultz B, Lyrenas S (1985) Cyclical mood changes as in the premenstrual tension syndrome using sequential oestrogen–progestagen postmenopausal replacement therapy. Acta Obstet Gynecol Scand 64: 393-397 Hampson E, Kimura D (1988) Reciprocal effects of hormonal fluctuations on human motor and perceptual-spatial skills. Behav Neurosci 102: 456-459 Henderson VW (1997) Estrogen, cognition, and a woman’s risk of Alzheimer’s disease. Am J Med 103 (Suppl 3A): 11-18 Henshaw C, O’Brien PMS, Foreman D, Belcher J, Cox J (1993) An experimental model for PMS. Neuropsychopharmacology 9: 713 Holte A (1992) Influences of natural menopause on health complaints; a prospective study of healthy Norwegian women. Maturitus 14: 127-141 Hunter M (1992) The South-east England longitudinal study of the climacteric and postmenopause. Maturitus 14: 117-126 Hussain SY, Massil JH, Matta WH, Shaw RW, O’Brien PMS (1992) Buserelin in premenstrual syndrome. Gynaecol Endocrinol 6: 57-64 Jarrett RB (1995) Comparing and combining short-term psychotherapy and pharmacotherapy for depression. In: Bekham EE, Leber WR (eds) Handbook of depression, 2nd ed. The Guilford Press, New York, 435-464 Joels M (1997) Steroid hormones and excitability in the mammalian brain. Front Neuroendocrinol 18: 2-48 Kampen DL, Sherwin BB (1994) Estrogen use and verbal memory in healthy postmenopausal women. Obstet Gynecol 83: 979-983 Kaufert PA, Gilbert P, Tate R (1992) The Manitoba project: a re-examination of the link between menopause and depression. Maturitas 14: 143-155 Kendell RE, Chalmers JC, Platz C (1987) Epidemiology of puerperal psychosis. Br J Psychiatry 150: 662-673 Kendell RE, MccGuire RJ, Connor Y, Cox JL (1981) Mood changes in the first three weeks after childbirth. J Affect Disord 3: 317-326 Klaiber EL, Broverman DM, Vogel W, Kobayashi Y (1979) Estrogen therapy for severe persistent depressions in women. Arch Gen Psychiatry 36: 550-554 Kumar R, Robson KM (1984) A prospective study of emotional disorders in child beating women. Br J PsychiatryI 144: 35-47

285

Khaled M. K. Ismail, G. V. Sunanda and P. M. Shaughn O’Brien Lawrie TA, Herxheimer A, Dalton K (2000) Oestrogens and progestogens for preventing and treating postnatal depression (Cochrane review). In: The Cochrane Library, issue 4, Oxford: Update Software McEwen BS (1994) Steroid hormone actions on the brain: when is the genome involved? Horm Behav 28: 396-405 McKinlay SM, Brambilla DJ, Posner JG (1992) The normal menopause transition. Maturitus 14: 103-115 Nolen-Hoeksema S (1987) Sex difference in unipolar depression: evidence and theory. Psychol Bull 101: 259-282 O’Brien PMS (1993) Helping women with premenstrual syndrome. BMJ 307: 1471-1475 O’Hara MW, Neunaber DJ, Zekoski EM (1984) Prospective study of postpartum depression: prevalence, course and predictive factors. J Abnorm Psychol 93: 158-171 O’Keane V, O’Hanlon M, Webb M, Dinan T (1991) d-Fenfluramine/prolactin response throughout the menstrual cycle: evidence for an estrogen induced alteration. Clin Endocrinol 34: 289-292 Pearce J, Hawton K, Blake F (1995) Psychological and sexual symptoms associated with the menopause and the effects of hormone replacement therapy. Br J Psychiatry 167: 163-173 Pecins-Thompson M, Brown NA, Kohama SG, Bethea CL (1996) Ovarian steroid regulation of tryptophan hydroxylase mRNA expression in rhesus macaques. J Neurosci 16: 7021-7029 Phillips DL, Segal BE (1969) Sexual status and psychiatric symptoms. Am Sociol Rev 34: 58-72 Phillips SM, Sherwin BB (1992) Effects of estrogen on memory function in surgically menopausal women. Psychoneuroendocrinology 17: 485-495 Rapkin AJ (1992) The role of serotonin in premenstrual syndrome. Clin Obstet Gynaecol 35: 658–666 Rapkin AJ, Morgan M, Goldman L, Brann DW, Simone D, Mahesh VB (1997) Progesterone metabolite allopregnanolone in women with premenstrual syndrome. Obstet Gynecol 90: 709-714 Rojansky N, Halbreich U, Zander K, Barkai A, Goldstein S (1991) Imipramine receptor binding and serotonin uptake in platelets of women with premenstrual changes. Gynecol Obstet Invest 31: 146-152 Sanders D, Warner P, Bäckstrom T, Bancroft J (1983) Mood, sexuality, hormones and the menstrual cycle. I: Changes in mood and physical state: description of subjects and method. Psychosom Med 45: 487-501 Sherwin BB (1994) Estrogenic effects on memory in women. Ann NY Acad Sci 743: 213-230 Sichel DA (2000) Postpartum psychiatric disorders. In: Steiner M, Yonkers KA, Eriksson E (eds) Mood disorders in women. Martin Dunitz, London, 313-328 Sichel DA, Cohen LS, Robertson LM, Ruttenberg A, Rosenbaum JF (1995) Prophylactic oestrogen in recurrent postpartum affective disorder. Biol Psychiatry 38: 814-818 Smith SS (1994) Female sex steroid hormones from receptors to networks to performance-actions on the sensorimotor system. Progr Neurobiol 44: 55-86

