Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Biopsychosocial factors in major depressive disorder
(USC Thesis Other)
Biopsychosocial factors in major depressive disorder
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
BIOPSYCHOSOCIAL FACTORS IN MAJOR DEPRESSIVE DISORDER Copyright 2002 by Ariane Marie A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements of the Degree DOCTOR OF PHILOSOPHY (EPIDEMIOLOGY) May 2002 Ariane Marie Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 3093419 Copyright 2002 by Marie, Ariane All rights reserved. ® UMI UMI Microform 3093419 Copyright 2003 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UNIVERSITY OF SOUTHERN CALIFORNIA THE GRADUATE SCHOOL UNIVERSITY PARK LOS ANGELES. CALIFORNIA 90007 This dissertation, w ritten by under the direction o f hx.r. Dissertation Committee, and approved by all its members, has been presented to and accepted by The Graduate School, in partial fulfillm ent of re quirements for the degree of DO CTO R OF PH ILOSOPH Y Dean of Graduate Studies Date A ugust 6 , 2002 -SERTATIQN COMM ! JLhairpersoi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ACKNOWLEDGEMENTS With sincere appreciation for Stanley P. Azen, Andrew F. Leuchter, Donna Spruijt- Metz, Chih-Ping Chou, and Margaret Gatz for their role in making this dissertation possible, and to Jim Gauderman, Kim Siegmund, Steve Cole, and Tom Newton for their guidance and support. Last but not least, a special thank you to my family for their persistent faith in me. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TABLE OF CONTENTS Page Acknowledgements ii List of Tables v List of Figures vi Abstract vii 1. Introduction 1.1. Clinical Significance of Studying Biopsychosocial Factors 1 1.2. Importance of Studying BPS Factors in MDD 2 1.3. Dissertation Research Aims 3 1.4. References 5 2. Review: Overview 6 2 .1. Definition of Major Depressive Disorder 6 2.2. Epidemiology 8 2.2.1. Temporal Trends 9 2.2.2. Prevalence Rates 10 2.2.3. Gender Differences 11 2.2.4. Additional Demographic Differences 12 2.3. Prognosis 13 2.4. Etiology and Pathophysiology 15 2.4.1. Genetics 16 2.4.2. Neurophysiology 17 2.4.3. Neuroendocrine System 18 2.4.4. Immune System 19 2.4.5. Stressful Life Events 20 2.4.6. Social Support 21 2.4.7. Cognitive-Behavioral Processes 23 2.4.8. Personality 24 2.4.9. Diet 25 2.4.10. Physical Activity 27 2.5. Treatment 27 2.5.1. Treatment Options 27 2.5.2. Response Rates 29 2.6. References 32 3. Data Analysis: Predictors of Treatment Outcome in Major Depression 41 3.1. Abstract 41 3.2. Introduction 42 iii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Page 3.3. Methods 46 3.4. Results 52 3.4.1. Sample Characteristics 52 3.4.2. Growth Curve Models 5 8 3.4.3. Logistic Regression Models 65 3.5. Discussion 69 3.6. References 75 4. Grant proposal: Mood and Physiologic Reactivity in Major Depression 79 4.1. Abstract 79 4.2. Specific Aims 80 4.3. Background and Significance 81 4.4. Research Design and Methods 90 4.5. Human Subjects/Vertebrate Animals 104 4.6. References 105 5. Review: Toward a Biopsychosocial Theory of Major Depression 113 5.1. Introduction 113 5.2. Biomedical Approach 114 5.2.1. Biomedical approach to medicine 114 5.2.2. Biomedical approach to maj or depression 115 5.2.3. Strengths and weaknesses of the biomedical approach 116 5.3. Biopsychosocial Approach 120 5.3.1. Biopsychosocial approach to medicine 120 5.3.2. Biopsychosocial approach to major depression 122 5.3.3. Strengths and weaknesses of the biopsychosocial approach 123 5.4. Developing a Biopsychosocial Theory of Major Depression 126 5.4.1. Review data on biopsychosocial factors in major depression 126 5.4.2. Generate biopsychosocial hypotheses about major depression 128 5.4.3. Conduct multifactorial longitudinal studies 132 5.4.4. Reassess diagnostic criteria 13 5 5.5. Summary 136 5.6. References 138 7. Bibliography 147 8. Appendix A 167 iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF TABLES Page Table 2.1. Demographic Factors Associated with MDD 13 Table 2.2. Findings Comparing Drug and Psychosocial Treatments 31 Table 3.1. Demographic and Medical Characteristics by Treatment Groups 53 Table 3.2. Baseline Prefrontal Cordance and Psychological Tests by Treatment Groups 54 Table 3.3. Categorical HDRS Measures of Depression by Treatment Groups 58 Table 3.4. Individual Predictors of HDRS Change in All Subjects 59 Table 3.5. Individual Predictors of HDRS Change by Treatment Groups 59 Table 3.6. Multi-variable Model for FIDRS Change in All Subjects 63 Table 3 .7. Multi-variable Model for HDRS Change by Treatment Groups 63 Table 3.8. Individual Predictors of HDRS Response for All Subjects 65 Table 3.9. Individual Predictors of HDRS Response by Treatment Groups 67 Table 3.10. Multi-variable Model for HDRS Response for All Subjects 68 Table 3.11. Multi-variable Model for HDRS Response by Treatment Groups 68 Table 4.1. Procedure for Reactivity Testing 94 V Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. LIST OF FIGURES Page Figure 3.1. Change in FIDRS Total over Time 56 Figure 3.2. Change in Natural Log of HDRS after Randomization 57 Figure 3.3. Mediation by Prefrontal Cordance 64 Figure 5.1. A General BPS Hypothesis of MDD 131 Figure 5.2. A Research Cycle for Developing a BPS Theory of MDD 136 vi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT Biological, psychological and social factors in major depressive disorder (MDD) were examined in order to facilitate the development of a comprehensive theory of MDD. The literature indicated that numerous biopsychosocial (BPS) factors have been associated with MDD. However, it remains to be determined how these factors act together to contribute to the etiology and progression of MDD. An exploratory analysis using data from two double-blind, placebo-controlled antidepressant clinical trials was conducted in order to identify potential biological or psychological predictors of treatment outcome. Several factors were indicated as predictors of a decrease in depressed mood over time in both drug and placebo groups, including lower ratings of depressed mood and somatic anxiety at baseline, as well as a greater change in prefrontal cordance (a quantitative electroencephalographic measure of brain perfusion) between baseline and wash-in. Analyses by treatment suggested that degree of depressed mood, somatization, and negative cognitions might distinguish drug and placebo outcomes. Change in prefrontal cordance during placebo lead-in appeared to mediate the effects of depressed mood, family history of mood disorders, and somatization. To facilitate further exploration of relationships between BPS factors in MDD, a grant proposal was developed. The proposed research design involves assessing changes in cardiovascular and immune measures in response to acute psychological stressors in patients with MDD and normal controls, before and after pharmacological treatment with venlafaxine or placebo. vii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Unique features of this proposed study include the use of a with-in subjects design to capitalize on the change in mood that occurs during a clinical trial in depressed patients, as well as the simultaneous examination of data on cardiovascular, immune and psychosocial variables. Finally, it was suggested that a specific BPS theory of MDD could be developed by future research that includes reviewing data on BPS factors in MDD, generating BPS hypotheses, conducting multifactorial longitudinal studies, and reassessing diagnostic criteria as needed. The future development of a comprehensive theory of MDD will help elucidate mechanisms relevant to understanding the etiology, treatment and prognosis of MDD. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1. Introduction 1.1. Clinical Significance of Studying Biopsychosocial Factors Human beings are biological organisms whose functioning is continuously influenced by psychologically mediated interactions with the social environment (Levin & Solomon, 1990). Therefore, a comprehensive understanding of health and disease requires the simultaneous study of biological, psychological and social factors (Brody, 1980; Gillett, 1990). Biological factors refer to material structures or processes in the human body, and include constructs such as genetic inheritance, neurological volumes, or immune cell activity. Psychological factors refer to mental and behavioral characteristics, such as cognitions, personality, and affect. Social factors describe interactions between the individual and human society, and include stressful life events and social support. The term “psychosocial” refers to both psychological and social factors. Many constructs have biological, psychological and social components, and so are best described by the combined term “biopsychosocial” (BPS). For instance, the influence of gender, a chronic illness, or a family history of a particular illness may be evaluated by a combination of biological, psychological, and social variables. Likewise, health behaviors, such as diet and exercise, are multifactorial constructs that may have reciprocal relationships with biological, psychological, and social variables. Research suggests that both biological and psychosocial factors are critical to understanding most “physical” diseases, such as atherosclerosis, type I diabetes, and asthma, as well as “mental” diseases, such as Major Depressive Disorder (MDD) 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (McEwen & Stellar, 1993; Whybrow, 1997). The challenge for clinical research is to find ways to study individual differences in BPS factors in order to reduce the amount of guessing in the practice of medicine, and to improve the odds of effective treatment for individual patients. Application of a BPS approach to clinical research offers the possibility of explaining why individuals with the same risk factors or the same symptoms follow different clinical courses (Engel, 1977; Temoshok, 1990). In addition, BPS studies have the potential to increase understanding of the relationship between mind and body, which could lead to a better understanding of health promotion and disease prevention. 1.2. Importance of Studying BPS Factors in MDD At present, only a limited number of BPS studies of MDD exist. Most clinical trials focus on changes in psychosocial symptoms, and do not measure any biological variables. Most experimental studies focus on one domain to the exclusion of others. For example, neuroimaging or neurophysiologic studies of patients with MDD usually do not measure psychosocial factors such as personality, exercise, or social support. Likewise, social science studies of MDD usually omit biological variables. Although there are a growing number of interdisciplinary studies examining depressed mood, this is not the same as interdisciplinary research on clinical depression. Lack of knowledge about the etiology and treatment of mental illnesses demands that BPS factors be examined even more closely than for illnesses of 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. known etiology. BPS studies of MDD are needed in order to determine the relative influence of each factor. It is plausible that biological, psychological, and social factors may account for different incident rates or treatment outcomes, but this can only be determined if biological, psychological and social factors are measured (Temoshok, 1990; Wallace, 1990). In addition, BPS studies are needed in order to examine relationships between factors, such as which factors mediate or moderate a specific effect, in order to understand the BPS mechanisms underlying MDD. The development of a BPS theory of MDD may help bridge the physical and psychosocial divisions of medicine with implications for the prevention and treatment of many diseases (McDaniel, 1995). 1.3. Dissertation Research Aims The purpose of this dissertation is to examine BPS factors in MDD and to suggest directions for future research. An extensive review of the literature on MDD is provided, beginning with an overview in Chapter 2 and continuing with a discussion about the development of a BPS theory of MDD in Chapter 5. In Chapter 3, an exploratory analysis of two antidepressant clinical trials is presented in order to describe predictors of general response after controlling for treatment, to describe predictors of drug versus placebo response, and to examine relationships between biological and psychological variables. Chapter 4 presents a grant proposal for a study that examines whether physiologic reactivity varies with mood in patients with MDD. The proposed study includes repeated measurements of biological, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. psychological, and social factors. Chapter 6 provides a summary of the significant conclusions from each chapter. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1.4. References Brody, H. (1980). Placebos and the philosophy o f medicine: Clinical, conceptual, and ethical issues. Chicago: University of Chicago Press. Engel, G. (1977). The need for a new medical model: A challenge for biomedicine. Science, 796(4286), 129-136. Gillett, G. (1990). Neuropsychology and meaning in psychiatry. Journal o f Medicine and Philosophy, 75(1), 21-40. Levin, D., & Solomon, G. (1990). The discursive formation of the body in the history of medicine. Journal o f Medicine and Philosophy, 15, 515-537. McDaniel, S. (1995). Collaboration between psychologists and family physicians: Implementing the biopsychosocial model. Professional Psychology: Research and Practice, 26(2), 117-122. McEwen, B., & Stellar, E. (1993). Stress and the individual: Mechanisms leading to disease. Archives o f Internal Medicine, 153, 2093-2101. Temoshok, L. (1990). On attempting to articulate the biopsychosocial model: Psychological-psychophysiological homeostasis. In H. S. Friedman (Ed.), Personality and disease, (pp. 203-225). New York: John Wiley & Sons. Wallace, E. (1990). Mind-body and the future of psychiatry. Journal o f Medicine and Philosophy, 75(1), 41-74. Whybrow, P. (1997). A M ood Apart: A thinker's guide to emotion and its disorders. New York: HarperPerennial. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2. Review: Overview 2.1. Definition of Major Depressive Disorder The term “depression” is frequently used to refer to several different phenomena in both the popular and scientific literature. For all phenomena, depression describes the state of feeling sad. However, different phenomena vary in etiology, severity, and duration. “Depressed mood” refers to the transient experience of feeling sad, helpless, apathetic, or demoralized, and is distinguished from normal sadness by the inability of the individual to exhibit positive affect or emotional involvement (Fombonne, 1994; McDowell & Newell, 1996). Many researchers have used instruments, such as the Center for Epidemiological Studies Depression Scale, to measure depressed mood in the general population. “Situational depression” is the experience of depressed mood that follows a stressor, such as a death in the family, and lasts between a few days and 6 months (McDowell & Newell, 1996). “Clinical depression” refers to one of several possible mood disorders, and is distinct from normal sadness, depressed mood, and situational depression. The research presented here will use the term “depression” to refer to clinical depression, unless otherwise indicated. The Diagnostic and Statistical Manual Fourth Edition (DSM-IV) of the American Psychiatric Association has established specific criteria to distinguish various mood disorders (Association, 1994). The DSM -IV defines four major categories of mood disorders: 1) Depressive Disorders, 2) Bipolar Disorders, 3) Mood Disorders Due to a General Medical Condition, and 4) Substance-Induced 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Mood Disorders. Under the category of Depressive Disorders are three possible diagnoses: 1) Major Depressive Disorder (MDD), 2) Dysthymic Disorder, and 3) Depressive Disorder Not Otherwise Specified. The focus of this research is on MDD. The diagnosis of MDD requires the experience of at least one or more major depressive episodes. A major depressive episode is identified by at least two weeks of depressed mood or loss of interest or pleasure in nearly all activities. In addition, at least four of the following symptoms must be present nearly every day: 1) significant weight loss, weight gain, or change in appetite; 2) insomnia or hypersomnia; 3) psychomotor agitation or retardation; 4) fatigue or loss of energy; 5) feelings of worthlessness or excessive or inappropriate guilt; 6) diminished ability to think or concentrate, or indecisiveness; or 7) recurrent thoughts of death. The symptoms must cause clinically significant distress or impairment in functioning, and cannot be due to bereavement, substance abuse, or a general medical condition. Other features are often associated with a major depressive episode, such as obsessive rumination, anxiety, complaints of pain, and excessive worry about physical health. There may also be abnormal laboratory findings, but currently there are no biological markers that are diagnostic of a major depressive episode (Association, 1994). Different specifiers may be used to describe a recent major depressive episode (Association, 1994). For example, the melancholic or classic pattern of depression is characterized by three or more of the following: 1) distinct quality of depressed mood (i.e. different from depressed mood experienced during 7 permission of the copyright owner. Further reproduction prohibited without permission. bereavement); 2) worsening of depression in the morning; 3) disturbed sleep with early morning awakening; 4) psychomotor retardation or agitation; 5) poor appetite with resultant weight loss; or 6) excessive or inappropriate guilt. By contrast, atypical depressions are characterized by mood reactivity and two or more of the following: 1) significant weight gain or increase in appetite; 2) hypersomnia; 3) leaden paralysis; or 4) interpersonal rejection sensitivity (Association, 1994). The relative importance of different specifiers to the prevention, treatment or prognosis of MDD is still under investigation. 2.2. Epidemiology A major obstacle for understanding the epidemiology of MDD is that different studies use different criteria to define and diagnose depressive disorders (Andreasen & Black, 1995; Lehtinen & Joukamaa, 1994). Use of the DSM-IV has helped increase the reliability of diagnosis among clinicians in the United States. However, there may still be variability among studies in the U.S. based upon the interview instruments used to diagnose MDD. Furthermore, outside the U.S. the diagnosis of depression is based upon the International Classification of Diseases (ICD), and operationalized by various interview instruments published by the World Health Organization (WHO) (Culbertson, 1997; Organization, 1992). Nonetheless, an attempt will be made to summarize current knowledge about MDD, with an emphasis upon studies that used DSM or ICD criteria to define mood disorders. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.2.1. Temporal Trends Several reviews of the literature indicate that over the past century there has been an increase in the prevalence rate of mood disorders and an earlier age at onset in more recent birth cohorts (Fombonne, 1994; Lehtinen & Joukamaa, 1994; Zarate & Tohen, 1996). The most recent epidemiological study to examine time trends and MDD was an analysis of data from the WHO study of Psychological Problems in General Health Care (Simon et al., 1995). A total of 26,421 general practice patients were screened at various centers around the world using the General Health Questionnaire, and from this sample 5603 patients were interviewed using the Composite International Diagnostic Interview. The results indicated that there was a higher cumulative prevalence of MDD in more recent birth cohorts and an earlier age of onset compared to older respondents. Specifically, there was almost a 2-fold increase in risk each decade, such that between the 1902 and 1962 birth cohorts there was a 50-fold increase in risk, and another doubling of risk between the 1962 and 1972 birth cohorts. The fact that this secular trend has been observed in several cross-sectional studies, in different countries and with different measures, suggests to some authors that there is a true birth cohort effect that may reflect changing economic and social stresses (Andreasen & Black, 1995; Fombonne, 1994). This point of view is corroborated by the observation that depression is highly correlated with objective outcomes that have shown a similar secular trend, such as rates of suicide, suicide attempts, and divorce/separation (Wickramaratne & Weissman, 1996). 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. However, Simon et al (1995) argue that the dramatic size of the temporal trend and its persistence across different cultures suggests that in fact the apparent birth cohort may be an artifact of measurement bias. Since all of the studies relied on interviews which used recall of past symptoms to determine lifetime prevalence and ages of onset, it is possible that a tendency to forget remote episodes may cause older individuals to seem to have lower lifetime risk and later age of onset. Reliability studies suggest that small memory errors are common, and can be sufficient to mimic a dramatic birth cohort effect. In addition to recall bias, changing attitudes and decreased stigma may be part of increased reporting in younger individuals. The observation of a similar temporal trend for other psychiatric disorders, such as panic disorder, agoraphobia, and schizophrenia, supports the argument that the apparent birth cohort effect is due to these potential biases (Fombonne, 1994; Simon et al., 1995). Thus, there may not be an increasing risk of depression over time, but there may be dramatic underestimation of lifetime risk (Simon et al., 1995). Unfortunately, reports from longitudinal studies yield mixed results, and are limited by changes in diagnostic methods over time. Until more reliable longitudinal studies are reported, interpretation of the data will remain largely a matter of opinion. 2.2.2. Prevalence Rates Recent reviews of the literature suggest that the current lifetime prevalence of depression is between 5% and 30% depending upon the population studied, how 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. depression is defined, and whether recall bias is considered (Andreasen & Black, 1995; Association, 1994; Musselman, Evans, & Nemeroff, 1998; Simon et al., 1995). Lifetime prevalence rates of depression are consistently higher in women than in men, with ranges from 7-38% for women and 3-29% for men (Association, 1994; Wilhelm, Parker, & Hadzi-Pavlovic, 1997). The point prevalence of depressive disorders tends to be higher in studies of primary care patients compared to studies of the general population. In the U.S., approximately 5% of the general population age 18 and older will experience a major depressive episode in a given year (Antonuccio, Danton, & DeNelsky, 1995). By contrast, as many as 6-10% of primary care patients may experience a major depressive episode in a given year, a prevalence rate which is even higher than hypertension in primary care patients (Gaynes et al., 1999; Simon et al., 1995). 2.2.3. Gender Differences Over the past 30 years, studies in the U.S. and internationally have found that clinical depression is equally common in pre-pubertal boys and girls, but becomes more common in girls during adolescence (Culbertson, 1997; Zarate & Tohen, 1996). By adulthood, studies of developed countries report that the female to male ratio is approximately 2:1 (Andreasen & Black, 1995; Association, 1994). However, most studies of developing countries report no gender differences (Culbertson, 1997). It is possible that reporting bias may be influencing results from developed countries, developing countries, or both. If reporting bias is significantly affecting 11 permission of the copyright owner. Further reproduction prohibited without permission. the results, than rates of different depressive disorders should show similar patterns. By contrast, several studies report that rates of MDD are 3-4 times more common in women than men, but bipolar disorder is equally common in women and men (Culbertson, 1997). More detailed analyses conclude that the gender difference observed in developed countries is a real phenomenon, and not an artifact of reporting bias or help-seeking behavior (Lehtinen & Joukamaa, 1994). 2.2.4. Additional Demographic Differences Limited research exists on ethnic differences in rates of depression. The National Comorbidity study found that Hispanics are the highest reporters of depression, and African-Americans are the lowest reporters (Culbertson, 1997; Kessler, Zhao, Blazer, & Swartz, 1997). Other research suggests that individuals with more traditional lifestyles (e.g. Koreans in Korea or Mexican-Americans in Los Angeles) may be relatively protected from the increasing incidence of depressive disorders (Fombonne, 1994). Further research is needed to sort out cultural influences on reporting behavior in order to determine whether actual differences in risk exist by ethnicity. Additional factors that have been reported to increase risk of depression include family history of mood disorders, chronic physical conditions, anxiety disorders, addictive disorders, and history of physical or sexual abuse (Association, 1994; Culbertson, 1997; Kessler et al., 1997). A summary of demographic factors associated with MDD is shown in Table 2.1. 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2.1. Demographic Factors Associated with MDD _____ ______________ • Possible increasing risk of depression over time • No gender difference in prevalence rate of depression prior to puberty • Higher lifetime prevalence in women • Increased risk associated with family history of mood disorders • Increased risk associated with chronic physical or mental illness • Increased risk associated with history of physical or sexual abuse_________________ 2.3. Prognosis If untreated, most episodes of major depression will remit within about 6 months (Andreasen & Black, 1995; Association, 1994). However, approximately 50-60% of patients will experience a second episode (Association, 1994; Greenberg & Fisher, 1997; Stangl & Greenhouse, 1998). Of those who experience a second episode, about 90% are likely to have a third episode (Stangl & Greenhouse, 1998). In addition, more than 15% of patients who experience a major depressive episode will experience chronic depression lasting 2 or more years (Lara & Klein, 1999). Several factors are predictive of persistent depressive illness, including the severity of the initial depressive episode, comorbid anxiety disorder, high neuroticism scores, comorbid medical illness, less education, less physical activity, unemployment, and low levels of social support (Gaynes et al., 1999). Treatment with antidepressants does not seem to reduce risk of persistent depression (Greenberg & Fisher, 1997). Clinical depression is a major cause of disability worldwide, and the leading cause of disability in the U.S. (Antonuccio et al., 1995; Clarke, 1998). Complications associated with depressive illness include poor performance at work or school, marital discord, and substance abuse (Andreasen & Black, 1995; Judd et Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. al., 2000). The most serious complication is suicide, which results in the death of as many as 15% of patients with MDD (Andreasen & Black, 1995; Association, 1994). Death rates due to all causes are about 4 times higher in individuals with MDD over age 55 (Association, 1994; Penninx et al., 1999). Thus, clinical depression is an important public health concern. MDD is associated with an increased risk of other mental disorders (Association, 1994). As many as 50% of patients with MDD will experience some type of anxiety or personality disorder (Jones, 1998; Wilhelm et al., 1997). However, the high frequency of co-morbid mental disorders may be an artifact of overlapping criteria, or failure to accurately describe distinct phenotypes (Jones, 1998). A history of depression also is associated with an increased risk for a number of somatic illnesses (e.g. hypertension, pulmonary embolism, stroke, muscular sclerosis, renal disease) (Stoudemire, 1995). Most recently there has been a lot of attention to the association between depression and cardiovascular disease, where depression appears to influence both etiology and prognosis (Lavoie & Fleet, 2000; Schwartzman & Glaus, 2000; Wulsin, Vaillant, & Wells, 1999). There also has been some speculation, as well as some evidence, that depressed mood may be associated with the course of some cancers (Holden, Pakula, & Mooney, 1998; McDaniel, Musselman, Porter, Reed, & Nemeroff, 1995; Spiegel, 1996; Watson, Haviland, Greer, Davidson, & Bliss, 1999). In addition, depressive symptoms may influence progression to AIDS (Page-Shafer, Delorenze, Satariano, & Winkelstein, 1996), although there is also conflicting evidence (Leserman et al., 1999). A recurring 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. theme in the literature is that depressed mood may increase risk of somatic illnesses through alterations in immune function. However, the association between depression and somatic illnesses is difficult to explain without understanding relationships between biological and psychosocial factors relevant to MDD. 2.4. Etiology and Pathophysiology At this point in time, there is no integrative theory that explains the etiology of MDD. Several factors are known to be associated with depression. Any of these biological, psychological or interpersonal factors may contribute to the etiology of depression in any given individual (Jones, 1998). However, it remains to be determined how these factors act together to increase an individual’s risk of MDD. Most studies only assess the outcome of depression, which makes it difficult to determine which factors are specific risk factors for depression. Lack of specificity is suggested by a recent longitudinal study, which found a similar pattern of relationships between social support and MDD as between social support and generalized anxiety disorder (Wade & Kendler, 2000). Further evidence of lack of specificity is provided by the observation that exposure to chronic stressors results in depression in only a minority of individuals exposed, and is associated with increased risk of a wide spectrum of later disorders, not just depression (Maughan & McCarthy, 1997). It is possible that psychosocial factors may create vulnerabilities to psychopathology in general, and that determinants of specific disorders may be the result of biological or genetic propensities. 