286

Estrogen Therapy: Interface Between Gynecology and Psychiatry Stuenkel CA (1989) Menopause and estrogen replacement therapy. Psychiatr Clin North Am 12: 133-152 Sumner BE, Fink G (1995) Estrogen increases density of 5-hydroxytryptamine-2A receptors in cerebral cortex and nucleus accumbens in the female rat. J Steroid Biochem Mol Biol 54: 15-20 Thomson J, Oswald I (1977) Effect of oestrogen on the sleep, mood and anxiety of menopausal women. BMJ 2: 1317-1319 Watson NR, Studd JW, Savvas M, Garnett T, Baber RJ (1989) Treatment of severe premenstrual syndrome with oestradiol patches and cyclical oral norethisterone. Lancet 2: 730-732 Weinberg MK, Tronick EZ (1998) The impact of maternal psychiatric illness on infant development. J Clin Psychiarty 59: 53-61 Weissman MM, Klerman JK (1992) Depression: current understanding and changing trends. Annu Rev Public Health 13: 319-339 Weissman MM, Myers JK (1978) Affective disorders in a US urban community: the use of research diagnostic criteria in an epidemiological survey. Arch Gen Psychiatry 35: 1304-1311 Willcox DL, Yovich JL, McColm SC, Phillips JM (1985) Progesterone, cortisol and oestradiol 17β in the initiation of human parturition: partitioning between free and bound hormone in plasma. Br J Obstet Gynaecol 92: 65-71 Williams RH, Wilson JD, Foster DW (1985) Williams Text book of Endocrinology. Saunders, Philadelphia World Health Organization (1981) Research on the Menopause. WHO, Geneva Yonkers KA, Bradshaw KD, Halbreich U (2000) Oestrogens, progestins and mood. In: Steiner M, Yonkers K, Eriksson E (eds) Mood diorders in women. Martin Dunitz, London, 207-232 Young MA, Fogg LF, Scheftner WA, Keller MB, Fawcett JA (1990) Sex differences in the lifetime prevalence in depression: does varying the diagnostic criteria reduce the female/male ratio? J Affect Disord 18: 187-192 Zuckerman M, Lubin B, Rinck CM (1983) Construction of new scales for the Multiple Affective Adjective Check List. J Behav Assessment 5: 19-29 Zweifel JE, O’Brien WH (1997) A meta-analysis of the effect of hormone replacement therapy upon depressed mood. Psychoendocrinology 22: 189-212

287

List of Contributors Christian Behl, Prof. Dr. rer. nat. Institute of Physiological Chemistry and Pathobiochemistry Duesbergweg 6 55099 Mainz Germany Niels Bergemann, Dr. Dr. med. Dipl.-Psych. Department of Psychiatry University of Heidelberg Voss-Str. 4 69115 Heidelberg Germany Leslie Born, MSc, PhD Womens’s Health Concerns Clinic St. Joseph’s Healthcare, Room FB-639 50 Charlton Ave. E. Hamilton, Ontario Canada L8N 4A6 Martina Dören, Prof. Dr. med. Klinisches Forschungszentrum Frauengesundheit Universitätsklinikum Benjamin Franklin Freie Universität Berlin, Klingsorstr. 109a 12203 Berlin Germany Tony Edwin, MD Biobehavioral Program SUNY Clinical Center 462 Grider Street Buffalo NY 14215 USA Alain Gregoire, MD West Hampshire NHS Trust, Maples, Tatchbury Mount, Calmore, Southampton SO40 2RZ England

Heinz Häfner, Prof. em. Dr. Dr. Dres. h.c. Central Institute of Mental Health Schizophrenia Research Unit J5 68159 Mannheim Germany Uriel Halbreich, MD, Prof. School of Medicine and Biomedical Sciences SUNY at Buffalo Biobehavioral Program Hayes C, Suite 1 3435 Main St., Building 5 Buffalo NY 14214-3016 USA Robin Z. Hayeems, MD Centre for Addiction and Mental Health 250 College St. Toronto M5T 1R8 Canada Marcus Henze, Dr. med. Im Neuenheimer Feld 400 69115 Heidelberg Germany Aoife Hunt, MSc Department of Psychiatry University of Heidelberg Voss-Str. 4 69115 Heidelberg, Germany Khaled M. K. Ismail, MD Division of Women and Child Health Academic Department of Obstetrics and Gynaecology Maternity Building City General Hospital North Staffordshire Hospital NHS Trust

289

List of Contributors Stoke-on-Trent ST4 6QG United Kingdom Linda S. Kahn, PhD Research Assistant Professor Biobehavioral Program SUNY Clinical Center, Room BB170 462 Grider Street Buffalo, NY 14215 USA Jayashri Kulkarni, MD, PhD, Prof. Alfred Psychiatry Research Centre Old Baker Building Alfred Hospital Commercial Rd. Prahran, Victoria 3181 Australia Pauline M. Maki, PhD University of Illinois at Chicago Department of Psychiatry Center for Cognitive Medicine (MC 913) 912 S. Wood Street Suite 235 Chicago, Ill 60612–7327 USA Christoph Mundt, Prof. Dr. med. Department of Psychiatry University of Heidelberg Voss-Str. 4 69115 Heidelberg, Germany P. M. Shaughn O’Brien, MD, Prof. Division of Women and Child Health Academic Department of Obstetrics and Gynaecology Maternity Building City General Hospital North Staffordshire Hospital NHS Trust Stoke-on-Trent ST4 6QG United Kingdom Johannes Pantel, Prof. Dr. med. Department of Psychiatry and

Psychotherapy I University of Frankfurt Heinrich-Hoffmannstr. 10 60528 Frankfurt/Main Germany Peter Parzer, Dipl.-Psych. Department of Child and Adolescent Psychiatry University of Heidelberg Blumenstr. 4 69115 Heidelberg Germany Franz Resch, Prof. Dr. med Department of Child and Adolescent Psychiatry University of Heidelberg Blumenstr. 8 69115 Heidelberg Germany Susan M. Resnick, PhD National Institute on Aging GRC/LPC/Box 3 5600 Nathan Shoch Dr. Baltimore, MD21224 USA Anita Riecher-Rössler, Prof. Dr. med. Psychiatric Outpatient Department University of Basel Petersgraben 4 4031 Basel Switzerland Benno Runnebaum, Prof. em. Dr. med. Dr. h.c. Department of Obstetrics and Gynaecology University of Heidelberg Voss-Str. 9 69115 Heidelberg Germany Peter Schönknecht, Dr. med. Department of Psychiatry University of Heidelberg Voss-Str. 4

290

List of Contributors 69115 Heidelberg Germany

4025 Basel Switzerland

Johannes Schröder, Prof. Dr. med. Section for Geriatric Psychiatry Department of Psychiatry University of Heidelberg Voss-Str. 4 69115 Heidelberg Germany

Thomas Strowitzki, Prof. Dr. med. Department of Obstetrics and Gynecology University of Heidelberg Voss-Str. 9 69115 Heidelberg Germany

Mary V. Seeman, MD, Prof. em. Centre for Addiction and Mental Health 250 College St. Toronto M5T 1R8 Canada Meir Steiner, MD, PhD, FRCPC, Prof. Womens’s Health Concerns Clinic St. Joseph’s Healthcare, Room FB-639 50 Charlton Ave. E. Hamilton, Ontario Canada L8N 4A6 Gabriela Stoppe, Prof. Dr. med. Psychiatrische Universitätsklinik Basel Wilhelm-Klein-Straße 27

G. V. Sunanda, MD Division of Women and Child Health Academic Department of Obstetrics and Gynaecology Maternity Building City General Hospital North Staffordshire Hospital NHS Trust Stoke-on-Trent ST4 6QG United Kingdom Helmut Vedder, Priv.-Doz. Dr. med. Department of Psychiatry and Psychotherapy Laboratory for Neurobiology Philipps-University of Marburg Rudolf-Blutmann-Str. 8 35033 Marburg Germany

291