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In addition, evidence for the association between MDD and psychological or social factors is often debated due to the limitations of the study designs employed. The same methodological concerns make it difficult to interpret the biological correlates of depression as well. In general, there is a preponderance of cross- sectional studies and a paucity of longitudinal studies, which limits conclusions about temporal relationships. Many studies use self-report measures, which are subject to misperceptions due to the current experience of depression (Paykel, 1983). The cyclical nature of MDD adds to the complexity of interpreting the relative effects of biological, psychological and social factors from cross-sectional studies, and underscores the need for longitudinal studies. Presented below is an overview of a large literature, which suggests that a number of factors that may be associated with the etiology and/or pathophysiology of MDD. Future multifactorial longitudinal studies will help clarify how these factors are related to MDD. At this point in time, it should be kept in mind that these associations may represent risk factors for MDD specifically, risk factors for psychopathology in general, and/or consequences of MDD. 2.4.1. Genetics Twin and adoption studies suggest that there is a genetic component to depression. For example, the monozygotic to dizygotic twin ratio is estimated to be 4:1 (Andreasen & Black, 1995). In addition, adoptees are at higher risk for MDD if their biological relatives had a history of affective disorder (Whybrow, Akiskal, & 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. McKinney, 1984). However, even among monozygotic twins there is not 100% concordance, suggesting a lack of full penetrance. Given the complexity of MDD, it is likely that multiple genes may interact with each other and/or the environment to influence risk of depression (Fava & Kendler, 2000). 2.4.2. Neurophysiology Neurotransmitter abnormalities implicated in MDD include decreased norepinephrine (NE), decreased serotonin (5-HT), increased acetylcholine (ACh), and dysregulation of dopamine (DA) and gamma-aminobutyric acid (GABA) (Andreasen & Black, 1995; Association, 1994; McDaniel et al., 1995). Structural neuroimaging studies of patients with MDD have found subcortical hyperintensities and reductions in gray matter volumes for areas of the prefrontal cortex, the caudate, and ventral striatum (Drevets, 2000; Fava & Kendler, 2000; Naisberg, 1996). Functional neuroimaging studies report decreased metabolism in the dorsolateral prefrontal cortex, and increased metabolism in the orbital frontal cortex and amygdala (Brody, Barsom, Bota, & Saxena, 2001; Drevets, 2001; Fava & Kendler, 2000; Leuchter et al., 1997). Some of the functional abnormalities appear to be state-dependent, while other abnormalities persist after remission of the major depressive episode (Drevets, 2000). Findings from neuroimaging studies are supported by epidemiological studies reporting high rates of MDD in patients with strokes or tumors that affect the left prefrontal cortex or striatum, in patients with degenerative diseases affecting the basal ganglia (e.g. Parkinson’s or Huntington’s 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. disease), and in patients with seizure disorders affecting temporal lobe structures (Andreasen, 1997; Drevets, 2000). There is also evidence for sleep EEG abnormalities in 40-60% of outpatients and up to 90% of inpatients with MDD (Association, 1994). The most common findings are decreased delta (slow-wave) sleep, reduced REM latency, and longer periods of REM sleep, which are consistent with complaints of disrupted sleep (Andreasen & Black, 1995; Kupfer, 1995). 2,4.3. Neuroendocrine System Neuroendocrine studies indicate that some depressed patients exhibit overactivity of the hypothalamic pituitary adrenal (HP A) axis (Krishnan, Gadde, & Kim, 1998; McDaniel et al., 1995). For example, about 50% of patients with MDD have increased secretion of cortisol throughout a 24-hour period, and depression is observed in about 50% of patients with Cushing’s syndrome (a syndrome characterized by extremely high cortisol levels) (Checkley, 1996). Numerous studies indicate that depressed patients often have increased urinary free cortisol, higher blood levels of cortisol, elevated concentration of CRF in CSF, adrenal and pituitary hypertrophy, and abnormal diurnal variation in cortisol production (Andreasen & Black, 1995; McDaniel et al., 1995). Furthermore, approximately 30-70% of patients with MDD show abnormal suppression of cortisol secretion after administration of dexamethasone (DST), although DST nonsupression is not specific to depression (Andreasen & Black, 1995). Other studies indicate that depressed patients have reduced thyroid hormone response to thyrotropin releasing hormone 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (TRH), and blunting of growth hormone output by the pancreas in response to insulin challenge (Andreasen & Black, 1995; McDaniel et al., 1995; Naisberg, 1996). The pattern of abnormal hormone secretions and the fact that the abnormalities occur across a variety of neuroendocrine target organs suggests that the primary problem may involve the hypothalamus and limbic system (Andreasen & Black, 1995; Drevets, 2000). In addition, the high prevalence of depression in women, postpartum depression, and premenstrual mood changes suggest that changes in sex hormones influence mood, possibly via effects on serotonin regulation or cortisol levels (Legato, 1997; Pearlstein, 1995). 2.4.4. Immune System Recent research suggests that changes in the immune system are part of the pathophysiology of clinical depression. The best supported finding is that there is a decrease in natural killer (NK) cell activity in depressed patients (Herbert & Cohen, 1993; Irwin, 1995; Irwin et al., 1990; Segerstrom, 1997). NK cell activity and number appear to normalize after successful treatment (Irwin, 1995; Ravindran, Griffiths, Merali, & Anisman, 1998). There is also evidence for alterations in the numbers or percentages of neutrophils, B-lymphocytes, and T-lymphocytes (Herbert & Cohen, 1993; Segerstrom, 1997). Another finding is that patients with MDD have an increased release of proinflammatory cytokines and acute phase reactants, such as gamma-IFN and IL-1 beta (Connor & Leonard, 1998; Holden et al., 1998). Proinflammatory cytokines can stimulate the HPA axis, alter glucocorticoid receptor 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. function, and alter monoamine neurotransmission, thus contributing to some of the other pathophysiological dysfunctions observed in depressed patients (Miller, 1998). Thus, preliminary research suggests that there is an association between clinical depression and several immune variables, although several additional factors such as gender, stress, substance abuse, social support, sleep, and physical activity may modify this relationship (Irwin & Friedman, 1999; McDaniel et al., 1995). It is also not yet clear whether immune changes associated with depression are clinically meaningful. Longitudinal studies, or studies on specific populations such as AIDS patients, may help to clarify the clinical significance of the relationship between depressed mood and immune variables (Leserman et al., 1999). 2.4.5. Stressful Life Events A common belief is that stress associated with the loss of a tangible object, such as a loved one, or loss of an intangible object, such as self-esteem, may provoke or promote the continuation of a depressive episode in some individuals (McDaniel et al., 1995). Supportive evidence includes findings that chronic stressors during childhood, such as lack of adequate parental care, increase risk of developing depression (Heim, Owens, Plotsky, & Nemeroff, 1997; Maughan & McCarthy, 1997). For example, women with a history of childhood abuse appear to be four times more likely to develop MDD than women who have not been abused, and are more likely to attempt suicide (Nemeroff et al., 1999). Although the findings of some studies may be influenced by recall bias, the preponderance of evidence Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. suggests that poor early home environment is a risk factor for MDD (Lara & Klein, 1999). In addition, literature reviews suggest that a major depressive episode is often preceded by acute stressful life events (Heim et al., 1997; Maughan & McCarthy, 1997; Paykel, 1994). The risk of suffering a depressive episode is 5-6 times higher within the six months following a stressful life event, such as bereavement, divorce, moving, etc. (Connor & Leonard, 1998). The association between acute stressful life events and MDD appears to be stronger in younger patients with mild-moderate depression, nonmelancholic depression, and fewer episodes of depression (Kendler, Thornton, & Gardner, 2000; Kohn et al., 2001). There is also evidence that chronic low-grade distress due to factors such as unemployment, marital strain, or caregiver stress can increase risk of a depressive episode (Connor & Leonard, 1998; Willner, 1985). 2.4.6. Social Support Social support is another factor that may influence risk of depression. Studies of patients with MDD suggest that lack of social support (e.g. lack of a confidant) is a risk factor for clinical depression (Gruen, 1993; Paykel, 1994; Wade & Kendler, 2000). Researchers have proposed that social support may affect risk of depression via an interaction with stress (the vulnerability hypothesis), or may have independent effects on risk of depression (the strain hypothesis). A recent re analysis of 12 cross-sectional community studies suggests that there is evidence for Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. both hypotheses depending upon the statistical method used (i.e. linear, logit, or probit specifications). The authors argue in favor of the logit or probit specifications, which support the strain hypothesis (Vilhjalmsson, 1993). Thus, there seems to be evidence that social support can have direct effects on depression, independent of stress. Most of the literature conceptualizes social support as an environmental variable that causes depression. However, the relationship between social support and MDD is not likely to be simple. Recent findings suggest that at least 30% of the variance in social support is created by intra-individually stable components, and therefore social support may be more appropriately thought of as a variable with environmental as well as individual difference components (Wade & Kendler, 2000). Support for this hypothesis was provided by a recent longitudinal study of Caucasian female twins that found evidence for a bi-directional relationship between social support and MDD (Wade & Kendler, 2000). Furthermore, the associations were stronger in monozygotic than dizygotic twins, suggesting that social support and depression may be affected by a common set of genetically influenced traits. Findings regarding the influence of social support on depression should be considered preliminary. As one author pointed out, "it is possible to obtain practically any result concerning stressor, support and depression depending on how depression and support have been defined" (Brown & Bifulco, 1985, p350). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.4.7. Cognitive-Behavioral Processes MDD is characterized by a number of cognitive and behavioral symptoms, including recurrent negative thoughts, inability to concentrate, and loss of interest or pleasure in nearly all activities. While these symptoms are traditionally viewed as outcomes of the disorder, it is also possible that cognitive and behavioral processes may contribute to the etiology or persistence of depression (Lara & Klein, 1999; Tewes, 1999). Two well-known theories are the cognitive theory of Beck (Beck, 1967) and Seligman’s theory of learned helplessness (Seligman, 1975). Beck proposed that errors in thought could lead to depressive episodes. Errors in thought might include extensive pessimism, hopelessness, overstatement of failures and understatement of achievements. Seligman proposed that situations where an individual has no control over threatening stimuli could lead to generalized feelings of helplessness resulting in depressed mood. Human and animal experiments suggest that the state of helplessness can result in depressed mood and symptoms characteristic of depression (Tewes, 1999). Further evidence for the etiologic role of cognitive and behavioral processes is provided by the successful treatment of MDD by cognitive-behavioral therapies (Antonuccio et al., 1995; Fava & Kendler, 2000). Future longitudinal studies will help identify the onset of MDD and clarify the relative importance of cognitive-behavioral influences. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.4.8. Personality Certain personality characteristics may create a vulnerability to a depressive episode if an individual encounters a sufficiently stressful life event (Coyne & Whiffen, 1995). For example, the characteristic of sociotropy or dependency refers to needing to establish secure interpersonal relationships in order to increase self esteem. Individuals high in sociotropy/dependency are at risk for depression if relationships fail, and they can be identified by their obsession with themes of loss and abandonment (Coyne & Whiffen, 1995). Autonomous or self-critical individuals are concerned with achievement of personal standards and goals, and may become depressed when they feel that they have failed (Coyne & Whiffen, 1995). Correlational studies suggest that dependency and self-criticism are associated with depression (Gruen, 1993). Some research also suggests that dependency and similar personality characteristics may predict response to antidepressant treatment (Joyce, Mulder, & Cloninger, 1994; Nelson & Cloninger, 1997). A third personality characteristic discussed in the literature is neuroticism i.e. the tendency to be anxious, fearful, or irritable. Studies of primary care patients have found high neuroticism scores to be associated with the persistence of depressive symptoms (Barrett et al., 1999). Additional evidence for the role of neuroticism, as well as stressful life events, comes from a prospective study of 680 Caucasian female-female monozygotic and dizygotic twin pairs. The twins were assessed three times over a period of three years. At least one member of the twin pairs was 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. diagnosed with MDD. Structural equation modeling indicated that 50.1% of the variance in liability to MDD was predicted by stressful life events, genetic factors, previous history of MDD, and neuroticism (Kendler, Kessler, Neale, Heath, & Eaves, 1993). Some studies suggests that neuroticism is highly correlated with dependency and self-criticism (Coyne & Whiffen, 1995). However, more studies are needed that simultaneously measure dependency, self-criticism, and neuroticism in order to understand the relative interdependence of these characteristics and the relationship to MDD. 2.4.9. Diet Recent research suggests the possibility that relative intake of macronutrients and vitamins may be associated with depression. Studies of macronutrient indicate that clinically depressed patients report higher carbohydrate (CHO) intake during depressive episodes (Christensen & Somers, 1996; Patton, 1993). The prevalence of CHO craving appears to be higher in individuals with seasonal affective disorder (67%) compared with depressed individuals without a seasonal pattern (23%) (Patton, 1993). CHO craving is also present in subgroups of populations with high rates of depression, including individuals with obesity, postpartum depression, premenstrual syndrome, and nicotine withdrawl (Christensen & Somers, 1996; Spring, Chiodo, & Bowden, 1987). The tryptophan/serotonin hypothesis proposes that eating high CHO meals helps reduce depressed mood by increasing brain tryptophan and serotonin (Spring et al., 1987). While studies using a tryptophan 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. depletion technique have shown that tryptophan depletion has mood lowering effects, particularly in individuals with a personal or family history of depression, it remains to be determined whether spontaneous changes in macronutrient intake are large enough to significantly alter mood (Klaassen et al., 1999; Moreno et al., 1999; Young, 1993). Research on the role of vitamins in depression suggests that folic acid deficiency is present in some depressed individuals. Folic acid deficiency is rare in the general population, but is present in 33-50% of psychiatric patients, especially those with depression, dementia, or schizophrenia (Young, 1993). It is estimated that folate deficiency is present in about 15-38% of depressed patients, whereas other deficiencies are less common (e.g. vitamin B12 deficiency is present in only 12-14% of depressed patients) (Alpert & Fava, 1997). Conversely, depressive symptoms are manifested by about 56% of patients with folate deficiency (Bottiglieri, 1996). Folate is needed for the one-carbon cycle, which is an essential pathway to transmethylation reactions involving SAM (a methyl donor) within the central nervous system (Alpert & Fava, 1997; Young, 1993). The rate-limiting step in the synthesis of dopamine, norepinephrine, and serotonin requires a cofactor (tetrahydrobiopterin) that is influenced by both folate and SAM (Alpert & Fava, 1997). There is evidence that folate supplementation can contribute to the treatment of some patients with psychiatric disorders (Alpert & Fava, 1997; Young, 1993). Attempts to determine whether folate deficiency is a causal factor in the development of depression, or a consequence of depression-induced malnutrition, 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. have been inconclusive, possibly due to heterogeneity across patient populations (Alpert & Fava, 1997). 2.4.10. Physical Activity A new area of research suggests that there may be a causal link between physical activity and clinical depression. Three prospective studies found that lack of physical activity increased risk of developing clinical depression and provided evidence for a dose-response relationship (Mutrie, 2000). In addition, results from several clinical trials illustrated the effectiveness of exercise programs to treat MDD (Blumenthal et al., 1999; Miser, 2000; Mutrie, 2000). However, one prospective study did not find a relationship between exercise and clinical depression over a 5- year time period (Mutrie, 2000). In addition, some researchers have pointed out methodological weaknesses in the clinical trials of exercise programs (Lawlor & Hopker, 2001). More well-designed studies are needed before a conclusion can be made about the relationship between exercise and MDD. 2.5. Treatment 2.5.1. Treatment Options Two types of treatments are commonly considered for patients with MDD: pharmacological and psychosocial. A third type of treatment, electroconvulsive therapy (ECT), may be considered for patients who have a severe depression or who are unable to take antidepressants. There are a variety of pharmacologic treatments 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. for depression. Tricyclic antidepressants (TCAs) are believed to affect mood by blocking reuptake of norepinephrine and serotonin at presynaptic nerve endings (Andreasen & Black, 1995). Imipramine (Tofranil) is a prototypical example of a TCA. Monoamine Oxidase Inhibitors (MAOIs) inhibit the enzyme that is responsible for oxidation of catecholamines leading to an increase in CNS levels of norepinephrine and serotonin. MAOIs are less commonly prescribed today due to their side-effect profile, but may be considered for patients with atypical depression (Andreasen & Black, 1995; Goodwin, 1993). The newer antidepressants form a third class of drugs, and include Serotonin Reuptake Inhibitors (SRIs) such as fluoxetine (Prozac). SRIs are believed to act by preferentially blocking serotonin reuptake at presynaptic receptors. Venlafaxine is another example of a newer antidepressant that blocks serotonin reuptake, but it also blocks norepinephrine and to a lesser extent dopamine reuptake. Differences in side effects and previous drug history are the main factors considered when selecting an antidepressant to prescribe (Andreasen & Black, 1995). Psychosocial treatments may be offered as an alternative or as an adjunct to pharmacological treatment. Types of psychosocial treatments include behavior therapy, cognitive therapy, various types of psychotherapy, group therapy, family therapy, and social skills training (Andreasen & Black, 1995). Psychosocial treatments may be particularly helpful in addressing issues such as compliance, education about symptoms, provision of insight, and support to help cope with a severe illness (Andreasen & Black, 1995). Support for the efficacy of interpersonal 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. therapy and cognitive therapy has been consistently reported in the literature (Fava & Kendler, 2000). In addition, recent research suggests that exercise programs may provide an effective treatment option for some patients with MDD (Blumenthal et al., 1999; Miser, 2000; Mutrie, 2000). 2.5.2. Response Rates There does not appear to be a significant difference in response rates between TCAs and the newer class of antidepressants (Greenberg & Fisher, 1997; Joflfe, Sokolov, & Streiner, 1996). In general, about 45-70% of patients who are given an antidepressant will respond to treatment, meaning that by the end of the study depressed mood will have decreased below a certain threshold determined by the investigators (Goodwin, 1993; Khan, Warner, & Brown, 2000). Antidepressant trials also show that about 30-50% of patients will respond to placebo (Greenberg & Fisher, 1997; Khan et al., 2000). This is in contrast to other illnesses, such as obsessive-compulsive disorder, where the placebo response rate tends to be negligible (Brown, Johnson, & Chen, 1992). These results suggest that approximately 18-25% more patients will respond to antidepressant than to placebo (Khan et al., 2000). However, the estimation of effect sizes for antidepressants and placebos in depressed patients is controversial due to different opinions about the influence of methodological limitations such as publication bias, unblinding, or inadequate clinical trial length (Fava & Kendler, 2000; Fisher & Greenberg, 1997; Greenberg & Fisher, 1997; Kirsch & Sapirstein, 1998; Swartzman & Burkell, 1998). 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A summary of recent reviews comparing the efficacy of antidepressants and psychosocial treatments is presented in Table 2.2. The psychosocial treatments examined by most studies were cognitive-behavioral therapy or interpersonal therapy. The majority of evidence suggests that psychosocial treatments are at least as effective as pharmacotherapy. The finding of comparable response rates to pharmacotherapy and psychosocial treatments appear to be true for both inpatient and primary care populations (Schulberg, Katon, Simon, & Rush, 1998). In addition, combined drug and psychosocial treatment does not appear to offer a significant advantage over monotherapy for depressed patients in general. However, there may be an advantage to combined treatment for patients with severe depression, comorbidity, chronicity, or treatment resistance (Thase, 1999). The comparison of the efficacy of pharmacotherapy and psychosocial treatments is controversial and conclusions are influenced by investigator allegiance (Deckersbach, Gershuny, & Otto, 2000). Nonetheless, most clinicians would agree that substantial evidence supports the use of antidepressants or psychosocial therapies as viable options for the treatment of MDD. What is less clear, and an important area for future research, is how to determine which treatment will be effective for a given individual at a given time. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2.2. Findings Comparing Drug and Psychosocial Treatments Reference Article Type Conclusions about Efficacy (Thase, 1999) Review C may be better for complex cases (Greenberg & Fisher, 1997) Review D < P (Antonuccio et al., 1995) Review D < P (Wexler & Cicchetti, 1992) Meta-analysis C > D; C = P (Hollon, Shelton, & Loosen, 1991) Meta-analysis D = P; C = P (Robinson, Berman, & Neimeyer, 1990) Meta-analysis D < P; C = P (Dobson, 1989) Meta-analysis D < P (Conte, Plutchik, Wild, & Karasu, 1986) Review D < P; C > P (Steinbrueck, Mawell, & Howard, 1983) Meta-analysis D < P D=drug only, P=psychosocial only, C=combined drug and psychosocial Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.6. References Alpert, J., & Fava, M. (1997). Nutrition and depression: the role of folate. [Review], Nutrition Reviews, 55(5), 145-9. Andreasen, N. (1997). Linking mind and brain in the study of mental illnesses: A project for a psychopathology. Science, 275, 1586-1593. Andreasen, N., & Black, D. (1995). Introductory Textbook o f Psychiatry (2nd ed.). Washington, DC: American Psychiatric Press, Inc. Antonuccio, D., Danton, W., & DeNelsky, G. (1995). Psychotherapy versus medication for depression: Challenging the conventional wisdom with data. Professional Psychology: Research and Practice, 26, 574-585. Association, A. P. (1994). Diagnostic and Statistical Manual o f Mental Disorders (Fourth (DSM-IV) ed ). Washington, DC: Author. Barrett, J. E., Williams, J. W., Jr., Oxman, T. E., Katon, W., Frank, E., Hegel, M. T., Sullivan, M., & Schulberg, H. C. (1999). The treatment effectiveness project. A comparison of the effectiveness of paroxetine, problem-solving therapy, and placebo in the treatment of minor depression and dysthymia in primary care patients: background and research plan. General Hospital Psychiatry, 21(4), 260-73. Beck, A. (1967). Depression: Clinical, experimental, and theoretical aspects. New York: Harper & Row. Blumenthal, J. A., Babyak, M. A., Moore, K. A., Craighead, W. E., Herman, S., Khatri, P., Waugh, R., Napolitano, M. A., Forman, L. M., Appelbaum, M., Doraiswamy, P. M., & Krishnan, K. R. (1999). Effects of exercise training on older patients with major depression. Archives o f Internal Medicine, 759(19), 2349-56. Bottiglieri, T. (1996). Folate, vitamin B12, and neuropsychiatric disorders. Nutrition Reviews, 54(12), 382-390. Brody, A. L., Barsom, M. W., Bota, R. G., & Saxena, S. (2001). Preffontal- subcortical and limbic circuit mediation of major depressive disorder. Seminars in Clinical Neuropsychiatry, 6(2), 102-12. Brown, G., & Bifiilco, A. (1985). Social Support, Life Events and Depression. In I. Saronson & B. Saronson (Eds ), Social Support: Theory, Research, and Applications (pp. 349-370). Boston, MA: Martinus Nijhoff. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Brown, W., Johnson, M., & Chen, M. (1992). Clinical features of depressed patients who do and do not improve with placebo. Psychiatry Research, 41, 203-214. Checkley, S. (1996). The neuroendocrinology of depression and chronic stress. Br M ed Bull, 52, 597-617. Christensen, L., & Somers, S. (1996). Comparison of nutrient intake among depressed and nondepressed individuals. International Journal o f Eating Disorders, 20(1), 105-9. Clarke, D. M. (1998). Psychological factors in illness and recovery. New Zealand Medical Journal, 777(1076), 410-2. Connor, T. J., & Leonard, B. E. (1998). Depression, stress and immunological activation: The role of cytokines in depressive disorders. Life Sciences, 62(1), 583-606. Conte, H., Plutchik, R., Wild, K., & Karasu, T. (1986). Combined psychotherapy and pharmacotherapy for depression: A systematic analysis of the evidence. Archives o f General Psychiatry, 43, 471-479. Coyne, J. C., & Whiffen, V. E. (1995). Issues in personality as diathesis for depression: the case of sociotropy-dependency and autonomy-self-criticism. Psychological Bulletin, 118(f), 358-78. Culbertson, F. (1997). Depression and gender. An international review. American Psychologist, 52(1), 25-31. Deckersbach, T., Gershuny, B. S., & Otto, M. W. (2000). Cognitive-behavioral therapy for depression. Applications and outcome. Psychiatric Clinics o f North America, 23(4), 795-809, VII. Dobson, K. (1989). A meta-analysis of the efficacy of cognitive therapy for depression. Journal o f Consulting and Clinical Psychology, 57, 414-419. Drevets, W. (2000). Functional anatomical abnormalities in limbic and prefrontal cortical structures in major depression. Progress in Brain Research, 126, 413-431. Drevets, W. C. (2001). Neuroimaging and neuropathological studies of depression: implications for the cognitive-emotional features of mood disorders. Current Opinion in Neurobiology, 77(2), 240-9. 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fava, M., & Kendler, K. S. (2000). Major depressive disorder. Neuron, 28(2), 335- 41. Fisher, S., & Greenberg, R. (1997). The curse of the placebo: fanciful pursuit of a pure biological therapy. In S. Fisher & R. Greenberg (Eds.), From Placebo to Panacea: Putting Psychiatric Drugs to the Test (pp. 3-56). New York: John Wiley & Sons, Inc. Fombonne, E. (1994). Increased rates of depression: update of epidemiological findings and analytical problems. Acta Psychiatrica Scandinavica, 90(3), 145-56. Gaynes, B. N., Magruder, K. M., Bums, B. J., Wagner, H. R., Yamall, K. S., & Broadhead, W. E. (1999). Does a coexisting anxiety disorder predict persistence of depressive illness in primary care patients with major depression? General Hospital Psychiatry, 21(3), 158-67. Goodwin, F. (1993). Predictors of antidepressant response. Bulletin o f the Menninger Clinic, 57(2), 146-160. Greenberg, R., & Fisher, S. (1997). Mood-mending medicines: probing drug, psychotherapy, and placebo solutions. In S. Fisher & R. Greenberg (Eds.), From Placebo to Panacea: Putting Psychiatric Drugs to the Test (pp. 115- 172). New York: John Wiley & Sons, Inc. Gruen, R. (1993). Stress and Depression: Toward the development of integrative models. In L. Goldberger & S. Breznitz (Eds.), Handbook o f Stress: Theoretical and Clinical Aspects . New York: The Free Press. Heim, C., Owens, M., Plotsky, P., & Nemeroff, C. (1997). The role of early adverse life events in the etiology of depression and posttraumatic stress disorder: Focus on corticotropin-releasing factor. Annals New York Academy o f Sciences, 821, 194-207. Herbert, T., & Cohen, S. (1993). Depression and immunity: a meta-analytic review. Psychological Bulletin, 113, 472-86. Holden, R , Pakula, I., & Mooney, P. (1998). An immunological model connecting the pathogenesis of stress, depression and carcinoma. Medical Hypotheses, 51(4), 309-14. Hollon, S., Shelton, R , & Loosen, P. (1991). Cognitive therapy and pharmacotherapy for depression. Journal o f Consulting and Clinical Psychology, 59, 88-99. 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Irwin, M. (1995). Psychoneuroimmunology of depression. In F. Bloom & D. Kupfer (Eds), Psychopharmacology: The Fourth Generation o f Progress (pp. 983- 998). New York: Raven Press, Ltd. Irwin, M., Caldwell, C., Smith, T., Brown, S., Schuckit, M., & Gillin, C. (1990). Major depressive disorder, alcoholism, and reduced natural killer cell cytotoxicity. Archives o f General Psychiatry, 47, 713-719. Irwin, M., & Friedman, E. (1999). Does Psychological Depression Cause Immune Suppression in Humans? In M. Schedlowski & U. Tewes (Eds.), Psychoneuroimmunology: An interdisciplinary introduction (pp. 327-340). New York: Kluwer Academic/Plenum Publishers. Joffe, R., Sokolov, S., & Streiner, D. (1996). Antidepressant treatment of depression: A metaanalysis. Canadian Journal o f Psychiatry, 41, 613-616. Jones, E. (1998). Depression: Intervention as assessment. In J. Barron (Ed.), Making Diagnosis Meaningful: Enhancing evaluation and treatment o f psychological disorders (pp. 267-297). Washington, DC: American Psychological Association. Joyce, P. R., Mulder, R. T., & Cloninger, C. R. (1994). Temperament predicts clomipramine and desipramine response in major depression [see comments]. Journal o f Affective Disorders, 30(1), 35-46. Judd, L., Akiskal, H., Zeller, P., Paulus, M., Leon, A., Maser, J., Endicott, J., Coryell, W., Kunovac, J., Mueller, T., Rice, J., & Keller, M. (2000). Psychosocial disability during the long-term course of unipolar major depressive disorder. Archives o f General Psychiatry, 57(4), 375-80. Kendler, K., Kessler, R., Neale, M., Heath, A., & Eaves, L. (1993). The prediction of major depression in women: Toward an integrated etiologic model. American Journal o f Psychiatry, 150, 1139-1148. Kendler, K., Thornton, L., & Gardner, C. (2000). Stressful life events and previous episodes in the etiology of major depression in women: an evaluation of the "kindling" hypothesis. American Journal o f Psychiatry, 757(8), 1243-1251. Kessler, R. C., Zhao, S., Blazer, D. G., & Swartz, M. (1997). Prevalence, correlates, and course of minor depression and major depression in the National Comorbidity Survey. Journal o f Affective Disorders, 45(1-2), 19-30. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Khan, A., Warner, H., & Brown, W. (2000). Symptom reduction and suicide risk in patients treated with placebo in antidepressant clinical trials: An analysis of the Food and Drug Administration database. Archives o f General Psychiatry, 57, 311-317. Kirsch, I., & Sapirstein, G. (1998). Listening to Prozac and hearing placebo: a meta analysis of antidepressant medications. Prevention and Treatment, /(article 2a), 1-17. Klaassen, T., Riedel, W.-J., van-Someren, A., Deutz, N.-E. P., Honig, A., & van- Praag, H.-M. (1999). Mood effects of 24-hour tryptophan depletion in healthy first-degree relatives of patients with affective disorders. Biological Psychiatry, 46(4), 489-497. Kohn, Y., Zislin, J., Agid, O., Hanin, B., Troudart, T., Shapira, B., Bloch, M., Gur, E., Ritsner, M., & Lerer, B. (2001). Increased prevalence of negative life events in subtypes of major depressive disorder. Comprehensive Psychiatry, 42(1), 57-63. Krishnan, K., Gadde, K., & Kim, Y. (1998). Psychoneuroendocrinology and brain imaging in depression. Psychoneuroendocrinology, 21(2), 465-472. Kupfer, D. (1995). Sleep research in depressive illness: clinical implications— a tasting menu. Biological Psychiatry, 38(6), 391-403. Lara, M. E., & Klein, D. N. (1999). Psychosocial processes underlying the maintenance and persistence of depression: implications for understanding chronic depression. Clinical Psychology Review, 19(5), 553-70. Lavoie, K., & Fleet, R. (2000). The impact of depression on the course and outcome of coronary artery disease: review for cardiologists. Canadian Journal o f Cardiology, 16(5), 653-662. Lawlor, D., & Hopker, S. (2001). The effectiveness of exercise as an intervention in the management of depression: Systematic review and meta-regression analysis of randomised controlled trials. British Medical Journal, 322(7289), 763-766. Legato, M. (1997). Gender-specific physiology: how real is it? How important is it? International Journal o f Fertility & Womens Medicine, 42(1), 19-29. Lehtinen, V., & Joukamaa, M. (1994). Epidemiology of depression: Prevalence, risk factors and treatment situation. Acta Psychiatr Scand, Suppl 377, 7-10. 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Leserman, J., Jackson, E., Petitto, J., Golden, R., Silva, S., Perkins, D., Cai, J., Folds, J., & Evans, D. (1999). Progression to AIDS: The effects of stress, depressive symptoms, and social support. Psychosomatic Medicine, 61(3), 397-406. Leuchter, A., Cook, I., Uijtdehaage, S., Dunkin, J., Lufkin, R., Anderson-Hanley, C., Abrams, M L , Rosenberg-Thompson, S., O'Hara, R., Simon, S., Osato, S., & Babaie, A. (1997). Brain structure and function and the outcomes of treatment for depression. Journal o f Clinical Psychiatry, 58(16), 22-31. Maughan, B., & McCarthy, G. (1997). Childhood adversities and psychosocial disorders. British Medical Bulletin, 53(1), 156-169. McDaniel, J., Musselman, D., Porter, M., Reed, D., & Nemeroff, C. (1995). Depression in patients with cancer: Diagnosis, biology, and treatment. Archives o f General Psychiatry, 52, 89-99. McDowell, I., & Newell, C. (1996). Measuring Health: A guide to rating scales and questionnaires (2nd ed.). New York: Oxford University Press. Miller, A.-H. (1998). Neuroendocrine and immune system interactions in stress and depression. Psychiatric Clinics o f North America, 21(2), 443-463. Miser, W. F. (2000). Exercise as an effective treatment option for major depression in older adults. Journal o f Family Practice, 49(2), 109-10. Moreno, F., Gelenberg, A., Heninger, G., Potter, R., McKnight, K., Allen, J., Phillips, A., & Delgado, P. (1999). Tryptophan depletion and depressive vulnerability. Biological Psychiatry, 46, 498-505. Musselman, D. L., Evans, D. L., & Nemeroff, C. B. (1998). The relationship of depression to cardiovascular disease: Epidemiology, biology, and treatment. Archives o f General Psychiatry, 55(1), 580-592. Mutrie, N. (2000). The relationship between physical activity and clinically defined depression. In S. Biddle, K. Fox, & S. Boutcher (Eds.), Physical Activity and Psychological Well-being (pp. 46-62). New York: Routledge. Naisberg, Y. (1996). Homeostatic disruption and depression. Medical Hypotheses, 47(6), 415-22. Nelson, E., & Cloninger, C. R. (1997). Exploring the TPQ as a possible predictor of antidepressant response to nefazodone in a large multi-site study. Journal o f Affective Disorders, 44(2-3), 197-200. 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Nemeroff, C., Heim, C., Owens, M., Newport, D., Miller, A., & Plotsky, P. (1999, ). Neurochemical mechanisms underlying depression and anxiety disorders: The influence of early trauma. New Developments in Understanding Depression and Its Treatment, pp. 3-6. Organization, W. H. (1992). The ICD classification o f mental and behavioural disorders: Clinical descriptions and diagnostic guidelines. Geneva, Switzerland: Author. Page-Shafer, K., Delorenze, G., Satariano, W., & Winkelstein, W. J. (1996). Comorbidity and survivial in HIV-infected men in the San Francisco Men's Health Survey. Ann Epidemiol, 6, 420-430. Patton, G. (1993). Eating problems. In C. Costello (Ed.), Symptoms o f Depression (pp. 227-242). New York: John Wiley & Sons, Inc. Paykel, E. (1983). Methodological aspects of life events research. Journal o f Psychosomatic Research, 27(5), 341-352. Paykel, E. S. (1994). Life events, social support and depression. Acta Psychiatrica Scandinavica, Supplementum, 377, 50-8. Pearlstein, T. (1995). Hormones and depression: What are the facts about premenstrual syndrome, menopause, and hormone replacement therapy? Am J Obstet Gynecol, 173, 646-653. Penninx, B. W., Geerlings, S. W., Deeg, D. J., van Eijk, J. T., van Tilburg, W., & Beekman, A. T. (1999). Minor and major depression and the risk of death in older persons. Archives o f General Psychiatry, 56(10), 889-95. Ravindran, A.-V., Griffiths, J., Merali, Z., & Anisman, H. (1998). Circulating lymphocyte subsets in major depression and dysthymia with typical or atypical features. Psychosomatic Medicine, 60(3), 283-289. Robinson, L., Berman, J., & Neimeyer, R. (1990). Psychotherapy for the treatment of depression. A comprehensive review of controlled outcome research. Psychological Bulletin, 108, 30-49. Schulberg, H. C., Katon, W., Simon, G. E., & Rush, A. J. (1998). Treating major depression in primary care practice: An update of the Agency for Health Care Policy and Research practice guidelines. Archives o f General Psychiatry, 55(12), 1121-1127. 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Schwartzman, J., & Glaus, K. (2000). Depression and coronary heart disease in women: Implications for clinical practice and research. Professional Psychology: Research and Practice, 57(1), 48-57. Segerstrom, S.-C. (1997). Optimism and coping with enduring stress: Psychological and immune effects. Dissertation Abstracts International: Section B: The Sciences and Engineering, >V(6-B), 3355. Seligman, M. (1975). Helplessness. San Francisco: Freeman. Simon, G., Vonkorff, M., Ustun, T., Gater, R., Gureje, O., & Sartorius, N. (1995). Is the lifetime risk of depression actually increasing? Journal o f Clinical Epidemiology, 48(9), 1109-1118. Spiegel, D. (1996). Cancer and depression. British Journal o f Psychiatry - Supplement(30), 109-16. Spring, B., Chiodo, J., & Bowden, D. (1987). Carbohydrates, tryptophan, and behavior: a methodological review. Psychol Bull, 102, 234-56. Stangl, D. K., & Greenhouse, J. B. (1998). Assessing placebo response using Bayesian hierarchical survival models. Lifetime Data Analysis, 4(1), 5-28. Steinbrueck, S., Mawell, S., & Howard, G. (1983). A meta-analysis of psychotherapy and drug therapy in the treatment of unipolar depression with adults. Journal o f Consulting and Clinical Psychology, 51, 856-863. Stoudemire, A. (Ed.). (1995). Psychological Factors Affecting Medical Conditions. Washington, DC: American Psychiatric Press, Inc. Swartzman, L. C., & Burkell, J. (1998). Expectations and the placebo effect in clinical drug trials: why we should not turn a blind eye to unblinding, and other cautionary notes. Clinical Pharmacology & Therapeutics, 64(1), 1-7. Tewes, U. (1999). Concepts in Psychology. In M. Schedlowski & U. Tewes (Eds.), Psychoneuroimmunology: An interdisciplinary introduction (pp. 93-111). New York: Kluwer Academic/Plenum Publishers. Thase, M. E. (1999). When are psychotherapy and pharmacotherapy combinations the treatment of choice for major depressive disorder? Psychiatric Quarterly, 70(4), 333-46. Vilhjalmsson, R. (1993). Life stress, social support and clinical depression: a reanalysis of the literature. Social Science & Medicine, 37(3), 331-42. 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wade, T. D., & Kendler, K. S. (2000). The relationship between social support and major depression: cross-sectional, longitudinal, and genetic perspectives. Journal o f Nervous & Mental Disease, 188(5), 251-8. Watson, M., Haviland, J., Greer, S., Davidson, J., & Bliss, J. (1999). Influence of psychological response on survival in breast cancer: a population-based cohort study. Lancet, 354(9187), 1331-1336. Wexler, B., & Cicchetti, D. (1992). The outpatient treatment of depression: Implications of outcome research for clinical practice. Journal o f Nervous and Mental Disease, 180, 277-286. Whybrow, P., Akiskal, H., & McKinney, W. J. (1984). Toward a psychobiological integration: affective illness as a final common path to adaptive failure, Mood Disorders: towarda New Psychobiology (pp. 173-203). New York: Plenum. Wickramaratne, P., & Weissman, M. (1996). Time Trends in Risk of Depression. Journal o f Clinical Epidemiology, 49(9), 1077-1078. Wilhelm, K., Parker, G., & Hadzi-Pavlovic, D. (1997). Fifteen years on: evolving ideas in researching sex differences in depression. Psychological Medicine, 27(4), 875-83. Willner, P. (1985). Depression: A psychobiological synthesis. New York: John Wiley and Sons. Wulsin, L. R., Vaillant, G. E., & Wells, V. E. (1999). A systematic review of the mortality of depression. Psychosomatic Medicine, 67(1), 6-17. Young, S .-N. (1993). The use of diet and dietary components in the study of factors controlling affect in humans: A review. Journal o f Psychiatry and Neuroscience, 18(5), 235-244. Zarate, C., & Tohen, M. (1996). Epidemiology of mood disorders throughout the life cycle. In K. Shulman, M. Tohen, & S. Kutcher (Eds.), Mood Disorders Across the Life Span (pp. 16-33). New York: Wiley-Liss, Inc. 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3. Data Analysis: Predictors of treatment outcome in major depression 3.1. Abstract We performed a secondary analysis of data from two antidepressant clinical trials in order to explore biological and psychological factors that might predict treatment outcome. Subjects were 51 patients with major depressive disorder (MDD) who participated in one of two nine-week, double-blind placebo-controlled treatment studies including a one-week placebo lead-in. The active medication for one study was fluoxetine, and the active medication for the other study was venlafaxine. A variety of factors were examined in order to identify predictors of treatment outcome as determined by the Hamilton Depression Rating Scale (HDRS). Subjects in both drug and placebo groups were more likely to demonstrate a decrease in HDRS over time if they had lower ratings of depressed mood at baseline, were younger, expressed more negative thoughts at baseline, self-reported fewer somatic symptoms at baseline, were rated with less somatic anxiety at baseline, demonstrated a greater change in prefrontal cordance between baseline and wash-in, reported less awakening at night at baseline, and did not have a family history of mood disorders. Analyses by treatment suggested that some factors might distinguish drug and placebo outcomes such as degree of depressed mood, somatization, somatic anxiety, negative cognitions, or changes in prefrontal brain activity. We also found that change in prefrontal cordance during placebo lead-in appeared to mediate the effects of depressed mood, family history of mood disorders, and somatization. The results 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of this analysis should be considered preliminary hypotheses to be tested in future experimental studies. 3.2. Introduction Major depressive disorder (MDD) is a serious medical illness and the leading cause of disability in the United States (Antonuccio, Danton, & DeNelsky, 1995; Clarke, 1998). MDD will affect approximately 7-38% of women and 3-29% of men over the course of their lifetimes (Association, 1994; Wilhelm, Parker, & Hadzi- Pavlovic, 1997). Thus, treatment of MDD is an important public health issue. While several treatment options are available to patients with MDD, little is known about individual differences in response to specific treatments. Research that examines biological and psychological factors in MDD is needed in order to facilitate prediction of treatment outcomes. In a perfect double-blind randomized clinical trial, 100% of patients would respond to the antidepressant, and 100% of patients would not respond to the placebo, thereby suggesting that the pharmacological action of the drug causes response and the lack of pharmocological action of the placebo causes failure to respond. However, in a typical clinical trial, about a third of patients respond to antidepressants, a third respond to placebo, and a third do not respond to antidepressants or placebos (Greenberg & Fisher, 1997). This makes it difficult to adequately test the efficacy of a new treatment. It also raises questions about factors that contribute to the high placebo response rate. The ability to predict drug response 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. is fundamental to improving the chances of successful treatment for individual patients. Knowledge of factors that predict placebo response could increase understanding of individual differences in recovery from depression, and could be used to improve the efficiency of clinical trials by excluding likely placebo responders. Previous research on predictors of treatment outcome has been more successful at identifying predictors of general response, that is response to either drug or placebo, as opposed to predictors specific to drug or placebo response (Fisher & Greenberg, 1997). Factors known to affect general response in clinical trials include the quality of the physician-patient relationship, expectations of the physician or patient, and environmental context variables such as the color, size, and shape of the pill (Brown, Johnson, & Chen, 1992; Evans, 1985; Fisher & Greenberg, 1997; Harrington, 1997; Moerman & Jonas, 2000; Shapiro & Shapiro, 1997). In addition, psychological characteristics of individuals affect general response. Most notably, individuals who tend to display anxiety, neuroticism, or negative affect are likely to respond positively to drugs or placebo (Fisher & Greenberg, 1997; Shapiro, Struening, Barten, & Shapiro, 1975; Swartzman & Burkell, 1998). The influence of psychological distress on treatment response may be part of the reason for the relatively high placebo response of depressed patients. Factors associated with prediction of general response in antidepressant trials have included severity of baseline depression, presence of psychotic symptoms, and improvement during the placebo lead-in. In particular, studies suggest that individuals with moderate 43 permission of the copyright owner. Further reproduction prohibited without permission. depression severity at baseline are more likely to respond to either drug or placebo, whereas individuals with very low or very high severity scores are less likely to respond (Abou-Saleh & Coppen, 1983; Brown et al., 1992; Paykel, Hollyman, Freeling, & Sedgwick, 1988). Studies also suggest that individuals with delusions or other psychotic symptoms have a poor prognosis, although this could be confounded by the fact that patients with psychotic depression are more depressed on average (Joyce & Paykel, 1989; Kocsis, 1990; Widiger et al., 1996). There is also evidence that patients who improve during the single-blind placebo lead-in are more likely to be responders whether they are given drug or placebo (Quitkin et al., 1998). Only a few factors are known to predict response specific to antidepressant medications. Individuals with the melancholic subtype of MDD are more likely to respond to antidepressants, particularly tricyclic antidepressants (TCAs) (Andreasen & Black, 1995; Goodwin, 1993). Important predictive characteristics of the melancholic subtype appear to be psychomotor retardation, early morning awakening, and weight loss. Individuals with atypical depression, including characteristics such as high anxiety and rejection sensitivity, appear to respond better to monoamine oxidase inhibitors (MAOIs) (Goodwin, 1993; Widiger et al., 1996). Attempts to identify biological measures of antidepressant treatment response, such as indicators of central neurotransmitters or hypothalamic-pituitary-adrenal axis activity, have not been successful (Goodwin, 1993; Tueting, 1991) Even less is known about factors that predict placebo response. A recent retrospective analysis of placebo subjects from three randomized controlled trials of 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a new antidepressant found that the best predictor of placebo response was duration of the current episode, with placebo responders more likely to be depressed for less than 3 months while non-responders were likely to be depressed for more than a year (Brown et al., 1992). There is also evidence that placebo response is predicted by lower baseline ratings of melancholic features, such as psychomotor retardation (Bialik, Ravindran, Bakish, & Lapierre, 1995; Widiger et al., 1996). Other possible predictors of placebo response include female gender and lower somatic anxiety at baseline (Bialik et al., 1995). Recent use of pattern analysis on repeated measures suggests that drug response may be characterized by onset of improvement in weeks 3-5 followed by a non-fluctuating course (Quitkin, 1999). By contrast, placebo response may be characterized by early, abrupt, or non-persistent responses (Quitkin et al., 1991). In summary, several factors are known to predict general response to treatment, but only a few factors have been indicated as possible predictors specific to drug or placebo response in antidepressant clinical trials. Previous research from our laboratory examined brain changes associated with clinical response in patients with MDD. Data was combined from two double blind randomized controlled clinical trials that used the same experimental procedures and inclusion/exclusion criteria. One study (N=24) tested the efficacy of fluoxetine, an antidepressant that preferentially blocks serotonin reuptake by presynaptic receptors. Another study (N=27) tested the efficacy of venlafaxine, another antidepressant that blocks serotonin reuptake, although it also blocks norepinephrine reuptake. Clinical response, the primary outcome variable, was 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. defined as having a Hamilton Depression Rating Scale (HDRS) score <10 at 8-week follow-up. Quantitative electroencephalography (QEEG) was used to measure brain changes, including prefrontal cordance. We found that antidepressant responders demonstrated a significant decrease in prefrontal cordance at 48 hours and 1 week after randomization to treatment. Whether drug subjects experienced a decrease or increase in prefrontal cordance at 48 hours and 1 week predicted medication response with 74% and 70% accuracy respectively (Cook et al., in press). In addition, placebo responders showed a significant increase in prefrontal cordance starting at 1 week after randomization and continuing throughout treatment, whereas antidepressant responders showed a decrease in cordance (Leuchter, Cook, Witte, & Abrams, 2002). The specific aims of this study are to do an exploratory, secondary analysis of the same data in order to: 1) identify predictors of general response; 2) identify predictors of drug versus placebo response; and 3) examine relationships among biological and psychological variables. 3.3. Methods Subjects: Adults diagnosed with a current major depressive episode were recruited for one of two randomized clinical trials conducted at the UCLA Neuropsychiatric Institute. Both studies were nine-week, double-blind placebo-controlled treatment studies including a one-week placebo lead-in. The active medication for one study 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. was fluoxetine, and the active medication for the other study was venlafaxine. Recruitment mechanisms and inclusion/exclusion criteria were identical for both studies. Adults were eligible for the trials if they met criteria for a major depressive episode as determined by a Structured Clinical Interview for DSM-IV (SCID) (Williams et al., 1992), and had not taken psychotropic medication for at least two weeks prior to enrollment. In addition, all subjects had baseline HDRS scores > 16. Subjects were excluded if their HDRS scores fell below 16 and/or dropped more than 10 points after the one-week placebo wash-in. Subjects also were excluded if they had a history of suicidal ideation, had previously failed treatment with the antidepressant being studied, had a medical illness, or were taking a medication that significantly affects brain function. Only subjects who completed the nine-week protocol were examined in this analysis (N=24 for the fluoxetine study, and N=27 for the venlafaxine study). Design and Procedure: After one week of single-blind placebo lead-in treatment, subjects were randomized to receive 8 weeks of either placebo or active medication (fluoxetine 20 mg in one protocol, venlafaxine 150 mg in the other protocol). All subjects also received brief sessions of supportive psychotherapy. Follow-up visits for supportive therapy, symptom monitoring, and data collection occurred tw o days after randomization, and then at weekly intervals. Thus, subjects were tested at the initial visit (baseline), 1-week after placebo treatment (wash-in), two days after 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. randomization (48-hours), and every week after randomization for 8 weeks (weeks 1- 8). Symptoms of a current major depressive episode were assessed during the SCID interview, and added to create a summary score (SCID total). Depressed mood was assessed using the HDRS (Hamilton, 1967), the Beck Depression Inventory (BDI) (Beck, Rush, Shaw, & Emery, 1979; Beck, Ward, Mendelsohn, Mock, & Erbaugh, 1961), and the Montgomery and Asberg Depression Rating Scale (MADRS) (Montgomery, Asberg, Traskman, & Montgomery, 1978). Subjects also completed the Hopkins Symptom Checklist (SCL-90-R) (Derogatis, 1977) and the Hamilton Anxiety Scale (HAM-A) (Hamilton, 1959). In addition, subjects participated in QEEG recordings of cordance, which is a measure of brain perfusion and metabolism derived from absolute and relative EEG power (Leuchter et al., 1997). The UCLA Human Subjects Protection Committee approved both the venlafaxine and fluoxetine clinical trials. Data Analysis: All statistical tests were two-sided. Significant results were defined as a < .05, and marginal or borderline significance was defined as .05 < a < . 10. Demographic and medical characteristics at baseline were contrasted across each of the four treatment conditions (fluoxetine, venlafaxine, placebo from the fluoxetine trial, placebo from the venlafaxine trial), and across the two combined groups (drug vs. placebo). Baseline prefrontal cordance and psychological tests also 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. were compared across the two combined groups. Continuous variables were compared using a two-sample t-test if normally distributed, and a Wilcoxon rank sum test otherwise. Categorical variables were compared using Fisher’s exact tests. We considered both the HDRS and the BDI as potential measures of treatment outcome. Although the BDI has better psychometric properties than the FIDRS (Greenberg & Fisher, 1997), we decided to use the HDRS because it was used in our primary intent-to-treat analysis of the same data, has been widely used in clinical trials, and provides an external assessment of clinical improvement. In support of this decision, we found that the HDRS and BDI measures were statistically significantly correlated over the course of the 9-week trial (r= .69, p < .0001). In addition, test-retest reliability between baseline and wash-in was acceptable for the HDRS (r = .66, p < .0001). A variety of independent variables were considered as potential predictors of treatment outcome based upon a review of the literature. These variables included severity of baseline prefrontal cordance, depressed mood (BDI, MADRS, SCID), duration of the current episode (< 1 year vs. > 1 year), anxiety (HAM-A, SCL-90-R subscales), psychotic symptoms (SCL-90-R subscales), and melancholic symptoms (HDRS and SCID items). The difference between baseline and follow-up values at placebo wash-in were considered as individual predictors for prefrontal cordance, HAM-A, and depressed mood measures (HDRS, BDI, MADRS). Additional demographic and clinical risk factors were examined as potential predictors, including the following: age, gender, ethnicity (white vs. non-white), marital history 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (previously married vs. not), education, family history of mood disorders (yes vs. no), number of primary relatives, number of depressive episodes, chronicity of depression (single vs. recurrent), current medications (cardiac, thyroid, hormones; yes vs. no), additional HDRS items, SCID items, and SCL-90-R subscales. All continuous variables were centered around their means for modeling. Growth curve models were used to determine which variables were the best predictors of change in HDRS scores. The outcome evaluated was change in HDRS scores over six time points between wash-in (randomization to drug or placebo) and 8-week follow-up. A natural log transformation was performed for both HDRS scores and time in weeks in order to satisfy the linear assumption. All models considered fixed and random effects of time, and were adjusted for subject differences at baseline. Significant or marginally significant univariate and multi- variable models were identified for all subjects after adjusting for treatment, and separately by treatment group. In order to minimize Type I errors, backward selection was used for the determination of all multi-variable models. All significant or marginally significant risk factors based upon univariate results were considered for multi-variable models, which is consistent with methods used in other biopsychosocial research (Marusic, Gudjonsson, Eysenck, & Stare, 1999). The relationship between change in prefrontal cordance during the placebo lead-in and psychological predictors of treatment outcome was examined by testing for significant interactions with prefrontal cordance, or mediation by prefrontal cordance. Tests for interactions with prefrontal cordance were done by testing the 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. significance of multiplicative terms. Tests for mediation were by done by determining whether psychological variables were still significant predictors after adjusting for prefrontal cordance. In order to compare results from the growth curve analysis to a more traditional method of analysis, logistic regression models were used to determine which variables were the best predictors of clinical response at 8-week follow-up. The outcome evaluated was clinical response defined as having HDRS scores <10 at 8-week follow-up. Previous studies have used a more stringent cut-off of HDRS < 7, or a definition of response based upon 50% reduction from baseline HDRS total score (Nierenberg & Wright, 1999; Thase, Entsuah, & Rudolph, 2001). We considered both alternate definitions of clinical response, and found that there was a strong association between the cut-off of 10 and the cut-off of 7 (x2=23.5, p < .0001, 82% concordant), as well as the 50% reduction (% 2 = 47.1, p<0001, 98% concordant). We chose to use the cut-off of 10 for our logistic regression models in part because of our small sample size, the short duration of follow-up, and to be consistent with our earlier analyses of the same data (Cook et al., in press; Leuchter et al., 2002). The methods used for the logistic regression analysis were comparable to the methods described for growth curve modeling. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.4. Results 3.4.1. Sample Characteristics Did the subject groups differ on baseline characteristics? The two drug groups were statistically significantly different on number of primary relatives with mood disorders (p=.02). The two placebo groups were statistically significantly different on family history of mood disorders (p=.04), and number of primary relatives with mood disorders (p= 005). Although these differences existed, they did not suggest that the two drug groups or two placebo groups should not be combined. Table 3.1 contrasts baseline characteristics between the combined drug and combined placebo groups. The only baseline demographic difference between the combined drug and combined placebo groups was a slightly higher education level in the placebo group (p=.02). However, this difference did not appear to affect the analysis since education was not a significant predictor of treatment outcome. There were no statistically significant baseline differences in prefrontal cordance or psychological tests for the combined drug and combined placebo groups (Table 3.2). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.1. Demographic and Medical Characteristics by Treatment Groups Characteristics Drug Group (n=25) Placebo Group (n=26) P-value Age 42.2±12.1 41.8+11.9 .81 Gender (MalerFemale) 7:18 12:14 .25 Ethnicity (White: Non-White) 21:4 19:7 .50 Marital History (Yes:No) 9:16 11:15 .78 Education 14.2+2.1 15.6+1.9 .02 Family history of mood disorder (No:Yes)* 11:13 10:16 .77 Number of primary relatives 1.2+1.4 1.0+1.1 .99 Number of episodes 2.6+2.0 3.0+2.8 .85 Chronicity (Single episode:Recurrent) 5:20 6:20 1.00 Duration of current episode (<year: >year) Medications 4:21 3:23 .70 Cardiac (No:Yes) 23:2 23:3 1.00 Thyroid (No:Yes) 23:2 25:1 .61 Hormones (No:Yes) 22:3 24:2 .67 Values are mean±SD for continuous variables and frequencies for categorical variables. Continuous variables were compared using a two-sample t-test if normally distributed, and a Wilcoxon rank sum test otherwise. Categorical variables were compared using two-tail Fisher’s exact tests. * Missing data on 1 drug subject Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.2. Baseline Prefrontal Cordance and Psychological Tests by Treatment Groups Characteristics Drug Group Placebo Group P-value (n=25) (n=26) Prefrontal Cordance -0.4±1.2 -1.0+1.0 .10 HDRS 22.4±3.7 21.9+3.5 .60 BDI* 27.5±7.5 26.4+8.1 .63 MADRS 33.2±4.6 32.7+4.2 .68 SCID total 25.4±1.6 24.7+1.9 .22 HAM-A 21.1+5.2 20.8+6.4 .85 Melancholic Symptoms HDRS Item 2: Guilt (0:1:2:3) 1:8:12:4 2:3:17:4 .33 HDRS Item 5: Insomnia middle (0:1:2) 5:4:16 3:7:16 .56 HDRS Item 6: Insomnia late (0:1:2) 8:5:12 4:13:9 .07 HDRS Item 8: Retardation (0:1:2) 11:8:6 10:10:6 .94 HDRS Item 9: Agitation (0:1:2:3) 17:5:2:1 21:5:0:0 .44 HDRS Item 16: Loss of weight (0:1:2) 17:1:7 21:3:2 .15 SCID Item: Weight (1:2:3) 5:0:20 9:2:15 .16 SCID Item: Guilt (1:2:3) 5:4:16 7:2:17 .65 SCL-90-R** Somatization 1.0+0.7 1.1+0.7 .67 Obsessive-compulsive 2.0+0.6 2.1+0.7 .63 Interpersonal sensitivity 2.0+0.6 1.7+0.7 .11 Depression 2.6+0.6 2.4+0.6 .53 Anxiety 1.5+0.7 1.2+0.7 .26 Hostility 1.1+0.7 1.3+0.8 .52 Phobic anxiety 0.8+0.8 0.5+0.7 .07 Paranoid ideation 1.4+1.0 1.1+0.6 .33 Psychoticism 1.1+0.5 1.1+0.6 .90 General Severity Index 1.6+0.5 1.5+0.5 .48 Positive Symptom Distress Index 2.4+0.5 2.3+0.5 .36 Positive Symptom Total 59.6+12.0 55.9+15.4 .39 Additional HDRS and SCID items*** HDRS Item 1: Depressed mood (1:2:3) 1:10:14 0:10:16 .88 HDRS Item 11: Anxiety somatic (0:1:2:3) 11:4:5:5 8:8:7:3 .49 SCID Item: Suicide (1:2:3) 0:3:22 2:2:22 .67 Values are mean±SD for continuous variables and frequencies for categorical variables. Continuous variables were compared using a two-sample t-test if normally distributed, and a Wilcoxon rank sum test otherwise. Categorical variables were compared using two-tail Fisher’s exact tests. * Missing data on 1 drug subject and 2 placebo subjects. ** Missing data on 4 drug subjects and 4 placebo subjects. *** Only items relevant to later results are shown. 54 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Did the drug and placebo groups demonstrate different treatment outcomes? A graph of the change in the HDRS total over time stratified by treatment group is shown in Figure 3.1. The change in the natural log of HDRS by treatment group after randomization is shown in Figure 3.2. Growth curve models revealed that there was a statistically significant reduction in HDRS scores over time (p= .0006). However, there was no statistically significant interaction with treatment. Both drug and placebo groups exhibited a steep decrease in depressed mood between baseline and 1 week after randomization, followed by a less steep decrease over the following weeks. For the logistic regression analysis, treatment outcome was based upon the final HDRS score using a cut-off of 10. We found that there was not a statistically significant difference between the drug and placebo groups (p >.10) (Table 3.3). For exploration only, we also compared the drug and placebo groups using a more stringent cut-off of 7, as well as a definition of clinical response based upon a 50% reduction from baseline HDRS total score. There were no statistically significant differences in number of responders and non-responders between the drug and placebo groups, regardless of the definition of response used. Did the placebo lead-in reduce the placebo response rate? A total of 53 subjects were enrolled at baseline. Of these subjects, 2 were dropped from the study due to high placebo response at wash-in. However, 10 of the placebo subjects, or 38.5%, still responded to placebo at 8-week follow-up. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.1 Change in HDRS Total over Time 25 DRUG PLACEBO 15 4 4 5 7 4 7 0 W eek Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 3.2 Change in Natural Log of HDRS after Randomization 3.5 48 hours 1 week 2 weeks weeks 8 weeks — DRUG I — PLACEBO C f l pel e K a -J 2.5 1.5 2 1 0.5 0 Ln Time Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.3. Categorical HDRS Measures of Depression by Treatment Groups Definitions of Response Drag Group (n=25 ) N % Placebo Group (n=26) N % P (between groups) Endpoint cut-off of 10 Responders (< 10) 13 52.0 10 38.5 .40 Non-responders (>10) 12 48.0 16 61.5 Endpoint cut-off of 7 Responders (< 7) 8 32.0 6 23.1 .54 Non-responders (>7) 17 68.0 20 76.9 Change between baseline and endpoint Responders (> 50%) 13 52.0 1 1 42.3 .58 Non-responders (< 50%) 12 48.0 15 57.7 Variables were compared using two-tail Fisher’s exact tests. 3.4.2. Growth Curve Models Repeated measurements of BDI, MADRS, and HAM-A scores were recorded over time. As expected, changes in BDI, MADRS, and HAM-A scores over time were statistically significantly associated with changes in HDRS over time for all subjects and by treatment group (all p< .0001). Did baseline severity of depression predict HDRS change over time? Severities of baseline BDI, MADRS, and SCID total scores were not statistically significantly predictive of HDRS change over time, after adjusting for baseline HDRS and treatment. However, even after adjusting for baseline total HDRS as well as treatment, decreased depressed mood on HDRS item 1 was a statistically significant predictor of a decrease in HDRS over time (p=.03) (Table 3.4). Analysis by treatment 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. suggested that the HDRS item for depressed mood was predictive of clinical course in the drug group (p=.03), but not the placebo group (Table 3 .5). Table 3.4. Individual Predictors of HDRS Change in All Subjects HDRS Decrease Over Time 0 SE P 1 Depressed mood (HDRS item 1) -0.18 0.08 .03 i Anxiety somatic (HDRS item 11) -0.09 0.04 .02 t Guilt (SCID item) -0.10 0.05 .05 I Middle insomnia (HDRS item 5) -0.11 0.06 .06 t Prefrontal cordance during lead-in -0.08 0.04 .03 I Age -0.01 0.003 .07 Family history of mood disorder: No -0.15 0.09 .09 t Thoughts of death (SCID item) -0.16 0.09 .08 i Somatization (SCL-90-R) 0.12 0.07 .07 Growth curve models were used to determine which one-variable models were the best predictors of change in HDRS after adjusting for baseline HDRS and treatment. Results are shown for all variables with p<.10. 51 (25 drug, 26 placebo) subjects were used in the analysis. Due to missing data, only 43 (21 drug, 22 placebo) subjects were used in the analysis for SCL-90-R items. Table 3.5. Individual Predictors of HDRS Change by Treatment Groups_____________ HDRS Decrease Over Time Drug Group Placebo Group P (SE) P P (SE) P •I Depressed mood (HDRS item 1) -0.23 (0.11) .03 - - i Anxiety somatic (HDRS item 11) - - -0.10(0.05) .06 t Guilt (SCID item) - - -0.12 (0.06) .06 T Prefrontal cordance during lead-in -0.10(0.05) .07 - - f Thoughts of death (SCID item) - - -0.16(0.09) .08 i Somatization (SCL-90-R) -0.23 (0.09) .02 - - Growth curve models were used to determine which one-variable models were the best predictors of change in HDRS after adjusting for baseline HDRS. Only significant or marginally significant univariate predictors for all subjects were considered. Results are shown for all variables with p<.10. 25 drug and 26 placebo subjects were used in each analysis. Due to missing data, only 21 drug and 22 placebo subjects were used for the analysis of the SCL-90-R somatization subscale. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Did duration of the current episode predict HDRS change over time? Duration of the current episode was not a statistically significant predictor of change in HDRS over time, after adjusting for baseline HDRS and treatment (Table 3.4). Did baseline anxiety predict HDRS change over time? Baseline anxiety, as measured by the HAM-A and subscales of the SCL-90- R, was not a statistically significant predictor of HDRS change over time. However, decreased somatic anxiety on the HDRS (item 11) was associated with a decrease in HDRS over time, after adjusting for baseline HDRS and treatment (p=. 02) (Table 3.4). Analysis by treatment suggested that baseline somatic anxiety affected clinical course in the placebo group (p=.06), but not the drug group (Table 3.5). Did baseline psychotic symptoms predict HDRS change over time? Psychoticism and paranoid ideation, as defined by SCL-90-R subscales, were not statistically significant predictors of change in HDRS over time, after adjusting for baseline HDRS and treatment (Table 3.4). Did baseline melancholic symptoms predict HDRS change over time? Several items on the HDRS and SCID were indicators of melancholic symptoms. Increased guilt on the baseline SCID was a marginally significant predictor of a decrease in HDRS over time, after adjusting for baseline HDRS and 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. treatment (p=.05) (Table 3.4). Analysis by treatment suggested that baseline guilt affected clinical course in the placebo group (p=.06), but not the drug group (Table 3.5). Decreased middle insomnia on the HDRS was a marginally significant predictor of a decrease in HDRS over time, after adjusting for baseline HDRS and treatment (p=.06) (Table 3.4), but did not reach significance for the analysis by treatment group. Did change in clinical symptoms during placebo lead-in predict HDRS change over time? Differences between baseline and follow-up depressed mood (HDRS, BDI, MADRS) and anxiety (HAM-A) measures at wash-in were not statistically significant predictors of HDRS change over time, after adjusting for baseline HDRS and treatment (Table 3.4). Did change in brain activity during placebo lead-in predict HDRS change over time? An increase in prefrontal cordance during placebo lead-in was a statistically significant predictor of a decrease in HDRS over time, after adjusting for baseline HDRS and treatment (p=.03) (Table 3.4). The analysis by treatment suggested that an increase in prefrontal cordance during placebo lead-in was a marginally Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. significant predictor of clinical course in the drug group (p=. 07), but not the placebo group (Table 3.5). Did any additional demographic or clinical risk factors predict HDRS change over time? After adjusting for baseline HDRS and treatment, individuals were marginally significantly more likely to experience a decrease in HDRS over time if they were younger (p=.07), did not have a family history of depression (p=.09), expressed more thoughts about death during the SCID interview (p=.08), and reported less somatization on the SCL-90-R (p=.07) (Table 3.4). Age and family history did not reach significance when analyzed by treatment group. Increased thoughts about death appeared to have a marginal effect on clinical course in the placebo group (p=08), but not the drug group (Table 3.5). Decreased somatization was a statistically significant predictor of decreased HDRS over time in the drug group (p=02), but not the placebo group (Table 3.5). All other demographic and clinical risk factors tested were not statistically significant. What combination of variables best predicted HDRS change over time? After adjusting for baseline HDRS and treatment, the best multi-variable growth curve model indicated that individuals were more likely to experience a decrease in HDRS over time if they were rated with decreased depressed mood on the baseline HDRS (p=.01), were younger (p=.02), expressed more thoughts about 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. death during the SCID interview (p=. 03), and reported less somatization on the SCL- 90-R (p=.08) (Table 3.6). Table 3.6. Multi-variable Model for HDRS Change in All Subjects_____________ HDRS Decrease Over Time p____________SE____________P_ 1 Depressed mood (HDRS item 1) -0.20 0.08 .01 I Age -0.01 0.004 .02 t Thoughts of death (SCID item) -0.18 0.08 .03 I Somatization (SCL-90-R)_________________ -0.10_________ 0.06___________.08 Significant or marginally significant univariate predictors were entered into a growth curve model using alpha=. 10 and backward selection criteria. The best-fitting multiple model was adjusted for baseline HDRS and treatment. The analysis by treatment suggested that depressed mood on the baseline HDRS (p= 04) and somatization on the SCL-90-R (p=.03) were predictive of clinical course in the drug group (Table 3.7). Only one variable appeared to be marginally predictive of clinical course in the placebo group, which was guilt on the baseline SCID (p=06) (Table 3.7). Table 3.7. Multi-variable Model for HDRS Change by Treatment Groups______________ HDRS Decrease Over Time Drug Group Placebo Group P (SE)_______P______ P (SE) P 4 Depressed mood (HDRS item 1) -0.22 (0.11) .04 1 Somatization (SCL-90-R) -0.20 (0.09) .03 t Guilt (SCID item)________________________ - - -0.12(0.06) .06 Significant or marginally significant univariate predictors were entered into a growth curve model using alpha=. 10 and backward selection criteria. The best-fitting multiple model was adjusted for baseline HDRS. Only one variable met criteria for the placebo group. 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. What is the relationship between biological and psychological variables? Tests for interactions between prefrontal cordance during placebo lead-in and all other individual predictors were non-significant. However, after controlling for change in prefrontal cordance during placebo lead-in, effects of depressed mood (HDRS item 1), family history of mood disorders, and somatization (SCL-90-R) were no longer significant (Figure 3.3). This suggested that changes in prefrontal cordance might mediate the effects of these variables. Figure 3.3 Mediation by Prefrontal Cordance Before adjusting for prefrontal cordance Depressed mood A HDRS over time Somatization Family history After adjusting for prefrontal cordance A Prefrontal Cordance Depressed mood A HDRS over time Somatization Family history Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.4.3. Logistic Regression Models Baseline HDRS was marginally statistically significantly predictive of HDRS response in all subjects (p=.07), but not statistically significant by treatment group. Nonetheless, baseline HDRS was controlled for in all analyses. Like the growth curve models, results of the logistic regression analysis suggested that the following variables might be useful predictors of treatment outcome: guilt on the SCID (p=.05), change in prefrontal cordance during placebo lead-in (p=.04), middle insomnia on the HDRS (p=.02), and family history of mood disorders (p=.03) (Table 3.8). Additional predictors identified by logistic regression are described below. Table 3.8. Individual Predictors of HDRS Response for All Subjects______________ HDRS Response OR 95% Cl P Consistent with Growth Curve Models Guilt (SCID item) 2.21 1.00- 4.85 .05 A Prefrontal cordance during lead-in 1.89 1.03- 3.46 .04 Middle insomnia (HDRS item 5) 0.32 0.12- 0.84 .02 Family history of mood disorder: Yes 0.24 0.06- 0.87 .03 New predictors Duration of current episode: >Year 0.06 0.005 --0.61 .02 Late insomnia (HDRS item 6) 0.36 0.15- 0.87 .02 Number of primary relatives 0.62 0.37- 1.06 .08 Logistic regression was used to determine which one-variable models were the best predictors of HDRS response after adjusting for baseline HDRS and treatment. Results are shown for all variables with p< 10. 51 (25 drug, 26 placebo) subjects were used in the analysis. Due to missing data, only 50 (24 drug, 26 placebo) subjects were used for the analysis of family history. Did duration of the current episode predict HDRS response? Subjects were statistically significantly less likely to be HDRS responders if the duration of the current episode was greater than 1 year, after adjusting for 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. baseline HDRS and treatment (p=. 02) (Table 3.8). However, duration of the current episode did not reach significance for the analysis by treatment group. Did baseline melancholic symptoms predict HDRS response? Increased guilt on the baseline SCID was a marginally significant predictor of HDRS response, after adjusting for baseline HDRS and treatment (p=.05) (Table 3.8). Guilt did not reach significance in the analysis by treatment group. Individuals were statistically significantly less likely to be HDRS responders if they had middle or late insomnia on the baseline HDRS, after adjusting for baseline HDRS and treatment (both p=02) (Table 3.8). For the drug group, subjects were marginally significantly less likely to be HDRS responders if they reported middle insomnia (p=.07) (Table 3.9). For the placebo group, subjects were significantly less likely to be HDRS responders if they reported late insomnia (p=.03) (Table 3.9). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3.9. Individual Predictors of HDRS Response by Treatment Groups HDRS Response Drug Group Placebo Group OR P OR P (95% Cl) (95% Cl) Consistent with Growth Curve Models A Prefrontal cordance during lead-in 2.81 .03 - (1.10-7.19) Middle insomnia (HDRS item 5) 0.24 .07 - (0.05-1.10) New predictors Late insomnia (HDRS item 6) - .10 .03 (0.01-0.77) Logistic regression was used to determine which one-variable models were the best predictors of HDRS response after adjusting for baseline HDRS. Only significant or marginally significant univariate predictors for all subjects were considered. Results are shown for all variables with p<10. 25 drug and 26 placebo subjects were used in each analysis. Did any additional demographic or clinical risk factors predict HDRS response? Having a higher number of primary relatives with depression was marginally associated with a decreased likelihood of HDRS response, after adjusting for baseline HDRS and treatment (p=.08) (Table 3 .8). However, number of primary relatives did not reach significance in the analysis by treatment. What combination of variables best predicts HDRS response? After adjusting for baseline HDRS and treatment, the best multiple logistic model suggested that individuals were more likely to be HDRS responders if they reported less late insomnia on the HDRS (p=004), had a duration of the current episode less than 1 year (p=.01), expressed guilt on the baseline SCID (p-.Ol), and if they demonstrated a change in prefrontal cordance during the placebo lead-in (p=.06) 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Table 3.10). Specifically, an increase in prefrontal cordance during placebo lead-in appeared to increase likelihood of HDRS response. Table 3.10. Multi-variable Model for HDRS Response for All Subjects_____ HDRS Response OR_________ 95% Cl__________P _ Late insomnia (HDRS item 6) 0.08 0.01-0.44 .004 Duration of current episode: >Year 0.02 <001-0.38 .01 Guilt (SCID item) 7.07 1.59-31.33 .01 A Prefrontal cordance during lead-in__________ 2.22______ 0.98 - 5.03_________ 06 Significant or marginally significant univariate predictors were entered into a logistic regression model using alpha=. 10 and backward selection criteria. The best-fitting multiple model was adjusted for baseline HDRS and treatment. The analysis by treatment suggested that change in prefrontal cordance during placebo lead-in (p=04) and middle insomnia on the HDRS (p= 08) were predictive of clinical response in the drug group (Table 3.11). Only one variable appeared to be a statistically significant predictor of clinical response in the placebo group, which was late insomnia on the HDRS (p=03) (Table 3.11). Table 3.11. Multi-variable Model for HDRS Response by Treatment Groups_________ HDRS Response Drug Group Placebo Group OR (95% Cl) P OR (95% Cl) P A Prefrontal cordance during lead-in 2.94 (1.04-8.32) .04 - Middle insomnia (HDRS item 5) .20 (.03-1.19) .08 " Late insomnia (HDRS item 6) “ " .097 (.012-.771) .03 Significant or marginally significant univariate predictors were entered into a logistic regression model using alpha- 10 and backward selection criteria. The best-fitting multiple model was adjusted for baseline HDRS. Only one variable met criteria for the placebo group. 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. What is the relationship between biological and psychological variables? There was a marginally significant interaction between change in prefrontal cordance and middle insomnia (p=. 06), but no other interactions between change in prefrontal cordance and other individual predictors. Change in prefrontal cordance also appeared to mediate the effects of family history and number of primary relatives. 3.5. Discussion Several biological and psychological variables were identified as potential predictors of a decrease in depression over time. Subjects in both drug and placebo groups were more likely to demonstrate a decrease in HDRS over time if they had lower ratings of depressed mood on the HDRS at baseline, were younger, expressed more negative thoughts during the SCID interview at baseline, self-reported fewer somatic symptoms at baseline, were rated with less somatic anxiety on the HDRS at baseline, demonstrated a greater change in prefrontal cordance between baseline and wash-in, reported less awakening at night on the HDRS at baseline, and did not have a family history of mood disorders. Results of the logistic regression analysis corroborated findings from the growth curve analysis, and suggested that additional predictors of decreased depression by eight-week follow-up might include late insomnia, duration of the current episode, and number of primary relatives. The drug and placebo groups did not demonstrate statistically significant differences in treatment outcomes. Although the drug group tended to have lower Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. HDRS scores over time and a higher response rate than the placebo group, the placebo group had similar decreases in HDRS scores over time and a relatively high response rate. Despite the exclusion of 2 initial placebo responders during the placebo lead-in, 38.5% of the placebo subjects had HDRS scores less than or equal to 10 at eight-week follow-up. Other clinical trials also have found that elimination of placebo responders during the placebo lead-in has little effect on the placebo response rate (Greenberg & Fisher, 1997). Although the drug and placebo groups demonstrated similar treatment outcomes, results from this study suggested that drug and placebo response might be predicted by different factors. In particular, drug subjects a were more likely to show a decrease in HDRS over time if they had lower ratings of depressed mood on the HDRS at baseline, demonstrated a greater change in prefrontal cordance between baseline and wash-in, and reported fewer somatic symptoms at baseline. Placebo subjects were more likely to show a decrease in FIDRS over time if they were rated with less somatic anxiety on the HDRS at baseline or expressed more negative thoughts during the SCID interview. Future studies could examine whether drug and placebo outcomes are distinguished by factors such as degree of depressed mood, somatization, somatic anxiety, negative cognitions, or changes in prefrontal brain activity. We also examined relationships among biological and psychological variables. The biological variable available was prefrontal cordance, a measure of prefrontal brain activity. We found that change in prefrontal cordance during 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. placebo lead-in appeared to mediate the effects of depressed mood, family history of mood disorders, and somatization. Future studies could examine whether the prefrontal cortex mediates recovery from a major depressive episode. It is possible that genetic or environmental factors associated with family history of MDD may affect treatment outcome through processing in the prefrontal cortex. In addition, the effect of depressed mood and somatization on treatment outcome may be mediated by processing in the prefrontal cortex. Several of our findings are consistent with previous research. In our analysis, increased expression of guilt or thoughts about death on the baseline SCID predicted a decrease in HDRS scores over time. This finding suggests that negative cognitions at baseline predict a favorable response to drugs or placebo, and is consistent with previous studies (Fisher & Greenberg, 1997; Shapiro et al., 1975; Swartzman & Burkell, 1998). Our finding that middle to late insomnia at baseline is predictive of poor treatment outcome is consistent with reports of higher rates of sleep EEG abnormalities in outpatients compared to inpatients with MDD (Association, 1994). It is possible that difficulty sleeping is a behavioral marker of more severe illness that is more difficult to treat. In addition, we found that longer duration of the current episode was associated with failure to respond to treatment, which is consistent with previous research (Brown et al., 1992). We also found that lower somatic anxiety at baseline predicted placebo response (Bialik et al., 1995). However, some of our results appear to be inconsistent with previous research. We did not find that change in clinical symptoms during the placebo lead- 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. in predicted clinical course, in contrast to a previous study that found that improvement predicted response (Quitkin et al., 1998). However, the study by Quitkin and colleagues used the Clinical Global Impression Scale to assess improvement, whereas we examined clinical changes using the HDRS, BDI, MADRS, and HAM-A. Also, other research suggests that psychotic symptoms are associated with a poor prognosis, which we did not find (Joyce & Paykel, 1989; Widiger et al., 1996). It is possible that psychotic symptoms are actually a marker of depression severity, and that controlling for baseline HDRS scores reduced our ability to detect an association between psychotic symptoms and treatment outcome (Kocsis, 1990). With regard to brain changes, we found that increased prefrontal cordance during placebo lead-in was a significant predictor of a decrease in HDRS over time for all subjects. Since all of the subjects were on placebo during the placebo lead-in, this finding is consistent with our previous research, which indicated that response to placebo treatment is predicted by an increase in prefrontal cordance (Leuchter et al., 2002). The analysis by treatment suggested that an increase in prefrontal cordance during placebo lead-in was a predictor of drug response but not placebo response. The combined results of this analysis and our previous research suggests that antidepressant responders experience an initial increase in prefrontal cordance prior to a decrease at 48 hours and 1 week after randomization (Cook et al., in press). By contrast, increase in prefrontal cordance may not be predictive of placebo response unless it occurs 1 week after randomization (Leuchter et al., 2002). 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. One of the primary strengths of this study was that repeated measures of HDRS scores were analyzed using growth curve models. An advantage of growth curve modeling is that it is more powerful than logistic regression since several data points, rather than a single endpoint, are used to provide more stable estimates of clinical response. The use of repeated measures to define the outcome is particularly important for psychiatric research, where the outcome of interest is often a psychological variable with substantial intrasubject variability (Kraemer, 1991; Kraemer & Thiemann, 1989). Another advantage of growth curve modeling is that it does not require the arbitrary classification of subjects into responders and non responders. Instead, the degree of clinical response is measured by the degree of change in the slope (Jaccard & Wan, 1993). In addition, we were able to model time and subject differences at baseline as random variables in the growth curve models. Another strength of this study was that the relationship between a biological factor, prefrontal cordance, and several psychological variables was examined for both moderating and mediating effects. Further studies of relationships between biological and psychological variables will help elucidate mechanisms underlying recovery from MDD. Nevertheless, some limitations should be considered when evaluating the results of this study. Since this was an exploratory analysis, the potential for a high rate of Type I errors limits our conclusions. Also, due to the small sample size, our null findings might represent insufficient power, measurement error, or a true lack of association. The analysis by treatment should be considered with particular caution 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. since the sample size was reduced to 25 subjects in the drug group and 26 subjects in the placebo group. Furthermore, the drug group was a mixture of 13 subjects taking fluoxetine and 12 subjects taking venlafaxine. It is possible that factors predictive of treatment outcome vary by drug type, but we were unable to address this question due to insufficient power. In addition, we were unable to examine the influence of noncompliance or unblinding due to lack of information. Both noncompliance and unblinding have the potential to confound the relationship between treatment and depressed mood outcomes (Kocsis, 1990; Swartzman & Burkell, 1998). Finally, it is possible that the clinical trials examined in this analysis were of insufficient duration to detect clear distinctions between the differential effects of drug versus placebo. In conclusion, the results of this exploratory analysis should be considered preliminary hypotheses to be tested in future experimental studies. Further research that examines biopsychosocial predictors of treatment outcome will lead to a better understanding of how to appropriately treat individual patients with MDD. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3.6. References Abou-Saleh, M., & Coppen, A. (1983). Classification of depression and response to antidepressive therapies. British Journal o f Psychiatry, 143, 601-603. Andreasen, N., & Black, D. (1995). Introductory Textbook o f Psychiatry (2nd ed.). Washington, DC: American Psychiatric Press, Inc. Antonuccio, D., Danton, W., & DeNelsky, G. (1995). Psychotherapy versus medication for depression: Challenging the conventional wisdom with data. Professional Psychology: Research and Practice, 26, 574-585. Association, A. P. (1994). Diagnostic and Statistical Manual o f Mental Disorders (Fourth (DSM-IV) ed.). Washington, DC: Author. Beck, A., Rush, A., Shaw, B., & Emery, G. (1979). Cognitive therapy o f depression. New York: Guilford Press. Beck, A., Ward, C., Mendelsohn, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives o f General Psychiatry, 4, 561-571. Bialik, R. J., Ravindran, A. V., Bakish, D., & Lapierre, Y. D. (1995). A comparison of placebo responders and nonresponders in subgroups of depressive disorder. Journal o f Psychiatry & Neuroscience, 20(4), 265-70. Brown, W., Johnson, M., & Chen, M. (1992). Clinical features of depressed patients who do and do not improve with placebo. Psychiatry Research, 41, 203-214. Clarke, D. M. (1998). Psychological factors in illness and recovery. New Zealand Medical Journal, 777(1076), 410-2. Cook, I., Leuchter, A., Witte, E., Stubbeman, W., Abrams, M., & Rosenberg, S. (in press). Early changes in prefrontal activity characterize clinical responders to antidepressants. . Derogatis, L. (1977). Administration, Scoring and Procedures Manual fo r the SCL- 90-R. Baltimore: Clinical Psychometrics Research. Evans, F. (1985). Expectancy, therapeutic instructions, and the placebo response. In L. White, B. Tursky, & G. Schwartz (Eds.), Placebo: Theory, Research, and Mechanisms (pp. 215-228). New York: Guilford Press. 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fisher, S., & Greenberg, R. (1997). The curse of the placebo: fanciful pursuit of a pure biological therapy. In S. Fisher & R. Greenberg (Eds.), From Placebo to Panacea: Putting Psychiatric Drugs to the Test (pp. 3-56). New York: John Wiley & Sons, Inc. Goodwin, F. (1993). Predictors of antidepressant response. Bulletin o f the Menninger Clinic, 57(2), 146-160. Greenberg, R., & Fisher, S. (1997). Mood-mending medicines: probing drug, psychotherapy, and placebo solutions. In S. Fisher & R. Greenberg (Eds.), From Placebo to Panacea: Putting Psychiatric Drugs to the Test (pp. 115- 172). New York: John Wiley & Sons, Inc. Hamilton, M. (1959). The assessment of anxiety states by rating. Brit. J. o f Med. Psychol., 32, 50-59. Hamilton, M. (1967). Development of a rating scale for primary depressive illness. British Journal o f Social and Clinical Psychology, 6, 278-296. Harrington, E. (Ed.). (1997). The Placebo Effect: An Interdisciplinary Exploration. Cambridge: Harvard University Press. Jaccard, J., & Wan, C. (1993). Statistical analysis of temporal data with many observations: issues for behavioral medicine data. Annals o f Behavioral Medicine, 75(1), 41-50. Joyce, P., & Paykel, E. (1989). Predictors of drug response in depression. Archives o f General Psychiatry, 46, 88-89. Kocsis, J. (1990). New issues in the prediction of antidepressant. Psychopharmacology Bulletin, 26(1), 49-53. Kraemer, H. C. (1991). To increase power in randomized clinical trials without increasing sample size. Psychopharmacology Bulletin, 27(3), 217-24. Kraemer, H. C., & Thiemann, S. (1989). A strategy to use soft data effectively in randomized controlled clinical trials. Journal o f Consulting & Clinical Psychology, 57(1), 148-54. Leuchter, A., Cook, I., Uijtdehaage, S., Dunkin, J., Lufkin, R., Anderson-Hanley, C., Abrams, M., Rosenberg-Thompson, S., O'Hara, R., Simon, S., Osato, S., & Babaie, A. (1997). Brain structure and function and the outcomes of treatment for depression. Journal o f Clinical Psychiatry, 55(16), 22-31. 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Leuchter, A., Cook, I., Witte, E., & Abrams, M. Changes in brain function of depressed subjects during treatment with placebo. American Journal o f Psychiatry, 759(1), 122-129. Marusic, A., Gudjonsson, G. H., Eysenck, H. J., & Stare, R. (1999). Biological and psychosocial risk factors in ischaemic heart disease: Empirical findings and a biopsychosocial model. Personality and Individual Differences, 26(2), 285- 304. Moerman, D. E., & Jonas, W. B. (2000). Toward a research agenda on placebo. Advances in Mind-Body Medicine, 76(1), 33-46. Montgomery, S., Asberg, M., Traskman, L., & Montgomery, D. (1978). Cross cultural studies on the use of the CPRS in English and Swedish Depressed Patients. ActaPsychiat. Scand, Suppl. 271, 33-37. Nierenberg, A. A., & Wright, E. C. (1999). Evolution of remission as the new standard in the treatment of depression. Journal o f Clinical Psychiatry, 60(Suppl 22), 7-11. Paykel, E., Hollyman, J., Freeling, P., & Sedgwick, P. (1988). Predictors of therapeutic benefit from amitriptyline in mild depression: a general practice placebo-controlled trial. Journal o f Affective Disorders, 14, 83-95. Quitkin, F., McGraph, P., Rabkin, J., Stewart, J., Harrison, W., Ross, D., Tricamo, E., Gleiss, J., Markowitz, J., & Klein, D. (1991). Different types of placebo response in patients receiving antidepressants. American Journal o f Psychiatry, 148, 197-203. Quitkin, F., McGraph, P., Stewart, J., Ocepek-Welikson, K., Taylor, B., Nunes, E., Delivannides, D., Agosti, V., Donovan, S., Ross, D., Petkova, E., & Klein, D. (1998). Placebo run-in period in studies of depressive disorders. Clinical, heuristic and research implications. British Journal o f Psychiatry, 173(9), 242-248. Quitkin, F. M. (1999). Placebos, drug effects, and study design: a clinician's guide. American Journal o f Psychiatry, 156(6), 829-36. Shapiro, A., & Shapiro, E. (1997). The placebo: is it much ado about nothing? In E Harrington (Ed.), The Placebo Effect: An Interdisciplinary Exploration (pp. 12-36). Cambridge: Harvard University Press. Shapiro, A., Struening, E., Barten, H., & Shapiro, E. (1975). Correlates of placebo reaction in an outpatient population. Psychological Medicine, 5, 389-396. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Swartzman, L. C., & Burkell, J. (1998). Expectations and the placebo effect in clinical drug trials: why we should not turn a blind eye to unblinding, and other cautionary notes. Clinical Pharmacology & Therapeutics, 64(1), 1-7. Thase, M., Entsuah, A., & Rudolph, R. (2001). Remission rates during treatment with venlafaxine or selective serotonin reuptake inhibitors. British Journal o f Psychiatry, 178, 234-241. Tueting, P. (1991). Psychophysiological predictors of drug treatment response. Psychiatric Medicine, 9(1), 145-161. Widiger, T., Frances, A., Pincus, H., Ross, R., First, M., & Davis, W. (Eds ). (1996). DSM-IVSourcebook (Vol. 2). Washington, DC: American Psychiatric Association. Wilhelm, K., Parker, G., & Hadzi-Pavlovic, D. (1997). Fifteen years on: evolving ideas in researching sex differences in depression. Psychological Medicine, 27(4), 875-83. Williams, J., Gibbon, M., First, M., Spitzer, R., Davies, M., Borus, J., Howes, M., Kane, J., Pope, H., Rounsaville, B., & Wittchen, H. (1992). The Structured clinical Interview for DSM-III-R (SCID). II. Multisite test-retest reliability. Archives o f General Psychiatry, 49, 630-636. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4. Grant proposal: Mood and physiologic reactivity in major depression 4.1. Abstract Previous research has shown that healthy individuals have characteristic physiologic responses to acute psychological stressors. It is not known whether patients with major depressive disorder (MDD) have similar responses, or whether physiologic reactivity varies with mood in patients with major depression. We will assess changes in cardiovascular and immune measures in response to acute psychological stressors in patients with MDD at two time points: prior to beginning treatment (screening) and after 8 weeks of pharmacological treatment (follow-up). Similar data will be collected in normal controls prior to beginning treatment (screening) and after 4 weeks of pharmacological treatment (follow-up). Depressed patients (N=26) will be recruited from a randomized-controlled clinical trial of venlafaxine. Normal controls (N=26) will be recruited from a parallel randomized clinical trial of venlafaxine in healthy subjects, and will be matched to clinically depressed patients on age, gender, and education. Although all patients will have comparable levels of severely depressed mood at screening, we expect that at follow- up some of the patients will demonstrate a decrease in depressed mood. Therefore, we will be able to use a within-subjects design to determine whether a change in mood is associated with a change in physiologic reactivity. The primary objective of this pilot study will be to test the hypotheses that: 1) physiologic reactivity will be higher in depressed patients than in normal controls at screening; 2) physiologic reactivity at screening will predict depressed mood at follow-up in depressed 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. patients; and 3) physiologic reactivity will not change in normal controls, but will decrease as symptoms of depression decrease in depressed patients. All analyses of follow-up data will be stratified by treatment type (venlafaxine or placebo). This study will provide data about individual differences in the relationships between cardiovascular, immune, and psychosocial variables. The study of relationships between biological and psychosocial factors may help develop an integrated theory of MDD, as well as hypotheses about mechanisms by which depressed mood may influence health outcomes. 4.2. Specific Aims Previous research has shown that healthy individuals have characteristic cardiovascular and immune responses to acute psychological stressors. It is not known whether patients with major depressive disorder (MDD) have similar cardiovascular and immune responses to acute psychological stressors, or whether physiologic reactivity varies with mood in patients with major depression. In this pilot study, we propose to assess changes in cardiovascular and immune measures in response to acute psychological stressors in patients with MDD at two time points: prior to beginning treatment (screening) and after 8 weeks of treatment (follow-up). Similar data will be collected in normal controls prior to beginning treatment (screening) and after 4 weeks of pharmacological treatment (follow-up). The primary objective of this study will be to test the hypotheses that: 1) physiologic reactivity will be higher in depressed patients than in normal controls; Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2) physiologic reactivity at screening will predict depressed mood at follow-up in depressed patients; and 3) physiologic reactivity will not change in normal controls, but will decrease as symptoms of depression decrease in depressed patients. This study will provide data about individual differences in the relationships between cardiovascular, immune, and psychosocial variables. The study of relationships between biological and psychosocial factors may help develop an integrated theory of MDD, as well as hypotheses about mechanisms by which depressed mood may influence health outcomes. 4.3. Background and Significance The Problem: Major Depressive Disorder Recent reviews of the literature suggest that the current lifetime prevalence of depression is between 5% and 30% (Andreasen & Black, 1995; Association, 1994; Musselman, Evans, & Nemeroff, 1998; Simon et al., 1995). Lifetime rates of depression are consistently higher in women than in men, with ranges from 7-38% for women and 3-29% for men (Association, 1994; Wilhelm, Parker, & Hadzi- Pavlovic, 1997). Clinical depression is a major cause of disability worldwide, and the leading cause of disability in the U.S. (Antonuccio, Danton, & DeNelsky, 1995; Clarke, 1998). Complications associated with depressive illness include poor performance at work or school, marital discord, and substance abuse (Andreasen & Black, 1995; Judd et al., 2000). The most serious complication is suicide. About 15% of hospitalized depressed patients die by suicide (Andreasen & Black, 1995; 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Association, 1994). Death rates due to all causes are about 4 times higher in individuals with MDD over age 55 (Association, 1994). Thus, MDD is an important public health concern. Currently, the diagnosis of clinical depression is based upon the severity and duration of psychological and behavioral symptoms. Extensive research suggests that a variety of biological markers may be useful in the diagnosis of clinical depression. For example, MDD is associated with neurotransmitter abnormalities such as decreased norepinephrine and/or decreased serotonin (Andreasen & Black, 1995). Another common finding is that depressed patients exhibit overactivity of the hypothalamic pituitary adrenal (HPA) axis (e.g. high cortisol levels) (Krishnan, Gadde, & Kim, 1998; McDaniel, Musselman, Porter, Reed, & Nemeroff, 1995). The high prevalence of depression in women, postpartum depression, and premenstrual mood changes suggest that changes in sex hormones influence mood, possibly via effects on serotonin regulation or cortisol levels (Legato, 1997). In addition, recent research provides evidence for symptom dependent decrease in natural killer (NK) cell activity, and other possible alterations in the immune system of depressed patients (Herbert & Cohen, 1993; Irwin, 1995; Miller, 1998; Ravindran, Griffiths, Merali, & Anisman, 1998). Thus, clinical depression is associated with a number of biological disturbances. However, to date there is no biological characteristic with sufficient sensitivity and specificity to aid in the diagnosis of depression. The fundamental problem is our lack of knowledge about relationships between biological and psychosocial factors in patients with MDD. 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A corollary of our limited understanding of relationships between biological and psychosocial factors in MDD is our limited understanding of individual differences in response to antidepressants. In general, about a third of patients in clinical trials respond to antidepressants, a third respond to placebo, and a third do not respond to antidepressants or placebos (Greenberg & Fisher, 1997). However, only a few factors are known to predict drug response. Recent reviews suggest that melancholic symptoms such as psychomotor retardation, early morning awakening, and weight loss may be predictive of response to antidepressants, particularly TCAs (Goodwin, 1993; Joyce & Paykel, 1989; Kocsis, 1990; Tueting, 1991). In addition, individuals with moderately severe depressed mood may be more likely to respond to either treatment, whereas individuals with very low or very high severity scores may be less likely to respond (Brown, Johnson, & Chen, 1992; Goodwin, 1993; Joyce & Paykel, 1989). However, this does little to guide physicians in the appropriate treatment of individual patients with depression. Thus, more knowledge about relationships between biological and psychosocial factors in MDD is needed in order to advance our understanding of individual differences in treatment outcomes. An additional consequence of our failure to understand relationships between biological and psychosocial factors in MDD is that we are unable to explain the association between depression and somatic illnesses. A history of depression is associated with an increased risk for a number of somatic illnesses (e.g. hypertension, pulmonary embolism, stroke, muscular sclerosis, renal disease) (Stoudemire, 1995). Most recently there has been a lot of attention to the association 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. between depression and cardiovascular disease (CVD), where depression appears to influence both etiology and prognosis (Kop, 1997; Lavoie & Fleet, 2000; Musselman et al., 1998; Stoudemire, 1995). The prevalence of MDD is about three-times higher in patients with coronary artery disease compared to the general population (Rozanski, Blumenthal, & Kaplan, 1999). In addition, the prevalence of cardiac events is about four-times higher in individuals with a history of MDD compared to individuals with no history (Lane, Carroll, & Lip, 1999). Some studies also provide evidence for a gradient between the magnitude of depression and risk of future cardiac events (Rozanski et al., 1999). The association between depression and heart disease appears to be independent of gender, although more studies of men have been reported (Eaker, 1998; Schwartzman & Glaus, 2000). Elucidation of the mechanisms explaining the association between depression and CVD remains a topic for future research. Physiologic Reactivity A paradigm for studying relationships between biological and psychosocial factors is to examine physiological reactivity in response to acute psychological stressors in patients with MDD. A fairly extensive literature exists on the study of cardiovascular and immune reactivity in healthy non-clinical populations. However, to our knowledge, very few studies have examined physiologic reactivity in patients with MDD. This is surprising given the similarity between the physiology of Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. depression and the physiological changes observed during reactivity studies (Delehanty, Dimsdale, & Mills, 1991; Dimsdale, 1988). Cardiovascular reactivity (CR) refers to cardiovascular responses to acute physical or psychological stressors that are compared to baseline levels. Cardiovascular responses include changes in blood pressure (BP), heart rate (HR), stroke volume, or total peripheral resistance, where BP and HR are the most commonly used measures (Newman, McGarvey, & Steele, 1999). Psychological stressors that have been used to study individual differences in CR include mental arithmetic, Stroop color-word tests, public speaking, role-play stressors, and video games (Newman et al., 1999; Sherwood & Turner, 1992). Research suggests that acute stress is associated with activation of the sympathetic adrenal medullary (SAM) system. It is hypothesized that appraisal of a stressor activates the sympathetic nervous system via the limbic system and hypothalamus, and then the sympathetic nervous system activates the adrenal medulla which secretes catecholamines (Tewes, 1999). A number of studies of CR to acute psychological stress show that changes in BP and HR correlate with changes in plasma or urinary catecholamines (Benschop & Schedlowski, 1999; Cohen, 1986; Jemerin & Boyce, 1990; Uchino, Cacioppo, Malarkey, & Glaser, 1995). However, these studies provide only indirect evidence. Direct evidence for the role of SAM system and CR is provided by studies using beta-blockers. A review of 59 studies examining the effects of beta-blockers on CR to psychological stressor found that all classifications of beta-blockers reduced HR reactivity, but not BP reactivity (Mills & Dimsdale, 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1991). Additional research suggests that acute stress is associated with activation of the hypothalamic-pituitary axis (HPA) (Roy, Steptoe, & Kirschbaum, 1998; Uchino et al., 1995). Specifically, perception of a stressor stimulates the hypothalamus to release CRF, which causes the pituitary to release ACTH and stimulate the adrenal cortex to secrete glucocorticoids such as cortisol (Tewes, 1999). Thus, research in healthy subjects suggests that acute psychological stressors increase release of catecholamines that mediate HR reactivity, and that acute psychological stressors also are associated with increased activity of the HPA axis. Another line of research suggests that CR in response to acute psychological stress is associated with specific immune changes. Results from this literature indicate that brief stressful laboratory tasks are associated with increases in circulatory lymphocytes, especially NK cells and suppressor/cytotoxic T cells (Benschop & Schedlowski, 1999; Cacioppo et al., 1998; Naliboffet al., 1991; Naliboff et al., 1995; Pike et al., 1997; Schedlowski et al., 1993). Studies also indicate an increase in NK cell cytotoxicity within a few minutes following acute stressors (Naliboff et al., 1991; Schedlowski et al., 1993; Uchino et al., 1995), although this increase may be attributable to the increase in the number of NK cells (Cacioppo et al., 1998). Approximately one-hour post-stressor NK cell cytotoxicity may decrease below baseline levels before returning to normal (Brosschot et al., 1992; Schedlowski et al., 1993). Of particular interest is evidence that individuals who show the highest sympathetic activation also show the largest immune changes (Cohen & Herbert, 1996; Manuck, Cohen, Rabin, Muldoon, & Bachen, 1991; 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Uchino et al., 1995). Furthermore, beta-blockers can prevent stress-induced increases in NK cell numbers and activity (Benschop & Schedlowski, 1999; Cohen & Herbert, 1996). In addition, injection of epinephrine and norepinephrine is associated with increased NK cell numbers (Benschop & Schedlowski, 1999). Studies do not suggest that cortisol or opioids are involved in acute stress-induced immune changes (Cohen & Herbert, 1996). Thus, sympathetic activation may be the primary mechanism triggering immune reactivity in response to acute psychological stress. In summary, acute psychological stressors increase release of catecholamines, which mediate HR reactivity and affect immune function (Cacioppo, 1994). Thus, healthy individuals have characteristic cardiovascular and immune responses to acute psychological stressors, which we will refer to as “physiologic reactivity.” Significance of Studying Physiologic Reactivity in Patients with MDD Prospective studies indicate that increased physiologic reactivity is associated with increased risk of CVD (Jemerin & Boyce, 1990; Wilson et al., 1999). For example, a longitudinal study of 83 Samoan adolescents found that systolic BP reactivity predicted higher resting systolic BP 3-4 years later (Newman et al., 1999). Another longitudinal study of 504 male undergraduates found that systolic, diastolic and HR reactivity predicted laboratory and ambulatory BP 10-15 years later, after controlling for standard risk factors, baseline BP, and parental history of 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hypertension (Light, Dolan, Davis, & Sherwood, 1992). Physiologic reactivity is associated with SAM hyperactivity and may lead to CVD by causing hypercholesterolemia, hypertriglyceridemia, hypertension, vascular injury, or changes in hemodynamic factors (Musselman et al., 1998). The study of the relationship between physiologic reactivity and depressed mood in patients with MDD may help identify biological changes associated with depressed mood, and suggest a mechanism that may explain the association between depression and CVD. Several studies have examined beat-to-beat variability in HR in patients with MDD. A recent review of these studies indicated that patients with mood disorders have decreased HR variability, which is a known predictor of sudden death and ventricular arrhythmias (Gorman & Sloan, 2000). Reduction in HR variability is thought to be due to decreased parasympathetic input and unopposed sympathetic input to the heart. Finding that patients with MDD have higher CR compared to normal controls would be consistent with the literature on HR variability in MDD, and would provide corroborating evidence of sympathetic hyperactivity. To date, only a few studies have examined physiologic reactivity to psychological stressors in patients with MDD. One study examined physiologic reactivity to a 1-minute mental arithmetic task, and reported that depressed patients had increased sympathetic skin response but no change in HR or BP (Guinjoan, Bemabo, & Cardinali, 1995). Another study examined immune changes following a mathematical challenge and found that the extent of increase in circulating NK cells was comparable in depressive, dysthymic and control subjects, but that the greatest 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. rise in NK cells was exhibited by the patients with most severe depression scores (Ravindran, Griffeths, Merali, & Anisman, 1996). Both studies used only one task to induce psychological stress, which is not as reliable as using multiple tasks, and the duration of the mental arithmetic task in the study by Guinjoan and colleagues was only 1-minute. In an alternative approach, a few researchers have studied patients with cardiovascular disease to determine whether physiologic reactivity varies with depressed mood. These studies have reported strong associations between depressed mood and CR in patients with cardiovascular disease (Carney et al., 1999; Sheffield et al., 1998; Waked & Jutai, 1990). Thus, very few studies have examined physiologic reactivity in response to acute psychological stressors in depressed patients, and few if any studies have followed patients over time to determine whether changes in depressed mood are associated with changes in physiologic reactivity. By using a within-subjects design, we will be able to control for individual differences and capitalize on the change in mood that occurs during a clinical trial in a percentage of both antidepressant and placebo-treated patients. A similar experiment could be done in which static measures of HR, BP and immune variables are taken at screening and eight-week follow-up. However, by analyzing physiologic reactivity instead of static measures of physiological variables, we will be evaluating the dynamic functioning of the cardiovascular and immune systems. Studying the dynamic functioning of the cardiovascular and immune systems has the potential to provide meaningful information about how depressed patients respond to psychological stressors. In 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. addition, studying the dynamic functioning of the cardiovascular and immune systems is likely to provide meaningful information about the mechanisms underlying the association between depression and CVD. Thus, by exploring the relationship between biological and psychosocial processes in depressed patients and normal controls, our study will help increase knowledge about the fundamental science of depression, which may lead ultimately to better diagnosis, treatment and the prevention of complications associated with MDD. 4.4. Research Design and Methods Subjects Subjects for this study will be 26 outpatients with MDD, as well as 26 normal controls matched to age, gender, and education. All subjects will be recruited via newspaper ads and posters for participation in randomized clinical trials of venlafaxine at the University of California at Los Angeles (UCLA). Advertisements targeting normal controls will not indicate that the drug being tested is an antidepressant. Inclusion criteria: All subjects will be between the ages of 18 and 65 and in good health. All patients will meet DSM-IV criteria for major depression. In addition, patients must have a Hamilton Depression Rating Scale (HDRS) rating score >16 (with item #1 >2), and must meet criteria at recruitment and after the one- week single blind placebo wash-in. All normal controls must fail to meet DSM-IV criteria for major depression, and must have a HDRS rating score <16. 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Exclusion criteria: All subjects will be normotensive without a history of serious physical illnesses. Patients with MDD must meet DSM-IV axis I criteria for a major depressive episode, and normal controls must fail to meet the same criteria on the basis of a Structured Clinical Interview for DSM-IV (SCID). We will exclude subjects who meet criteria for axis I diagnoses other than MDD, as well as those meeting criteria for cluster A or B axis II diagnoses. Subjects with a current or past suicidal ideation or suicide attempts will be excluded, in addition to subjects who are pregnant or lactating, and subjects who exercise more than 10 hours per week. Concomitant treatment: Permitted: subjects may receive medications that are without significant CNS activity or cardiovascular activity. Prohibited: subjects may not receive sedative-hypnotics, other antidepressants, or other medications with significant CNS or cardiovascular activity. Design The UCLA Pharmacy will prepare matching capsules containing either venlafaxine 37.5 mg or placebo. After a one-week placebo lead-in, subjects will be randomly assigned to receive one capsule of either venlafaxine or placebo, with the dosage increase every two days until subjects receive four capsules daily (subjects will achieve a dose 150 mg. of venlafaxine after 7 days). The first dose will be administered in the morning, with subsequent capsules added on a b i d. schedule. We will assess physiologic responses to acute psychological stressors in patients with MDD at screening and follow-up. The screening assessment will occur 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. prior to the 1-week single-blind placebo lead in. The follow-up assessment will occur during the 8th week of random assignment to either venlafaxine or placebo. The normal control protocol will consist of a screening assessment that will occur prior to a 1-week single-blind placebo lead in, followed by 4 weeks of randomization. It will not be possible to randomize normal controls for longer than 4 weeks due to ethical considerations. Although depressed patients and normal controls will be followed for different amounts of time, our analysis will focus on a comparison of patients with major depression and matched controls at screening. Follow-up data will be used to examine whether reactivity appears to be a stable trait, with the limitation that normal controls were not followed for as long as depressed patients. All analyses of follow-up data will be stratified by treatment type (venlafaxine or placebo). Procedure Subjects will be asked to not exercise and to not ingest antiinflammatory agents, antihistamines, or alcohol 24 hours before testing. All subjects will receive the same standardized lunch one hour prior to testing. Subjects will be asked to not smoke for at least 2 hours prior to testing, and to not eat or drink any caffeine for at least 1 hour prior to testing. Immune data will not be analyzed for subjects with acute illnesses occurring within the last 2 weeks. At the beginning of the test session, a properly sized occluding cuff will be positioned over the brachial artery of the dominant arm, an intravenous (IV) catheter 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. will be placed in an antecubital vein of the opposite arm, and a HR monitor will be placed around the subject’s waist. After instrumentation, the subject will complete questionnaires regarding exercise and social support, and spend the remainder of 10 minutes resting in the upright position in order to allow adaptation. During the pre stressor baseline time period (SO), the subject will continue sitting upright and relaxing for 6 minutes. Resting BP readings will be taken at 2 minutes and 6 minutes, heart rate measures will be averaged over the final 5 minutes of SO, and a 15 ml. blood sample will be collected at the end of SO for immune assays. Instructions will be given for the Stroop Color-Word Test and Mental Arithmetic. Both tasks will be 6-minutes in duration and will immediately follow one another (SI and S2). The order of the tasks will be fixed with the Stroop always preceding the Mental Arithmetic. Since an important feature of this study is the use of a within-subjects design, we will not counterbalance the tasks across subjects because we do not want the counterbalancing to confound our examination of individual differences. BP readings will be taken at 2 minutes and 6 minutes into each task, and HR measures will be averaged over the final 5 minutes of each task. A 15 ml. blood sample will be collected for immune assays immediately after SI and S2. Subjects will also complete the 10-item Positive and Negative Affect Schedule (PANAS-A) at the end of SO , SI, and S2 (Watson, Clark, & Tellegen, 1988). The remaining questionnaires will be completed after S2 and the removal of the occluding cuff, IV, and HR monitor. 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A summary of the procedure is shown in Table 4.1. The same procedure will be repeated at follow-up. Table 4.1. Procedure for Reactivity Testing ~10 min. -» ■ 10 min. -» 6 min. -» 6 min. -> • 6 min. -» End Set-Up Adaptation Baseline (SO ) Stressor 1 (SI) Stressor 2 (S2) Instrumentation Questionnaires Resting upright Stroop Test Math Test Stroop Color-Word Test (Waldstein. Bachen, & Manuck, 1997) Subjects will perform a computerized Stroop Color-Word Test for 6 minutes at screening and follow-up. Words for four different colors (red, green, blue, yellow) will be displayed in the center of the computer screen in a color incongruent with the printed word (e.g., blue printed in red). Subjects will be asked to choose the color name corresponding to the color of the word printed in the center of the screen. Possible answer choices will be printed at the bottom of the screen in incongruent colors. The task will be titrated to the subject’s level of performance, becoming faster with three consecutively correct answers and slower with two consecutive errors. Mental Arithmetic Subjects will perform a computerized Mental Arithmetic test for 6 minutes at screening and follow-up after the Stroop Color-Word Test. A series of addition and subtraction problems will be presented in the center of the computer screen. For each problem, the subject will be asked to indicate “yes” or “no” depending upon whether they believe the answer given is correct or not. Like the Stroop Color-Word 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Test, this task will be titrated to the subject’s level of performance, becoming faster and more difficult (up to three digit problems) with three consecutively correct answers, and slower and less difficult (down to one digit problems) with two consecutive errors. Assessment Instruments We will inquire about risk factors, prior episodes of depression, and personality at screening only. We will inquire about menstrual history at follow-up only. All other measures listed below will be obtained at screening and again at follow-up. The order of measurements will be consistent across subjects. Demographic Measures a. Subjects will complete a self-report form including information about date of birth, gender, ethnicity, marital status, years of education, employment status, and family history of mood disorders. b. Menstrual History. Mood can vary with the menstrual cycle, and cycle phase may influence physiologic measures (Legato, 1997). Therefore, subjects will be asked questions about cycle phase, menstrual history, age at menarche, menstrual regularity, and age at menopause. Questions are taken from the same instrument that was developed at the University of Southern California for Dr. Giske Ursin's Women's LIFE study (IRB protocols number 981007, 986011 and 986047). 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Biological Measures c. Cardiovascular. An automatic Omron blood pressure monitor will be used to obtain systolic and diastolic BP readings. For each time period (SO, SI, S2), a reading will be taken at 2 minutes and 6 minutes. Both readings will be averaged to obtain mean BP for SO , SI, and S2. In addition, a Polar HR monitor will be used for continuous measurement of cardiovascular activity. We will average continuous measures of HR over 5 minute intervals to obtain mean measures at SO, SI, and S2. Aggregation of HR over repeated measures has been shown to increase reliability (Cacioppo, 1994). d. Immune. Peripheral blood samples will be obtained at the end of SO, SI, and S2, both at screening and follow-up. Blood samples will be taken to the Clinical Immunological Research Laboratory for Interdisciplinary Research in Immunology and Disease (CIRID) at UCLA for immunological analysis. Standard lymphocyte subset analyses and measurements of markers of immune activation will be performed by the CIRID laboratory, as described previously (Futterman, Kemeny, Shapiro, & Fahey, 1994). Briefly, immune assessments will include a) white blood cell (WBC) differentials, b) WBC phenotypic assessments, and c) immunofunctional testing of lymphocyte proliferative response. WBC differential counts on whole blood will be performed on a Coulter JT counter to obtain total counts for lymphocytes, monocytes and granulocytes. Cell-surface phenotype analyses will be carried out by flow cytometry. Data will be analyzed to obtain the percentage and absolute numbers of lymphocytes with specific phenotypic antigens including NK 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. cells (CD 16 and CD56), large granulocytes (CD57), T cells (CD3), and the major T- cell subsets (CD4 helper/inducer and CD8 suppressor/cytotoxic). Whole blood will be centrifuged to obtain the peripheral blood mononuclear cells and lymphocyte response to the mitogen phytohemagglutinin (PHA) will be determined. Excess plasma samples will be frozen for the consideration of additional factors in future studies. Psychological Measures e. Depression. Patients will be diagnosed as experiencing a current major depressive episode based upon the number of symptoms present during the SCID interview at screening. Number of prior episodes, age at first episode, and duration of the current episode will be obtained from the SCID used to diagnose subjects. The SCID questions about major depression will be repeated at follow-up in order to determine if patients still meet criteria for a major depressive episode. The SCID is a standardized interview used to diagnose current and past psychiatric disorders, and has high inter-rater and re-test reliability (Williams et al., 1992). The Hamilton Depression Rating Scale (HDRS) will be administered by a trained interviewer at screening and follow-up (Hamilton, 1967). The HDRS is widely used in clinical trials, but has some weaknesses since it is strongly influenced by somatic symptoms and has low internal construct validity. Therefore, we will do a secondary analysis using the Beck Depression Inventory (BDI) (Beck, Rush, Shaw, & Emery, 1979; Beck, Ward, Mendelsohn, Mock, & Erbaugh, 1961). The BDI is widely used in 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. clinical as well as general population studies, has high internal consistency and retest reliability, and is sensitive to changes in depressed mood (Beck, Steer, & Garbin, 1988). f. Anxiety. Anxiety is known to influence blood pressure measures (Jonas, Franks, & Ingram, 1997; Shapiro et al., 1996), and is frequently associated with depression (Barlow & Campbell, 2000; Widiger et al., 1996). In order to examine the influence of anxiety, we will evaluate the subscale for anxiety from the self- reported Hopkins Symptom Checklist (SCL-90-R) (Derogatis, 1977). g. Personality. Personality factors have been shown to predict as much as 35% of variance in response to antidepressants (Joyce, Mulder, & Cloninger, 1994). It is possible that personality factors may help discriminate individuals with high and low physiologic reactivity. In order to examine the role of personality, we will administer the Temperament and Character Inventory (TCI), a self-report measure of seven personality dimensions (Cloninger, Svrakic, & Przybeck, 1993). The short- version of the TCI consists of 144 true/false items. Social and Behavioral Measures h. Social Support. Several studies suggest that social support may influence the relationship between stress and immune dysfunction (Uchino, Cacioppo, & Keicolt-Glaser, 1996). Research also suggests that social support may be independently negatively associated with depressed mood (Brown & Bifialco, 1985; Brummett et al., 1998; Cohen, Mermelstein, Kamarck, & Hoberman, 1985). 98 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Therefore, subjects will complete the Interpersonal Support Evaluation List (ISEL) appraisal subscale (Cohen et al., 1985). The ISEL appraisal subscale consists of ten items rated on a 4 point scale, and is a standard measure of perceived availability of social support. Internal reliability is high for general population studies and six week test-retest reliability is satisfactory. i. Exercise. Research suggests that exercise decreases depressed mood (Byrne & Byrne, 1993). Frequency of exercise may also influence immune status (Cohen et al., 1998). Therefore, subjects will complete the Godin Leisure-Time Questionnaire, a 4-item scale indicating frequency of exercise during the last week (Godin & Shepard, 1985). The first three items ask the subject to fill in the times per week that s/he engaged in strenuous, moderate, or mild exercise. The fourth item inquires about strenuous exercise and is rated on a 3 point scale. j. Sleep. Sleep quality is associated with depressed mood, and may explain some or all of the association between depressed mood and immune variables (Irwin, 1995; Savard et al., 1999). Therefore, subjects will complete seven items from the Pittsburgh Sleep Quality Index (Buysse, Reynolds, Monk, Berman, & Kupfer, 1989). This instrument has high reliability, and measures subjective sleep quality, sleep latency, sleep duration, and habitual sleep efficiency (number of hours slept/number of hours spent in bed). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Data Analysis Physiologic reactivity will be calculated as the difference between the pre stressor measure (SO) and the post-stressor measure (SI or S2). For example, one measure of immune reactivity will be the difference between the percentage of NK cells at SO and SI. Likewise, measures of CR will be calculated based upon changes in HR pre- and post-stressor. Physiologic reactivity measures will be calculated for all depressed patients at screening and eight-week follow-up, and for all normal controls at screening and four-week follow-up. Changes in the PAN AS scores between SO and SI or S2 will be evaluated to verify that Stroop Test and Mental Arithmetic induced a degree of subjective stress. Hypothesis 1: Physiologic reactivity will be higher in depressed patients than in normal controls at screening. The difference in physiologic reactivity between normal controls and depressed patients will be calculated for both cardiovascular and immune variables. Physiologic measures will be analyzed in separate repeated measures ANOVAs with group (normal controls, depressed patients) as a between- subject factor and stressor (SO, SI, S2) as within-subject factors. Hypothesis 2: Physiologic reactivity at screening will predict depressed mood at follow-up in depressed patients. Physiologic reactivity and depressed mood from the HDRS and BDI will be kept as continuous variables. Univariate linear regression will be used to determine if there is a statistically significant association between physiologic reactivity at screening and depressed mood at follow-up, after controlling for mood at screening. BP, HR, and immune measures will be 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. considered separately. This analysis will be stratified by treatment type (venlafaxine or placebo). Hypothesis 3: Physiologic reactivity will not change in normal controls, but will decrease as symptoms of depression decrease in depressed patients. The difference between screening and follow-up reactivity will be calculated for both cardiovascular and immune variables. Physiologic measures will be analyzed in separate repeated measures ANOVAs with group (normal controls, depressed patients) as a between-subject factor and stressor (SO, SI, S2) and time (screening, follow-up) as within-subject factors. This analysis will be stratified by treatment type (venlafaxine or placebo). We will also perform exploratory analyses to evaluate whether additional variables may be confounders and/or modifiers of the relationship between physiologic reactivity and depressed mood. Third variables that will be considered include the following: a) demographic and clinical risk factors, b) psychosocial factors (i.e., personality, social support), and c) health behaviors (i.e., exercise, sleep). Information from these exploratory analyses will provide pilot data to help design future studies of individual differences. Power Analysis Since there has been little research on physiologic reactivity in depressed patients, and few if any studies examining repeated measures of physiologic reactivity in depressed patients, it is difficult to estimate the required sample size for 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. this study. Given the lack of previous information, as well as practical limitations, it is best to consider the results of this study as providing pilot data, including estimates that can be used in power calculations for future studies. However, an estimate of the power of this study can be based upon a simplification of the analyses (performing paired t-tests instead of repeated measures ANOVAs) and estimates of mean changes from studies of healthy subjects. A sample size of 10 will have 80% power to detect a difference in means of 5.00, assuming a standard deviation of differences of 5.00, using a paired t-test with a 0.05 two-sided significance level. Since we expect larger differences, the proposed sample size of 26 patients and 26 normal controls, or 13 patients and 13 normal controls for each of the analyses stratified by treatment, should be sufficient to detect statistically significant associations between mood and physiologic reactivity. Also, we expect to increase the power of our study by planning the analysis on continuous outcome variables (Kraemer, 1991). Limitations The small sample size will limit the generalizability of this study. While it would be ideal to conduct this study on a larger sample, this is not feasible given the limitations of time and funds. An alternative would be to focus the study on a specific demographic group, but this may make the difficult job of recruiting subjects, particularly clinical patients, even more difficult. Thus, we are forced to accept a small sample of convenience, and to view the results of this study as pilot 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. data that may be useful for hypothesis generation and planning future studies. The generalizability of this study will also be limited by the demographics of the study sample, which is expected to be predominantly Caucasian, because of the setting, and predominantly female, because depression is more common in women. An important feature of this study is that we will be using a within-subjects design to compare a control condition (pre-stressor SO) with the test condition (post stressors SI, S2). In addition, we will be using a repeated measures design to compare variables at screening and follow-up. The potential for information bias, also referred to as differential misclassification due to measurement bias, is eliminated since there will be no difference in the nature or quality of measurements in control and test conditions. Likewise, control and test conditions are being compared within subjects so selection bias in terms of the conditions is eliminated. However, the internal validity of this study design may be threatened by confounding. A number of third variables may be associated with both physiologic reactivity and depressed mood. For example, having a family history of mood disorders may predispose a subject to high physiologic reactivity as well as severe depressed mood, which could result in positive bias away from the null. In order to evaluate the potential influence of confounding, we will be measuring and examining the influence of a number of third variables including demographic factors (e.g. age, gender), clinical risk factors (e.g. number of episodes, family history), behavioral factors (e.g. exercise, sleep), and psychosocial factors (e.g. personality, social support). 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.5. Human Subjects/Vertebrate Animals The UCLA Human Subjects Protection committee will approve all procedures proposed in this study prior to the beginning of subject enrollment. The consent form, outlining procedures, potential risks and anticipated benefits, right to withdraw, and confidentiality, will be read with each participant, allowing time for questions. Participants will receive a copy of their signed consent form. The physiologic measures, administration of rating scales and completion of self-report forms are minimal risk procedures. Subject will be screened at recruitment, and any subjects expressing fear of needles, math, or speech tasks will be excluded. Potential risks associated with phlebotomy will be minimized using universal precautions. This study will be restricted to adults between the ages of 18 and 65. The age group was chosen to be representative of depression in adults, as distinct from depression in adolescence or older individuals. Due to substantially increased recruitment of minority and low socioeconomic subjects, we expect that approximately 25% of our subjects will be members of ethnic minorities or from low socioeconomic backgrounds. Vertebrate animals: Not applicable. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.6. References Andreasen, N., & Black, D. (1995). Introductory Textbook o f Psychiatry (2nd ed.). Washington, DC: American Psychiatric Press, Inc. Antonuccio, D., Danton, W., & DeNelsky, G. (1995). Psychotherapy versus medication for depression: Challenging the conventional wisdom with data. Professional Psychology: Research and Practice, 26, 574-585. Association, A. P. (1994). Diagnostic and Statistical Manual o f Mental Disorders (Fourth (DSM-IV) ed.). Washington, DC: Author. Barlow, D. H., & Campbell, L. A. (2000). Mixed anxiety-depression and its implications for models of mood and anxiety disorders. Comprehensive Psychiatry, 41(2 Suppl 1), 55-60. Beck, A., Rush, A., Shaw, B., & Emery, G. (1979). Cognitive therapy o f depression. New York: Guilford Press. Beck, A., Steer, R., & Garbin, M. (1988). Psychometric properties of the Beck Depression Inventory: Twenty-five years of evaluation. Clinical Psychology Review, 8,11-100. Beck, A., Ward, C., Mendelsohn, M., Mock, I , & Erbaugh, J. (1961). An inventory for measuring depression. Archives o f General Psychiatry, 4, 561-571. Benschop, R., & Schedlowski, M. (1999). Acute psychological stress. In M. Schedlowski & U. Tewes (Eds.), Psychoneuroimmunology: An interdisciplinary introduction (pp. 293-306). New York: Kluwer Academic/Plenum Publishers. Brosschot, J., Benschop, R., Godaert, G., de Smet, M., Olff, M., Heijnen, C., & Ballieux, R. (1992). Effects of experimental psychological stress on distribution and function of peripheral blood cells. Psychosomatic Medicine, 54, 396-406. Brown, G., & Bifulco, A. (1985). Social Support, Life Events and Depression. In I. Saronson & B. Saronson (Eds ), Social Support: Theory, Research, and Applications (pp. 349-370). Boston, MA: Martinus Nijhoff. Brown, W., Johnson, M., & Chen, M. (1992). Clinical features of depressed patients who do and do not improve with placebo. Psychiatry Research, 41, 203-214. 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Brummett, B. H., Babyak, M. A., Barefoot, J. C., Bosworth, H. B., Clapp- Channing, N. E., Siegler, I. C., Williams, R. B., Jr., & Mark, D. B. (1998). Social support and hostility as predictors of depressive symptoms in cardiac patients one month after hospitalization: a prospective study. Psychosomatic Medicine, 60(6), 707-13. Buysse, D., Reynolds, C., Monk, T., Berman, S., & Kupfer, D. (1989). The Pittsburgh Sleep Quality Index. Psychiatry Research, 28, 193-213. Byrne, A., & Byrne, D. (1993). The effect of exercise on depression, anxiety and other mood states: a review. Journal o f Psychosomatic Research, 17(6), 565-574. Cacioppo, J. (1994). Social neuroscience: Autonomic, neuroendocrine, and immune response to stress. Psychophysiology, 31, 113-128. Cacioppo, J., Poelmann, K., Kiecolt-Glaser, J., Malarkey, W., Burleson, M., Bernston, G., & Glaser, R. (1998). Cellular immune responses to acute stress in female caregivers of dementia patients and matched controls. Health Psychology, 17(2), 182-189. Carney, R. M., Freedland, K. E., Veith, R. C., Cryer, P. E., Skala, J. A., Lynch, T., & Jaffe, A. S. (1999). Major depression, heart rate, and plasma norepinephrine in patients with coronary heart disease. Biological Psychiatry, 45(4), 458-63. Clarke, D. M. (1998). Psychological factors in illness and recovery. New Zealand Medical Journal, 111(1616), 410-2. Cloninger, C., Svrakic, D., & Przybeck, T. (1993). A psychobiological model of temperament and character. Archives o f General Psychiatry, 50, 975-990. Cohen, S., Frank, E., Doyle, W. J., Skoner, D. P., Rabin, B. S., & Gwaltney, J. M., Jr. (1998). Types of stressors that increase susceptibility to the common cold in healthy adults. Health Psychology, 17(3), 214-23. Cohen, S., & Herbert, T (1996). Health Psychology: Psychological factors and physical disease from the perspective of human psychoneuroimmunology. Annual Reviews in Psychology, 47, 113- 142. 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cohen, S., Mermelstein, R., Kamarck, T., & Hoberman, H. (1985). Measuring the functional components of social support. In I. Saronson & B. Saronson (Eds.), Social Support: Theory, Research, and Applications (pp. 73-94). Boston, MA: Martinus Nijhoff. Cohen, S. I. (1986). Measurement of human adaptation to stressful environments— revisiting a 1958 presentation to the American Rocket Society— revising the stress model to a biopsychosocial systems model. American Journal o f Social Psychiatry, 6(3), 175-182. Delehanty, S. G., Dimsdale, J. E., & Mills, P. (1991). Psychosocial correlates of reactivity in black and white men. Journal o f Psychosomatic Research, 35(4-5), 451-60. Derogatis, L. (1977). Administration, Scoring and Procedures Manual fo r the SCL-90-R. Baltimore: Clinical Psychometrics Research. Dimsdale, J. (1988). The effect of depression on cardiovascular reactivity. In T. Field, P. McCabe, & e. al (Eds ), Stress and coping across development (pp. 215-225). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Eaker, E. D. (1998). Psychosocial risk factors for coronary heart disease in women. Cardiology Clinics, 76(1), 103-11. Futterman, A., Kemeny, M., Shapiro, D., & Fahey, J. (1994). Immunological and physiological changes associated with induced positive and negative mood. Psychosomatic Medicine, 56, 499-511. Godin, G., & Shepard, R. (1985). A simple method to assess exercise behavior in the community. Canadian Journal o f Applied Sport Science, 10(3), 141-146. Goodwin, F. (1993). Predictors of antidepressant response. Bulletin o f the Menninger Clinic, 57(2), 146-160. Gorman, J., & Sloan, R. (2000). Heart rate variability in depressive and anxiety disorders. American Heart Journal, 140(4), S77-S83. Greenberg, R., & Fisher, S. (1997). Mood-mending medicines: probing drug, psychotherapy, and placebo solutions. In S. Fisher & R. Greenberg (Eds.), From Placebo to Panacea: Putting Psychiatric Drugs to the Test (pp. 115-172). New York: John Wiley & Sons, Inc. 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Guinjoan, S., Beraabo, J., & Cardinali, D. (1995). Cardiovascular tests of autonomic function and sympathetic skin responses in patients with major depression. Journal o f Neurology, Neurosurgery, and Psychiatry, 58, 299-302. Hamilton, M. (1967). Development of a rating scale for primary depressive illness. British Journal o f Social and Clinical Psychology, 6, 278-296. Herbert, T., & Cohen, S. (1993). Depression and immunity: a meta-analytic review. Psychological Bulletin, 113, 472-86. Irwin, M. (1995). Psychoneuroimmunology of depression. In F. Bloom & D. Kupfer (Eds.), Psychopharmacology: The Fourth Generation o f Progress (pp. 983-998). New York: Raven Press, Ltd. Jemerin, J. M., & Boyce, W. T. (1990). Psychobiological differences in childhood stress response. II. Cardiovascular markers of vulnerability. Journal o f Developmental & Behavioral Pediatrics, 77(3), 140-50. Jonas, B., Franks, P., & Ingram, D. (1997). Are symptoms of anxiety and depression risk factors for hypertension?: longitudinal evidence from the National Health and Nutrition Examination Survey I Epidemiologic Follow-up Study. Archives o f Family Medicine, 6(1), 43-49. Joyce, P., & Paykel, E. (1989). Predictors of drug response in depression. Archives o f General Psychiatry, 46, 88-89. Joyce, P R., Mulder, R. T., & Cloninger, C. R. (1994). Temperament predicts clomipramine and desipramine response in major depression [see comments]. Journal o f Affective Disorders, 50(1), 35-46. Judd, L., Akiskal, H., Zeller, P., Paulus, M., Leon, A., Maser, J., Endicott, J., Coryell, W., Kunovac, J., Mueller, T., Rice, J., & Keller, M. (2000). Psychosocial disability during the long-term course of unipolar major depressive disorder. Archives o f General Psychiatry, 57(4), 375-80. Kocsis, J. (1990). New issues in the prediction of antidepressant. Psychopharmacology Bulletin, 26(1), 49-53. Kop, W. J. (1997). Acute and chronic psychological risk factors for coronary syndromes: Moderating effects of coronary artery disease severity. Journal o f Psychosomatic Research, 43(2), 167-181. 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Kraemer, H. C. (1991). To increase power in randomized clinical trials without increasing sample size. Psychopharmacology Bulletin, 27(3), 217-24. Krishnan, K., Gadde, K., & Kim, Y. (1998). Psychoneuroendocrinology and brain imaging in depression. Psychoneuroendocrinology, 21(2), 465- 472. Lane, D., Carroll, D., & Lip, G. Y. (1999). Psychology in coronary care. Qjm, 92(8), 425-31. Lavoie, K., & Fleet, R. (2000). The impact of depression on the course and outcome of coronary artery disease: review for cardiologists. Canadian Journal o f Cardiology, 16(5), 653-662. Legato, M. (1997). Gender-specific physiology: how real is it? How important is it? International Journal o f Fertility & Womens Medicine, 42(1), 19-29. Light, K. C., Dolan, C. A., Davis, M. R , & Sherwood, A. (1992). Cardiovascular responses to an active coping challenge as predictors of blood pressure patterns 10 to 15 years later. Psychosomatic Medicine, 54(2), 217-30. Manuck, S., Cohen, S., Rabin, B., Muldoon, M., & Bachen, E. (1991). Individual differences in cellular immune response to stress. Psychological Science, 2(2), 111-115. McDaniel, J., Musselman, D., Porter, M., Reed, D., & Nemeroff, C. (1995). Depression in patients with cancer: Diagnosis, biology, and treatment. Archives o f General Psychiatry, 52, 89-99. Miller, A.-H. (1998). Neuroendocrine and immune system interactions in stress and depression. Psychiatric Clinics o f North America, 21(2), 443-463. Mills, P. J., & Dimsdale, J. E. (1991). Cardiovascular reactivity to psychosocial stressors. A review of the effects of beta-blockade. Psychosomatics, 32(2), 209-20. Musselman, D. L., Evans, D. L., & Nemeroff, C. B. (1998). The relationship of depression to cardiovascular disease. Epidemiology, biology, and treatment. Archives o f General Psychiatry, 55(7), 580-592. 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Naliboff, B., Benton, D., Solomon, G., Morley, J., Fahey, J., Bloom, E., Makinodan, T., & Gilmore, S. (1991). Immunological changes in young and old adults during brief laboratory stress. Psychosomatic Medicine, 53, 121-132. Naliboff, B., Solomon, G., Gilmore, S., Fahey, J., Benton, D., & Pine, J. (1995). Rapid changes in cellular immunity following a confrontational role-play stressor. Brain, Behavior, and Immunity, 9, 207-219. Newman, J. D., McGarvey, S. T., & Steele, M. S. (1999). Longitudinal association of cardiovascular reactivity and blood pressure in Samoan adolescents. Psychosomatic Medicine, 61(2), 243-9. Pike, J., Smith, T., Hauger, R., Nicassio, P., Patterson, T., McClintick, J., Costlow, C., & Irwin, M. (1997). Chronic life stress alters sympathetic, neuroendocrine, and immune responsivity to an acute psychological stressor in humans. Psychosomatic Medicine, 59, 447- 457. Ravindran, A., Griffeths, J., Merali, Z., & Anisman, H. (1996). Variations in lymphocyte subsets associated with stress in depressive populations. Psychoneuroendocrinology, 21, 659-671. Ravindran, A.-V., Griffiths, J., Merali, Z., & Anisman, H. (1998). Circulating lymphocyte subsets in major depression and dysthymia with typical or atypical features. Psychosomatic Medicine, 60(3), 283-289. Roy, M.-P., Steptoe, A., & Kirschbaum, C. (1998). Life events and social support as moderators of individual differences in cardiovascular and cortisol reactivity. Journal o f Personality and Social Psychology, 75(5), 1273-1281. Rozanski, A., Blumenthal, J. A., & Kaplan, J. (1999). Impact of psychological factors on the pathogenesis of cardiovascular disease and implications for therapy. Circulation, 99(16), 2192-217. Savard, J., Miller, S., Mills, M., O'Leary, A., Harding, H., Douglas, S., Mangan, C., Belch, R., & Winokur, A. (1999). Association between subjective sleep quality and depression on immunocompetence in low-income women at risk for cervical cancer. Psychosomatic Medicine, 61(4), 496-507. 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Schedlowski, M., Jacobs, R., Stratmann, G., Richter, S., Hadicke, A., Tewes, U., Wagner, T., & Schmidt, R. (1993). Changes of natural killer cells during acute psychological stress. Journal o f Clinical Immunology, 13(2), 119-126. Schwartzman, J., & Glaus, K. (2000). Depression and coronary heart disease in women: Implications for clinical practice and research. Professional Psychology: Research and Practice, 31(1), 48-57. Shapiro, D., Jamner, L., Lane, J., Light, K., Myrtek, M., Sawada, Y., & Steptoe, A. (1996). Blood pressure publication guidelines. Psychophysiology, 33, 1-12. Sheffield, D., Krittayaphong, R., Cascio, W., Light, K., Golden, R., Finkel, J., Glekas, G., Koch, G., & Sheps, D. (1998). Heart rate variability at rest and during mental stress in patients with coronary artery disease: Differences in patients with high and low depression scores. International Journal o f Behavioral Medicine, 5(1), 31-47. Sherwood, A., & Turner, J. (1992). A conceptual and methodological overview of cardiovascular reactivity research. In J. Turner, A. Sherwood, & K. Light (Eds.), Individual differences in cardiovascular responses to stress (pp. 3-32). New York: Plenum Press. Simon, G., Vonkorff, M., Ustun, T., Gater, R., Gureje, O., & Sartorius, N. (1995). Is the lifetime risk of depression actually increasing? Journal o f Clinical Epidemiology, 48(9), 1109-1118. Stoudemire, A. (Ed.). (1995). Psychological Factors Affecting Medical Conditions. Washington, DC: American Psychiatric Press, Inc. Tewes, U. (1999). Concepts in Psychology. In M. Schedlowski & U. Tewes (Eds.), Psychoneuroimmunology: An interdisciplinary introduction (pp. 93-111). New York: Kluwer Academic/Plenum Publishers. Tueting, P. (1991). Psychophysiological predictors of drug treatment response. Psychiatric Medicine, 9(1), 145-161. Uchino, B., Cacioppo, J., Malarkey, W., & Glaser, R. (1995). Individual differences in cardiac sympathetic control predict endocrine and immune responses to acute psychological stress. Journal o f Personality and Social Psychology, 69, 763-743. Ill Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Uchino, B.-N., Cacioppo, J.-T., & Keicolt-Glaser, J.-K. (1996). The relationship between social support and physiological processes: A review with emphasis on underlying mechanisms and implications for health. Psychological Bulletin, 119(3), 488-531. Waked, E. G., & Jutai, J. W. (1990). Baseline and reactivity measures of blood pressure and negative affect in borderline hypertension. Physiology & Behavior, 47(2), 265-71. Waldstein, S., Bachen, E., & Manuck, S. (1997). Active coping and cardiovascular reactivity: A multiplicity of influences. Psychosomatic Medicine, 59(6). Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: the PAN AS scales. Journal o f Personality & Social Psychology, 54(6), 1063-70. Widiger, T., Frances, A., Pincus, H., Ross, R., First, M., & Davis, W. (Eds ). (1996). DSM-IV Sourcebook (Vol. 2). Washington, DC: American Psychiatric Association. Wilhelm, K., Parker, G., & Hadzi-Pavlovic, D. (1997). Fifteen years on: evolving ideas in researching sex differences in depression. Psychological Medicine, 27(4), 875-83. Williams, J., Gibbon, M., First, M., Spitzer, R., Davies, M., Borus, J., Howes, M., Kane, J., Pope, H., Rounsaville, B., & Wittchen, H. (1992). The Structured clinical Interview for DSM-III-R (SCID). II. Multisite test- retest reliability. Archives o f General Psychiatry, 49, 630-636. Wilson, D. K., Kliewer, W., Bayer, L., Jones, D., Welleford, A., Heiney, M., & Sica, D. A. (1999). The influence of gender and emotional versus instrumental support on cardiovascular reactivity in African- American adolescents. Annals o f Behavioral Medicine, 21(3), 235-43. 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5. Toward a Biopsychosocial Theory of Major Depression 5.1. Introduction Various authors have suggested that theory-driven research is needed in order to develop a comprehensive understanding of major depressive disorder (MDD) (Flett, Vredenburg, & Krames, 1997; Kraemer & Telch, 1992; Wallace, 1990; Willner, 1985). A “theory” can be defined as "a set of interrelated concepts, definitions, and propositions that presents a systematic view of events or situations by specifying relations among constructs in order to explain and predict the events or situations" (Spruijt-Metz, 1999, p43). Sometimes the term “theory” and “model” are used interchangeably. However, there is an important distinction that needs to be made between the two. A theory represents abstract ideas through the relationships of constructs. By contrast, models use variables (i.e. operational definitions of constructs) to illustrate concrete consequences that result from the application of a theory (McLaren, 1998; Spruijt-Metz, 1999). Thus, a theory can be used to provide an explanation for specific observations related to a particular phenomenon, and a model can be used to test the validity of that theory. A good theory of MDD would help explain clinical observations, predict risk of MDD, lead to more specific and effective treatments, and be testable in human and/or animal models (Andreasen, 1997; Bebbington, 1987; McLaren, 1998; Spruijt-Metz, 1999; Willner, 1985). The purpose of this paper is to propose a method for developing a comprehensive theory of MDD. It will be argued that the biomedical approach to research on MDD supports a focus on biological factors, and tends to exclude 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. consideration of psychological and social factors such as cognitive appraisal, personality, life events, social support, or social context. Consequently, biomedical research leads to an accumulation of information about biological factors associated with MDD, but fails to explain how psychological, interpersonal, and sociocultural factors contribute to MDD. By contrast, a biopsychosocial (BPS) approach to research on MDD encourages the simultaneous measurement of biological, psychological, and social factors. Therefore, evidence from BPS studies can provide information about relationships between biological and psychosocial factors in MDD, which has the potential to enhance understanding of the etiology and treatment of MDD. The final part of this paper describes a research cycle for developing a specific BPS theory of MDD. 5.2. Biomedical Approach 5.2.1. Biomedical approach to medicine In 1674 the Dutch biologist Anton van Leeuwenhoek peered through a microscope and discovered the world of bacteria and protozoa in a drop of water (Murray, Rosenthal, Kobayashi, & Pfaller, 1998). In 1840 the German pathologist Friedrich Henle proposed the so-called “germ theory,” which stated that microorganisms were responsible for causing human diseases (Murray et al., 1998). Eventually the germ theory was replaced with a more general somatic theory, which proposed that material agents are responsible for causing human disease. Thus, the 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. search began and continues today to identify microorganisms, genes, neurotransmitters, and other biological or biochemical agents responsible for disease. The biomedical approach is the application of somatic theory to clinical research and practice. According to the biomedical approach, diseases are the practical consequences of somatic causes (Engel, 1977). Biomedical research employs factor-analytic designs where all factors are held constant except for the one factor under study (Engel, 1980). The biomedical practice of medicine focuses on the identification and treatment of specific somatic agents. Therefore, the biomedical approach has been described as emphasizing formistic (categorical) and mechanistic (single-cause) thinking (Schwartz, 1982). In addition, Cartesian mind- body dualism is supported by a biomedical approach, with the body delegated to the domain of physicians and the mind left to philosophers and theologians (Engel, 1977; Taylor, 1999). 5.2.2. Biomedical approach to major depression As viewed from a biomedical approach, MDD is a disease caused by biological or biochemical dysfunctions. Since key symptoms of depression include disruptions in emotional, cognitive and behavioral functioning, biomedical research has focused on finding neurological causes of these symptoms. One of the earliest hypotheses was that depression was caused by a depletion of norepinephrine, a neurotransmitter, in the brain. The catecholamine hypothesis was based upon the finding that effective pharmacological treatments increased norepinephrine levels in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the brain (Andreasen & Black, 1995). The catecholamine hypothesis is now considered oversimplified since there is extensive evidence for changes in other neurotransmitter systems, including serotonergic and cholinergic (Fava & Kendler, 2000; Tewes, 1999). Therefore, pharmacological studies are searching for increasingly complex biochemical explanations of MDD (Bedi, 1999). In addition, the biomedical approach to MDD has stimulated attempts to find other biological explanations of MDD via structural and functional neuroimaging, neuroendocrine, and psychoneuroimmunology studies. 5.2.3. Strengths and weaknesses of the biomedical approach Over the past two centuries, the biomedical approach has proven very useful in helping to understand the causes of disease and to develop effective treatments (Antonovsky, 1989). Nonetheless, the biomedical approach has been less successful in helping to explain the etiology of MDD. Furthermore, although several pharmacological treatment options are available to patients with MDD, response rates are limited to 45-70% of patients, which is only about 18-25% higher than placebo response rates (Goodwin, 1993; Khan, Warner, & Brown, 2000). In addition, about 20-30% of patients who experience a major depressive episode will continue to have residual symptoms, about 50-60% of patients with MDD will experience a second episode, and about 90% of those who experience a second episode will have a third episode (Association, 1994; Stangl & Greenhouse, 1998). Thus, the biomedical approach has had limited success explaining and treating Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MDD, which may be due at least in part to inherent weaknesses in the biomedical approach. Perhaps the most common criticism of the biomedical approach is that it is reductionist (Schwartz, 1982; Taylor, 1999). The focus of biomedical research on identifying biochemical causes of MDD is reflective of the reductionism inherent in the biomedical approach. Research on the efficacy of various antidepressants and their known effect on specific neurotransmitters provides indirect evidence that MDD is associated with neurochemical changes. Based upon this indirect evidence, it has been proposed that MDD is caused by alterations in neurotransmitters that are corrected by antidepressants. However, evidence for alterations in biochemical processes does not necessarily mean that biochemical processes are causal agents (Gabbard, 1994). Pharmacological depletion studies have attempted to provide direct evidence for a causal relationship between specific biochemical states of the brain and the clinical state of MDD. Results suggest that depletion of brain neurotransmitters in normal healthy subjects does not result in depression (Bedi, 1999). However, depletion of neurotransmitters can result in depression in patients with a history of depression or subjects undergoing life distress (Bedi, 1999). This suggests that neurotransmitter alterations by themselves are not sufficient to cause depression. Rather, neurotransmitter alterations may interact with other factors to cause MDD, such as life distress. The fact that onset of MDD does not occur until variable ages in adulthood, and the extensive evidence that episodes of MDD are triggered by stressful life events, suggests that consideration of psychological and Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. social factors in addition to biological factors is necessary in order to explain the onset and course of MDD (Andreasen & Black, 1995; Maughan & McCarthy, 1997). A second criticism of the biomedical approach is that it supports Cartesian mind-body dualism. One problem with mind-body dualism is that it leads to the creation of false categories. For example, in the past a dualist division between “physical” and “mental” phenomena lead to an attempt to distinguish between depressions that were biological in origin, as opposed to psychological in origin (Willner, 1985). It has since been shown that there is limited, if any, evidence to support a distinction between a biological and psychosocial origin in treating MDD. A second problem with mind-body dualism is that it tends to support a focus on finding biological causes and treatments, and a devaluing of psychosocial issues and interventions. For example, despite the atheoretical organization of the DSM-IV, there is still an implicit assumption that mental disorders are predominantly biological in origin and therefore will respond better to biological therapies (Jones, 1998; Weissman, 1999). Although most currently practicing psychiatrists would agree that episodes of depression are influenced by a combination of biological and psychological factors, there is still a tendency in the field of psychiatry to focus on biological causality and pharmacological treatments (Weissman, 1999). The focus of medical research on biological phenomena is not supported by the extensive evidence indicating that biological and psychosocial factors are relevant to the etiology and treatment of so-called “physical” and “mental” illnesses. This is illustrated by evidence that individuals with various physical diseases are at 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. higher risk for mental illnesses. For example, the prevalence of psychiatric co morbidity is higher than the general population in patients with heart disease, cancer, neurological disorders, and lung disease (Clarke, 1998). Conversely, individuals with mental illnesses are at higher risk for various physical diseases. For example, patients with clinical depression have more physical illnesses than non-depressed individuals, and have a worse prognosis for a variety of medical conditions including myocardial infarction, subarachnoid hemorrhage, upper gastrointestinal hemorrhage, pulmonary embolism, stroke, muscular sclerosis, epilepsy, renal disease, and certain cancers (Stoudemire, 1995). Furthermore, a number of studies have shown that psychological interventions can reduce morbidity and mortality associated with diseases such as irritable bowel syndrome, cardiovascular disease, and certain types of cancer (Clarke, 1998; Stoudemire, 1995). In patients with MDD, the majority of research suggests that psychosocial treatments, such as cognitive or interpersonal therapy, are at least as effective as pharmacotherapy (Fava & Kendler, 2000; Schulberg, Katon, Simon, & Rush, 1998). In addition, antidepressant trials show that about 30-50% of patients respond to placebo, which suggests that psychological or social factors may have a large influence on recovery from MDD (Greenberg & Fisher, 1997; Khan et al., 2000). Thus, a growing body of literature supports the need to understand both biological and psychosocial factors in physical and mental illnesses. In summary, the biomedical approach emphasizes biological theories and the belief that advances in neurobiology will help to understand diseases such as MDD. 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. While neurobiological research has played a significant role in contributing to our understanding of the pathophysiology of MDD, it is unlikely that a purely neurobiological theory will be able to explain the psychological, interpersonal, and sociocultural factors that contribute to the etiology and treatment of depression (Wallace, 1990). A growing amount of evidence suggests that a new approach is needed in order to integrate knowledge about biological and psychosocial factors into a comprehensive theory of MDD. 5.3. Biopsychosocial Approach 5.3.1. Biopsychosocial approach to medicine The attempt to understand the relationship between biological and psychological factors dates back to the earliest days of medicine (Kagan, 1994). Hippocrates and Galen, ancient Greek and Roman philosophers, were among the first to describe the influence of psychological characteristics on physical well-being (Friedman, 1990). It was only recently, over the past few hundred years or so, that medicine became more biologically focused (Clarke, 1998). However, despite the change in the focus of medicine, the nature of human experience did not change. In the late 1970s, George Engel proposed the idea of a BPS approach to medicine, which added a consideration of social factors to the ancient Greek and Roman focus on biological and psychological phenomena (Engel, 1977). Engel suggested that General Systems Theory (GST) offered a basis for the BPS approach. According to GST, "all levels of organization are linked to each other in a hierarchical relationship 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. so that change in one affects change in the others" (Engel, 1977, p i96). Therefore, in accord with GST, human diseases could be defined as the product of interactions between macrolevel (e.g. social support) and microlevel (e.g. genetic) processes (Engel, 1977; Taylor, 1999). According to the BPS approach, the predisposition, onset, course, and outcome of most illnesses is influenced by biological, psychological and social factors (Cohen-Cole & Levinson, 1990). Biological factors refer to material structures or processes in the human body, and include constructs such as genetic inheritance, neurological volumes, or immune cell activity. Psychological factors refer to mental and behavioral characteristics, such as cognitions, personality, and affect. Social factors describe interactions between the individual and human society, and include stressful life events and social support. The term “psychosocial” refers to both psychological and social factors. Many constructs have biological, psychological and social components, and so are best described by the combined term “biopsychosocial” (BPS). For instance, the influence of gender, a chronic illness, or a family history of a particular illness may be evaluated by a combination of biological, psychological, and social variables. Likewise, health behaviors, such as diet and exercise, are multifactorial constructs that may have reciprocal relationships with biological, psychological, and social variables. In general, BPS research measures biological, psychological and social variables in order to develop effective therapeutic interventions (Cohen-Cole & Levinson, 1990). The BPS practice of medicine involves inquiring about 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. psychosocial factors in addition to measuring biological factors (Sadler & Hulgus, 1990). Thus, the BPS approach can be described as emphasizing contextual (relational) and organistic (multi-causal) thinking (Schwartz, 1982). Furthermore, an integrated understanding of mind and body is supported by a BPS approach, rather than mind-body dualism (Cohen-Cole & Levinson, 1990; Taylor, 1999). 5.3.2. Biopsychosocial approach to major depression As viewed from a BPS approach, MDD is a disease caused by a combination of biological, psychological and social factors. This suggests that BPS research should consider social science studies of depression, as well as the results of biomedical research. There are a number of psychological and social theories of depression, and a review of each of them is beyond the scope of this paper. In general, these theories point to the importance of cognitive processing and social context (Champion & Power, 1995). Individual differences in the evaluation of life events and social relationships may explain differences in vulnerability to depression, as well as recovery from depression (Johnson, Han, Douglas, Johannet, & Russell, 1998; Jones, 1998; Lara & Klein, 1999). Likewise, social circumstances and behaviors may influence the etiology and treatment of depression, either directly or through interactions with cognitive processes (Champion & Power, 1995; Lara & Klein, 1999). Thus, the BPS approach to MDD supports the investigation of a wide range of factors, but in particular supports studies that attempt to analyze relationships between biological and psychosocial variables. 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.3.3. Strengths and weaknesses of the biopsychosocial approach The BPS approach advocates the consideration of multiple factors and thereby avoids the reductionism associated with the biomedical approach. Over the past few decades, clinicians interested in a BPS approach have helped promote the practice of a more humane medicine. Medical education has initiated programs to improve physician communication skills and promote patient-centered interviewing (Cohen-Cole & Levinson, 1990; Zimmermann & Tansella, 1996). Physicians have been encouraged to collect more information about psychosocial factors from the patient’s current and past history (Leigh & Reiser, 1985; McHugh & Yallis, 1986; Wolkenstein & Butler, 1998). Physicians also have been encouraged to collaborate with allied health practitioners, such as health psychologists and social workers, in order to ensure that all of the patient’s health care concerns are addressed (Taylor, 1999). A second strength of the BPS approach is that it is consistent with evidence presented above that divisions between “biological” and “psychological” depressions are erroneous, as are attempts to divide “physical” and “mental” illnesses by biological and psychosocial factors (Whybrow, 1997). Rather, according to a BPS approach, human beings are from birth biologically organized self-regulatory systems whose functioning is inseparable from continuous, psychologically mediated interaction with the social environment (Levin & Solomon, 1990). Therefore, it is not possible to draw a clear boundary between a biological body and a psychosocial mind (Levin & Solomon, 1990). Consequently, evidence of the involvement of 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. multiple factors in the predisposition, onset, course, and outcome of diseases such as MDD is in accord with a BPS approach. In addition, a BPS approach to research on MDD can serve as a bridge between mind, body and social environment by providing correlations between cognitive appraisal, physical states and social context (Levin & Solomon, 1990). Consideration of the cognitive appraisal or personal meaning of events may help explain the role of stress in the etiology of MDD by identifying individual differences in response to stressors (Gabbard, 1994). In addition, the diagnosis of MDD might be improved by considering subjective meanings in order to evaluate whether an individual’s response to an event or circumstance is within normal range (Wakefield, 1998). Furthermore, measurement of personal meanings might explain differences in response to treatment (Brody, 1997). For example, studies of the placebo effect in antidepressant drug trials indicate that therapeutic response can be predicted by factors such as personal expectations of the treatment, experience of emotional support, and a sense of control over the depressed state (Fisher & Greenberg, 1997; Shapiro & Shapiro, 1997). Concurrent collection of biological and psychosocial data is supported by the BPS approach, and has the potential to integrate personal experiences of depression with physiological states, which in turn may lead to a better understanding of the etiology and treatment of MDD. Most mental health professionals agree that consideration of biological, psychological and social factors is fundamental to a complete understanding of MDD (Mullen, 1998). Nonetheless, it has been over 30 years since Engel proposed the 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BPS approach, and yet the biomedical approach has remained the status quo. One obstacle to the implementation of a BPS approach may be the belief that, because the biomedical approach has developed a number of effective treatments for many diseases, it will eventually be successful for all diseases (Antonovsky, 1989). However, this cultural bias ignores the fact that the biomedical approach has had limited success in identifying the etiology of MDD or attaining high treatment response rates, which suggests the need for a new approach to research on MDD. Another obstacle to implementing a BPS approach may be the difficulty of establishing collaborations between interdisciplinary teams due to differences in training, theoretical paradigms, languages, and working styles (Antonovsky, 1989; McDaniel, 1995). Establishing interdisciplinary teams does require increased effort in order to overcome professional differences. However, the growing body of evidence that relationships between biological, psychological and social factors are critical to a variety of diseases suggests that investing the effort to establish interdisciplinary teams will be a necessary step for future progress. A third obstacle is that the BPS approach fails to specify the nature of relationships between biological and psychosocial processes, and so offers little help in guiding clinical research or practice (McLaren, 1998). In addition, the BPS approach does not make predictions about specific disease processes (Temoshok, 1990). Thus, it is not obvious how to use a BPS approach to articulate a specific BPS theory of MDD that does specify relationships between factors and makes specific predictions about the 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. course of MDD. In the remainder of this paper, we will discuss what is known and what needs to be known in order to articulate a specific BPS theory of MDD. 5.4. Developing a Biopsychosocial Theory of Major Depression 5.4.1. Review data on biopsychosocial factors in major depression There are four key phases in a research cycle that would lead to the development of a BPS theory of MDD. The first phase is to review data on biological and psychosocial factors in patients with MDD. Extensive research suggests that MDD is associated with a number of biological factors, including genetic factors (Andreasen & Black, 1995), alterations in neurotransmitters (Andreasen & Black, 1995; Association, 1994; McDaniel, Musselman, Porter, Reed, & Nemeroff, 1995), changed metabolism in specific brain regions (Brody, Barsom, Bota, & Saxena, 2001; Drevets, 2001; Fava & Kendler, 2000; Leuchter et al., 1997), sleep EEG abnormalities (Andreasen & Black, 1995; Kupfer, 1995), neuroendocrine disruptions (Krishnan, Gadde, & Kim, 1998; McDaniel et al., 1995; Naisberg, 1996), and immune changes (Herbert & Cohen, 1993; Holden, Pakula, & Mooney, 1998; Irwin, 1995; Irwin et al., 1990; Segerstrom, 1997). Social science studies suggest the etiology and course of MDD may be influenced by factors such as stressful life events (Heim, Owens, Plotsky, & Nemeroff, 1997; Maughan & McCarthy, 1997; Paykel, 1994), personality (Coyne & Whiffen, 1995; Kendler, Kessler, Neale, Heath, & Eaves, 1993), social support (Gruen, 1993; Wade & Kendler, 2000), and exercise 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Blumenthal et al., 1999; Miser, 2000). For a more extensive discussion, see the review of MDD by Fava and Kendler (2000). However, only a limited number of studies have examined both biological and psychosocial factors in patients with MDD to date. Thus, it is not known how biological and psychosocial constructs are related. A few interdisciplinary studies of MDD provide support for the consideration of both biological and psychosocial factors. For example, a longitudinal study of 680 Caucasian female-female monozygotic and dizygotic twin pairs examined genetic factors, childhood parental loss, neuroticism, stressful life events, social support, and previous history of major depression (Kendler et al., 1993). At least one member of the twin pairs was diagnosed with MDD. The researchers found that 50.1% of the liability to MDD was predicted by stressful life events (total effect of 0.388), genetic factors (0.329), previous history of major depression (0.302), and neuroticism (0.245). Thus, this study suggests that the etiology of MDD may be influenced by both biological and psychosocial factors. In addition, several recent reviews indicate that depressive disorders are associated with cardiovascular morbidity and mortality (Kop, 1997; Lavoie & Fleet, 2000; Musselman, Evans, & Nemeroff, 1998; Stoudemire, 1995). The prevalence of MDD is about three-times higher in patients with coronary artery disease compared to the general population (Rozanski, Blumenthal, & Kaplan, 1999). In addition, the prevalence of cardiac events is about four-times higher in individuals with a history of MDD compared to individuals with no history (Lane, Carroll, & Lip, 1999). 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Some studies also provide evidence for a gradient between the magnitude of depression and risk of future cardiac events (Rozanski et al., 1999). The association between depression and heart disease appears to be independent of gender, although more studies of men have been reported (Eaker, 1998; Schwartzman & Glaus, 2000). By nature of the association, the literature on cardiovascular disease and depression suggests the importance of measuring biological and psychosocial factors that may explain the relationship by mediation, moderation or confounding. In summary, the results of studies of MDD conducted to date suggest that a number of biological and psychosocial factors are potentially relevant to consider in developing a BPS theory of MDD. However, more studies are needed that examine relationships between biological and psychosocial factors in MDD. 5.4.2. Generate biopsychosocial hypotheses about major depression A critical phase in developing a BPS theory of MDD is to generate hypotheses that explain data on BPS factors and specify relationships between factors. For example, biological diathesis-stress hypotheses propose that risk of psychiatric illness is increased by an interaction between biological traits and psychosocial stressors (Gabbard, 1994). Several researchers have described biological diathesis-stress hypotheses of MDD (Arborelius, Owens, Plotsky, & Nemeroff, 1999; Bebbington, 1987; Bedi, 1999; Checkley, 1996; Nemeroff et al., 1999; Tilders & Schmidt, 1999; Whybrow, Akiskal, & McKinney, 1984; Willner, 1985). All of these hypotheses assume that recent life stressors predict MDD. While 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. there is some debate in the literature, several literature reviews suggest that acute stressful life events increase risk of depressive episodes (Connor & Leonard, 1998; Heim et al., 1997; Maughan & McCarthy, 1997; Paykel, 1994). The association between stressful life events and MDD appears to be stronger in younger patients with mild-moderate depression, nonmelancholic depression, and fewer episodes of depression (Kendler, Thornton, & Gardner, 2000; Kohn et al., 2001). One recent diathesis-stress hypothesis is that early life stressors result in a vulnerability to dysfunction in neurons that secrete corticotropic releasing factor (CRF) (Arborelius et al., 1999; Checkley, 1996; Nemeroff et al., 1999). Simultaneously, the negative emotional memory may be stored in circuitry that involves the amygdala and hippocampus (Andreasen, 1997). Thus, future stressors that activate corticolimbic circuitry may result in high CRF, ACTH and cortisol responses. This hypothesis suggests that early life stressors and recent stressors increase risk of MDD, and that the effects of early and recent stressors are mediated by the corticolimbic circuitry. The CRF hypothesis has not been tested to date, but there is evidence in the literature to support its plausibility. Animal studies provide evidence that early life stress can result in long-term alterations in CRF neurons (Heim et al., 1997). Additional evidence is consistent with the CRF hypothesis, including findings of hippocampal atrophy in survivors of childhood abuse (Nemeroff et al., 1999), decreased density of CRF receptors in the frontal cortex of suicide victims (Heim et al., 1997), and HP A hyperactivity in some depressed patients (Andreasen & Black, 1995; Heim et al., 1997; McDaniel et al., 1995). In 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. addition, CRF-containing neurons modulate monoaminergic neurotransmitter systems, and so may provide a link with neurotransmitter abnormalities implicated in MDD (Nemeroff et al., 1999). Although a variety of evidence supports the plausibility of this hypothesis, one inconsistent piece of evidence is that some, but not all, depressed patients exhibit HPA hyperactivity, a marker of corticolimbic activity. This might be explained by various individual differences that moderate the degree of HPA hyperactivity. However, these factors are not included in the CRF hypothesis. Another diathesis-stress hypothesis, called the Final Common Pathway (FCP) hypothesis, makes a distinction between predisposing and precipitating factors that increase risk of MDD (Whybrow et al., 1984). The FCP hypothesis describes MDD as a dysregulated psychobiological state characterized by disruptions in biological, psychological and social functioning. Various biological and psychosocial factors may predispose an individual to increased risk of MDD, such as genetic inheritance, age, gender, chronic illness, temperament and character traits, loss of attachment during childhood, social class, or lack of social support (Whybrow et al., 1984; Willner, 1985). However, precipitating factors are needed to trigger the dysfunctional regulatory state, and these factors are usually psychosocial stressors. The FCP hypothesis also describes cognitive appraisal as a mediator of the effects of predisposing and precipitating factors. Unlike the CRF hypothesis, the FCP hypothesis considers the influence of a variety of individual differences over time. Various research supports the importance of predisposing factors such as family 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. history, gender, chronic illness, and childhood loss in predicting risk of MDD (Andreasen & Black, 1995; Culbertson, 1997; Lehtinen & Joukamaa, 1994). However, further research is needed to specifically test whether interactions exist between specific predisposing factors and recent stressors, as well as whether cognitive appraisal mediates risk of MDD. One limitation of the FCP hypothesis is that it does not distinguish between brain dysfunction and the clinical expression of MDD. Another limitation of both the CRF and FCP hypotheses is that they do not consider the possibility of reciprocal relationships between MDD and various risk factors. A general BPS hypothesis of MDD is shown in Figure 5.1. Figure 5.1. A General BPS Hypothesis of MDD Cognitive Appraisal Biological Factors e.g. genetics, physiology Psychosocial Factors e.g. personality, stressors Brain dysfunction Biological Factors Psychosocial Factors MDD Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. This hypothesis distinguishes brain dysfunction and the clinical expression of MDD. This allows for the possibility that different BPS risk factors may lead to different dysfunctional brain states, which may lead to different clinical manifestations of MDD. However, in all cases some type of brain dysfunction is part of the mechanism leading to MDD. Cognitive appraisal is predicted to be a mediator of the relationship between BPS factors and brain dysfunction. In addition, this hypothesis proposes that reciprocal relationships may exist between BPS factors and brain dysfunction, as well as between cognitive appraisal and BPS factors. While more specific BPS hypotheses about MDD could be generated, it is the hypotheses that examine relationships between biological and psychosocial factors that have the most potential to further understanding of mind-body relationships in MDD. 5.4.3. Conduct multifactorial longitudinal studies The next phase in building a BPS theory of MDD is to design studies to test BPS hypotheses about MDD. The testing of BPS hypotheses necessarily involves conducting multifactorial studies. Therefore, research designs should incorporate assessments across biological systems, as well as a range of psychosocial constructs and health-related behaviors (Adler & Matthews, 1994; Uchino, Cacioppo, & Keicolt-Glaser, 1996). Some researchers already have begun to apply a BPS approach to topics such as personality disorders, rheumatoid arthritis, gastrointestinal illness, and ischemic heart disease (Drossman, 1998; Marusic, Gudjonsson, Eysenck, & Stare, 1999; Paris, 1993; Schoenfeld Smith et al., 1996). Multifactorial studies of 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. MDD may help refine criteria for diagnosis, improve tracking of changes in symptoms, and generate hypotheses for more successful treatments. Ideally, many of the multifactorial studies would be prospective and include repeated measures in order to help quantify duration and frequency of exposure to various factors, as well as temporal relationships (Herbert & Cohen, 1993; Lepore, 1997; Light, Dolan, Davis, & Sherwood, 1992). Longitudinal studies could prospectively measure premorbid characteristics, age at onset, duration of an episode, or number of episodes, and thereby reduce the influence of recall bias (Kraemer & Telch, 1992). In addition, longitudinal studies that follow the severity of depressive symptoms over time will provide information about the relationship between subthreshold, minor and major depression. Better measurement of time dependent variables will facilitate distinctions between state and trait variables, and help identify opportunities for effective prevention (Kraemer, Gullion, Rush, Frank, & Kupfer, 1994). One study design would be to follow a general population sample in order to examine BPS relationships before and after MDD onset. Such a study would contribute significantly to understanding the etiology of MDD. Unfortunately, obtaining funding for such a study might be difficult due to the long duration, which would have to be at least 10 years or longer. An alternative study design would be to follow a population of 1s t episode patients in order to examine BPS relationships after the 1s t episode. This type of study would help determine the etiology of recurrent episodes, a particularly important question to answer in order to improve 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. treatment of patients with MDD. An even simpler study design could follow patients during treatment for MDD in order to examine BPS relationships within an episode of major depression. Information from such a study could be used to determine mind-body relationships relevant to treatment and prognosis. In addition, single case series could be used to examine relationships between multiple factors over time within single individuals, and thereby generate specific hypotheses for larger studies. The future increase in multifactorial longitudinal studies should occur simultaneously with the development and use of advanced statistical methods. Several analytic methods exist for analyzing longitudinal data, including repeated- measures ANOVA, generalized mixed models, structural equation modeling, time series methods, and survival analysis (Jaccard & Wan, 1993; Kraemer & Thiemann, 1989). The ANOVA approach is useful when the outcome variable is continuous, such as the total score on a depressed mood scale, but assumes normality and requires categorization of independent variables. Generalized mixed models are a more general form of linear regression and can be used for growth curve analyses, which consider repeated measures of an outcome, such as depressed mood, as a function of time, initial status and rate of growth (Chou, Bentler, & Pentz, 1998). Structural equation modeling provides a method for examining relationships between biological and psychosocial variables by examining the additive effects of multiple variables on the outcome of interest, including the determination of direct and indirect pathways. Time series analysis is useful for correlated data and can be used to examine multiple BPS factors in a single case series (Jaccard & Wan, 1993). 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Survival analysis is useful if the outcome of interest is the time to a certain event, such as the time to recurrent episodes of depression. Some researchers already have begun to apply advanced methods to understanding heterogeneous responses of depressed patients to antidepressants and placebos (Moller, Blank, & Steinmeyer, 1989; Stangl & Greenhouse, 1998). In addition, development of nonlinear mathematical models may be necessary to understanding BPS relationships in diseases such as MDD (Goldberger, 1999; Solomon, 1997). 5.4.4. Reassess diagnostic criteria The final phase toward developing a BPS theory of MDD is to reassess diagnostic criteria based upon findings from BPS studies. Currently, there is no biological marker with sufficient sensitivity and specificity to aid in the diagnosis of MDD. Instead, the diagnosis of MDD is based upon the presence, severity and duration of clinical symptoms as described in the Diagnostic and Statistical Manual Fourth Edition (DSM-IY) of the American Psychiatric Association (Association, 1994). Critics of the DSM-IV claim that it has many psychometric problems that limit both the validity and reliability of diagnoses (Carson, 1997; Greenberg & Fisher, 1997; Jones, 1998; Wakefield, 1998). In support of their position, critics suggest that the increasing need for “co-morbid” diagnoses may reflect erroneous categories that have failed to describe broader underlying syndromes (Brown, Chorpita, & Barlow, 1998; Carson, 1997). Furthermore, there is concern that the DSM-IV criteria do not sufficiently describe a patient’s complaints, and therefore are 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of limited usefulness in determining and evaluating effective treatments (Carson, 1997; Jones, 1998). Since accurate diagnosis of MDD is essential to correctly identifying BPS factors associated with MDD, it is critical that DSM-IV criteria be re-evaluated periodically. Reviewing BPS data and testing BPS hypotheses will lead to a better understanding of the heterogeneity in samples of patients with MDD, and this information could be used to reassess and redefine DSM-IV criteria as needed. Repeating phases of the research cycle will ultimately lead to a BPS theory of MDD that could help explain clinical observations, predict risk of MDD, and lead to more specific and effective interventions. A summary of a research cycle for developing a BPS theory of MDD is illustrated in Figure 5.2. Figure 5.2. A Research Cycle for Developing a BPS Theory of MDD Review data on BPS factors Generate BPS hypotheses Conduct multi factorial studies Reassess diagnostic criteria 5.5. Summary The biomedical approach emphasizes biological theories and the belief that advances in neurobiology will help to understand diseases such as MDD. Much of 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. current research on MDD is guided by the biomedical approach. However, a growing amount of evidence suggests that a BPS approach is needed in order to integrate knowledge about biological and psychosocial factors into a comprehensive theory of MDD. A specific BPS theory of MDD can be developed by reviewing data on BPS factors in MDD, generating BPS hypotheses, conducting multifactorial studies, and reassessing diagnostic criteria as needed. The combination of multifactorial studies and advanced statistical methods will provide data about associations between biological and psychosocial factors, which ultimately will lead to a better understanding of MDD. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.6. References Adler, N., & Matthews, K. (1994). Health Psychology: Why do some people get sick and some stay well? Annual Reviews in Psychology, 45, 229-259. Andreasen, N. (1997). Linking mind and brain in the study of mental illnesses: A project for a psychopathology. Science, 275, 1586-1593. Andreasen, N., & Black, D. (1995). Introductory Textbook o f Psychiatry (2nd ed.). Washington, DC: American Psychiatric Press, Inc. Antonovsky, A. (1989). Islands rather than bridgeheads: The problematic status of the biopsychosocial model. Family Systems Medicine, 7(3), 243-253. Arborelius, L., Owens, M., Plotsky, P., & Nemeroff, C. (1999). The role of corticotropin-releasing factor in depression and anxiety disorders. Journal o f Endocrinology, 160, 1-12. Association, A. P. (1994). Diagnostic and Statistical Manual o f Mental Disorders (Fourth (DSM-IV) ed ). Washington, DC: Author. Bebbington, P. (1987). Misery and beyond: The pursuit of disease theories of depression. The International Journal o f Social Psychiatry, 55(1), 13-20. Bedi, R. P. (1999). Depression: an inability to adapt to one's perceived life distress? Journal o f Affective Disorders, 54(1-2), 225-34. Blumenthal, J. A., Babyak, M. A., Moore, K. A., Craighead, W. E., Herman, S., Khatri, P., Waugh, R., Napolitano, M. A., Forman, L. M., Appelbaum, M., Doraiswamy, P. M., & Krishnan, K. R. (1999). Effects of exercise training on older patients with major depression. Archives o f Internal Medicine, 759(19), 2349-56. Brody, A. L., Barsom, M. W., Bota, R. G., & Saxena, S. (2001). Preffontal- subcortical and limbic circuit mediation of major depressive disorder. Seminars in Clinical Neuropsychiatry, 6(2), 102-12. Brody, H. (1997). The doctor as therapeutic agent: a placebo effect research agenda. In E. Harrington (Ed.), The Placebo Effect: An Interdisciplinary Exploration (pp. 77-92). Cambridge: Harvard University Press. Brown, T. A., Chorpita, B. F., & Barlow, D. H. (1998). Structural relationships among dimensions of the DSM-IV anxiety and mood disorders and dimensions of negative affect, positive affect, and autonomic arousal. Journal o f Abnormal Psychology, 107(2), 179-92. 138 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Carson, R. (1997). Costly compromises: a critique of the Diagnostic and Statistical Manual of Mental Disorders. In S. Fisher & R. Greenberg (Eds.), From Placebo to Panacea: Putting Psychiatric Drugs to the Test (pp. 98-114). New York: John Wiley & Sons, Inc. Champion, L. A., & Power, M. J. (1995). Social and cognitive approaches to depression: towards a new synthesis. British Journal o f Clinical Psychology, 34(Pt 4), 485-503. Checkley, S. (1996). The neuroendocrinology of depression and chronic stress. Br M ed Bull, 52, 597-617. Chou, C.-P., Bentler, P., & Pentz, M. (1998). Comparisons of Two Statistical Approaches to Study Growth Curves: The multilevel model and the latent curve analysis. Structural Equation Modeling, 5(3), 247-266. Clarke, D. M. (1998). Psychological factors in illness and recovery. New Zealand Medical Journal, 777(1076), 410-2. Cohen-Cole, S., & Levinson, R. (1990). The biopsychosocial model in medical practice. In A. Stoudemire (Ed.), Human behavior: An introduction fo r medical students (pp. 22-63). Philadelphia, PA: J. B. Lippincott Company. Connor, T. J., & Leonard, B. E. (1998). Depression, stress and immunological activation: The role of cytokines in depressive disorders. Life Sciences, 62(7), 583-606. Coyne, J. C., & Whiffen, Y. E. (1995). Issues in personality as diathesis for depression: the case of sociotropy-dependency and autonomy-self-criticism. Psychological Bulletin, 118(f), 358-78. Culbertson, F. (1997). Depression and gender. An international review. American Psychologist, 52(1), 25-31. Drevets, W. C. (2001). Neuroimaging and neuropathological studies of depression: implications for the cognitive-emotional features of mood disorders. Current Opinion in Neurobiology, 77(2), 240-9. Drossman, D. A. (1998). Gastrointestinal illness and the biopsychosocial model. Psychosomatic Medicine, 60(3), 258-267. Eaker, E, D. (1998). Psychosocial risk factors for coronary heart disease in women. Cardiology Clinics, 16(1), 103-11. 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Engel, G. (1977). The need for a new medical model: A challenge for biomedicine. Science, 796(4286), 129-136. Engel, G. (1980). The clinical application of the biopsychosocial model. American Journal o f Psychiatry, 137(5), 535-544. Fava, M., & Kendler, K. S. (2000). Major depressive disorder. Neuron, 28(2), 335- 41. Fisher, S., & Greenberg, R. (1997). The curse of the placebo: fanciful pursuit of a pure biological therapy. In S. Fisher & R. Greenberg (Eds.), From Placebo to Panacea: Putting Psychiatric Drugs to the Test (pp. 3-56). New York: John Wiley & Sons, Inc. Flett, G., Vredenburg, K., & Krames, L. (1997). The continuity of depression in clinical and nonclinical samples. Psychological Bulletin, 121(3), 395-416. Friedman, H. (Ed.). (1990). Personality and Disease. New York: John Wiley and Sons. Gabbard, G. (1994). Mind and brain in psychiatric treatment. Bulletin o f the Menninger Clinic, 5#(4), 427-446. Goldberger, A. (1999). Nonlinear dynamics, fractals, and chaos theory: Implications for neuroautonomic heart rate control in health and disease. In C. Bolis & J. Licinio (Eds.), The Autonomic Nervous System (pp. 135-152). Geneva: World Health Organization. Goodwin, F. (1993). Predictors of antidepressant response. Bulletin o f the Menninger Clinic, 57(2), 146-160. Greenberg, R , & Fisher, S. (1997). Mood-mending medicines: probing drug, psychotherapy, and placebo solutions. In S. Fisher & R. Greenberg (Eds.), From Placebo to Panacea: Putting Psychiatric Drugs to the Test (pp. 115- 172). New York: John Wiley & Sons, Inc. Gruen, R. (1993). Stress and Depression: Toward the development of integrative models. In L. Goldberger & S. Breznitz (Eds.), Handbook o f Stress: Theoretical and Clinical Aspects . New York: The Free Press. Heim, C., Owens, M., Plotsky, P., & Nemeroff, C. (1997). The role of early adverse life events in the etiology of depression and posttraumatic stress disorder: Focus on corticotropin-releasing factor. Annals New York Academy o f Sciences, 821, 194-207. 140 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Herbert, T., & Cohen, S. (1993). Depression and immunity: a meta-analytic review. Psychological Bulletin, 113, 472-86. Holden, R., Pakula, I., & Mooney, P. (1998). An immunological model connecting the pathogenesis of stress, depression and carcinoma. Medical Hypotheses, 57(4), 309-14. Irwin, M. (1995). Psychoneuroimmunology of depression. In F. Bloom & D. Kupfer (Eds.), Psychopharmacology: The Fourth Generation o f Progress (pp. 983- 998). New York: Raven Press, Ltd. Irwin, M., Caldwell, C., Smith, T., Brown, S., Schuckit, M., & Gillin, C. (1990). Major depressive disorder, alcoholism, and reduced natural killer cell cytotoxicity. Archives o f General Psychiatry, 47, 713-719. Jaccard, J., & Wan, C. (1993). Statistical analysis of temporal data with many observations: issues for behavioral medicine data. Annals o f Behavioral Medicine, 75(1), 41-50. Johnson, J. G., Han, Y. S., Douglas, C. J., Johannet, C. M., & Russell, T. (1998). Attributions for positive life events predict recovery from depression among psychiatric inpatients: an investigation of the Needles and Abramson model of recovery from depression. Journal o f Consulting & Clinical Psychology, 66(2), 369-76. Jones, E. (1998). Depression: Intervention as assessment. In J. Barron (Ed ), Making Diagnosis Meaningful: Enhancing evaluation and treatment o f psychological disorders (pp. 267-297). Washington, DC: American Psychological Association. Kagan, J. (1994). Galen's Prophecy: Temperament in Human Nature. New York:: Basic Books. Kendler, K., Kessler, R , Neale, M., Heath, A., & Eaves, L. (1993). The prediction of major depression in women: Toward an integrated etiologic model. American Journal o f Psychiatry, 150, 1139-1148. Kendler, K., Thornton, L., & Gardner, C. (2000). Stressful life events and previous episodes in the etiology of major depression in women: an evaluation of the "kindling" hypothesis. American Journal o f Psychiatry, 157(8), 1243-1251. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Khan, A., Warner, H., & Brown, W. (2000). Symptom reduction and suicide risk in patients treated with placebo in antidepressant clinical trials: An analysis of the Food and Drug Administration database. Archives o f General Psychiatry, 57, 311-317. Kohn, Y., Zislin, J., Agid, O., Hanin, B., Troudart, T., Shapira, B., Bloch, M., Gur, E., Ritsner, M., & Lerer, B. (2001). Increased prevalence of negative life events in subtypes of major depressive disorder. Comprehensive Psychiatry, 42( 1), 57-63. Kop, W. J. (1997). Acute and chronic psychological risk factors for coronary syndromes: Moderating effects of coronary artery disease severity. Journal o f Psychosomatic Research, 43(2), 167-181. Kraemer, H. C., Gullion, C. M., Rush, A. J., Frank, E., & Kupfer, D. J. (1994). Can state and trait variables be disentangled? A methodological framework for psychiatric disorders. Psychiatry Research, 52(1), 55-69. Kraemer, H. C., & Telch, C. F. (1992). Selection and utilization of outcome measures in psychiatric clinical trials. Report on the 1988 MacArthur Foundation Network I Methodology Institute. Neuropsychopharmacology, 7(2), 85-94. Kraemer, H. C., & Thiemann, S. (1989). A strategy to use soft data effectively in randomized controlled clinical trials. Journal o f Consulting & Clinical Psychology, 57(1), 148-54. Krishnan, K., Gadde, K., & Kim, Y. (1998). Psychoneuroendocrinology and brain imaging in depression. Psychoneuroendocrinology, 21(2), 465-472. Kupfer, D. (1995). Sleep research in depressive illness: clinical implications— a tasting menu. Biological Psychiatry, 38(6), 391-403. Lane, D., Carroll, D., & Lip, G. Y. (1999). Psychology in coronary care. Qjm, 92(8), 425-31. Lara, M. E., & Klein, D. N. (1999). Psychosocial processes underlying the maintenance and persistence of depression: implications for understanding chronic depression. Clinical Psychology Review, 19(5), 553-70. Lavoie, K., & Fleet, R. (2000). The impact of depression on the course and outcome of coronary artery disease, review for cardiologists. Canadian Journal o f Cardiology, 16(5), 653-662. 142 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Lehtinen, V., & Joukamaa, M. (1994). Epidemiology of depression: Prevalence, risk factors and treatment situation. ActaPsychiatr Scand, Suppl 377, 7-10. Leigh, H., & Reiser, M. (1985). The Patient: Biological, psychological, and social dimensions o f medical practice. New York: Plenum Press. Lepore, S. (1997). Measurement of Chronic Stressors. In S. Cohen, R. Kessler, & L. Gordon (Eds.), Measuring Stress: A Guide fo r Health and Social Scientists (pp. 102-120). Oxford: Oxford University Press. Leuchter, A., Cook, I., Uijtdehaage, S., Dunkin, J., Lufkin, R., Anderson-Hanley, C., Abrams, M., Rosenberg-Thompson, S., O'Hara, R., Simon, S., Osato, S., & Babaie, A. (1997). Brain structure and function and the outcomes of treatment for depression. Journal o f Clinical Psychiatry, 58( 16), 22-31. Levin, D., & Solomon, G. (1990). The discursive formation of the body in the history of medicine. Journal o f Medicine and Philosophy, 15, 515-537. Light, K. C., Dolan, C. A., Davis, M. R., & Sherwood, A. (1992). Cardiovascular responses to an active coping challenge as predictors of blood pressure patterns 10 to 15 years later. Psychosomatic Medicine, 54(2), 217-30. Marusic, A., Gudjonsson, G. H., Eysenck, H. J., & Stare, R. (1999). Biological and psychosocial risk factors in ischaemic heart disease: Empirical findings and a biopsychosocial model. Personality and Individual Differences, 26(2), 285- 304. Maughan, B., & McCarthy, G. (1997). Childhood adversities and psychosocial disorders. British Medical Bulletin, 55(1), 156-169. McDaniel, J., Musselman, D., Porter, M., Reed, D., & NemerofF, C. (1995). Depression in patients with cancer: Diagnosis, biology, and treatment. Archives o f General Psychiatry, 52, 89-99. McDaniel, S. (1995). Collaboration between psychologists and family physicians: Implementing the biopsychosocial model. Professional Psychology: Research and Practice, 26(2), 117-122. McHugh, S., & Vallis, M. (1986). Illness behaviour: Operationalization o f the biopsychosocial model. In S. McHugh & T. Vallis (Eds.), Illness behavior: A multidisciplinary model (pp. 1-31). New York, NY: Plenum Press. McLaren, N. (1998). A critical review of the biopsychosocial model. Australian and New Zealand Journal o f Psychiatry, 52(1), 86-92. 143 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Miser, W. F. (2000). Exercise as an effective treatment option for major depression in older adults. Journal o f Family Practice, 49(2), 109-10. Moller, H. J., Blank, R., & Steinmeyer, E. M. (1989). Single-case evaluation of sleep-deprivation effects by means of nonparametric time-series analysis (according to the HTAKA model). European Archives o f Psychiatry & Neurological Sciences, 239(2), 133-9. Mullen, P. E. (1998). "A Critical Review of the Biopsychosocial Model": Comment. Australian and New Zealand Journal o f Psychiatry, 32(1), 95-96. Murray, P., Rosenthal, K., Kobayashi, G., & Pfaller, M. (1998). Medical Microbiology (3rd ed.). St. Louis: Mosby. Musselman, D. L., Evans, D. L., & Nemeroff, C. B. (1998). The relationship of depression to cardiovascular disease: Epidemiology, biology, and treatment. Archives o f General Psychiatry, 55(7), 580-592. Naisberg, Y. (1996). Homeostatic disruption and depression. Medical Hypotheses, 47(6), 415-22. Nemeroff, C., Heim, C., Owens, M., Newport, D., Miller, A., & Plotsky, P. (1999,). Neurochemical mechanisms underlying depression and anxiety disorders: The influence of early trauma. New Developments in Understanding Depression and Its Treatment, pp. 3-6. Paris, J. (1993). Personality disorders: A biopsychosocial model. Journal o f Personality Disorders, 7(3), 255-264. Paykel, E. S. (1994). Life events, social support and depression. Acta Psychiatrica Scandinavica, Supplementum, 377, 50-8. Rozanski, A., Blumenthal, J. A., & Kaplan, J. (1999). Impact of psychological factors on the pathogenesis of cardiovascular disease and implications for therapy. Circulation, 99(16), 2192-217. Sadler, J. Z., & Hulgus, Y. F. (1990). Knowing, valuing, acting: Clues to revising the biopsychosocial model. Comprehensive Psychiatry, 31(3), 185-195. Schoenfeld Smith, K., Petroski, G. F., Hewett, J. E., Johnson, J. C., Wright, G. E., Smarr, K. L., Walker, S. E., & Parker, J. C. (1996). A biopsychosocial model of disability in rheumatoid arthritis. Arthritis Care and Research, 9(5), 368- 375. 144 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Schulberg, H. C., Katon, W., Simon, G. E., & Rush, A. J. (1998). Treating major depression in primary care practice: An update of the Agency for Health Care Policy and Research practice guidelines. Archives o f General Psychiatry, 55(12), 1121-1127. Schwartz, G. (1982). Testing the biopsychosocial model: The ultimate challenge facing behavioral medicine? Journal o f Consulting and Clinical Psychology, 50, 1040-1053. Schwartzman, J., & Glaus, K. (2000). Depression and coronary heart disease in women: Implications for clinical practice and research. Professional Psychology: Research and Practice, 57(1), 48-57. Segerstrom, S.-C. (1997). Optimism and coping with enduring stress: Psychological and immune effects. Dissertation Abstracts International: Section B: The Sciences and Engineering, 55(6-B), 3355. Shapiro, A., & Shapiro, E. (1997). The placebo: is it much ado about nothing? In E. Harrington (Ed.), The Placebo Effect: An Interdisciplinary Exploration (pp. 12-36). Cambridge: Harvard University Press. Solomon, G. (1997). Stress, hormones, immunity, the complexity of intertwined systems, and the simplicity of humanism. Journal o f Intensive Care Medicine, 12, 219-222. Spruijt-Metz, D. (1999). Adolescence, affect and health. London: Psychology Press. Stangl, D. K., & Greenhouse, J. B. (1998). Assessing placebo response using Bayesian hierarchical survival models. Lifetime Data Analysis, 4( 1), 5-28. Stoudemire, A. (Ed.). (1995). Psychological Factors Affecting Medical Conditions. Washington, DC: American Psychiatric Press, Inc. Taylor, S. (1999). Health Psychology (4th ed.). Boston: McGraw Hill. Temoshok, L. (1990). On attempting to articulate the biopsychosocial model: Psychological-psychophysiological homeostasis. In H. S. Friedman (Ed.), Personality and disease, (pp. 203-225). New York: John Wiley & Sons. Tewes, U. (1999). Concepts in Psychology. In M. Schedlowski & U. Tewes (Eds ), Psychoneuroimmunology: An interdisciplinary introduction (pp. 93-111). New York: Kluwer Academic/Plenum Publishers. 145 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Tilders, F. J., & Schmidt, E. D. (1999). Cross-sensitization between immune and non-immune stressors. A role in the etiology of depression? Advances in Experimental Medicine & Biology, 461, 179-97. Uchino, B.-N., Cacioppo, J.-T., & Keicolt-Glaser, J.-K. (1996). The relationship between social support and physiological processes: A review with emphasis on underlying mechanisms and implications for health. Psychological Bulletin, 119(3), 488-531. Wade, T. D., & Kendler, K. S. (2000). The relationship between social support and major depression: cross-sectional, longitudinal, and genetic perspectives. Journal o f Nervous & Mental Disease, 188(5), 251-8. Wakefield, J. (1998). Meaning and melancholia: Why the DSM-IV cannot (entirely) ignore the patients intentional system. In J. Barron (Ed.), Making diagnosis meaningful: Enhancing evaluation and treatment o f psychological disorders (pp. 29-72). Washington, DC: American Psychological Association. Wallace, E. (1990). Mind-body and the future of psychiatry. Journal o f Medicine and Philosophy, 75(1), 41-74. Weissman, S. (1999). Psychiatric Diagnosis. In S. Weissman, M. Sabshin, & H. Eist (Eds.), Psychiatry in the New Millenium (pp. 57-66). Washington, DC: American Psychiatric Press, Inc. Whybrow, P. (1997). A Mood Apart: A thinker's guide to emotion and its disorders. New York: HarperPerennial. Whybrow, P., Akiskal, H., & McKinney, W. J. (1984). Toward a psychobiological integration: affective illness as a final common path to adaptive failure, Mood Disorders: Toward a New Psychobiology (pp. 173-203). New York: Plenum. Willner, P. (1985). Depression: A psychobiological synthesis. New York: John Wiley and Sons. Wolkenstein, A., & Butler, D. (1998). The pre-encounter psychosocial review: The anticipatory integration of the biopsychosocial model in primary care. Families, Systems and Health, 16(A), 337-346. Zimmermann, C., & Tansella, M. (1996). Psychosocial factors and physical illness in primary care: Promoting the biopsychosocial model in medical practice. Journal o f Psychosomatic Research, 40(4), 351-358. 146 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. BIBLIOGRAPHY Abou-Saleh, M., & Coppen, A. (1983). Classification of depression and response to antidepressive therapies. British Journal o f Psychiatry, 143, 601-603. Adler, N., & Matthews, K. (1994). Health Psychology: Why do some people get sick and some stay well? Annual Reviews in Psychology, 45, 229-259. Alpert, J., & Fava, M. (1997). Nutrition and depression: the role of folate. [Review], Nutrition Reviews, 55(5), 145-9. Andreasen, N. (1997). Linking mind and brain in the study of mental illnesses: A project for a psychopathology. Science, 275, 1586-1593. Andreasen, N., & Black, D. (1995). Introductory Textbook o f Psychiatry (2nd ed.). Washington, DC: American Psychiatric Press, Inc. Antonovsky, A. (1989). Islands rather than bridgeheads: The problematic status of the biopsychosocial model. Family Systems Medicine, 7(3), 243-253. Antonuccio, D., Danton, W., & DeNelsky, G. (1995). Psychotherapy versus medication for depression: Challenging the conventional wisdom with data. Professional Psychology: Research and Practice, 26, 574-585. Arborelius, L., Owens, M., Plotsky, P., & Nemeroff, C. (1999). The role of corticotropin-releasing factor in depression and anxiety disorders. Journal o f Endocrinology, 160, 1-12. Association, A. P. (1994). Diagnostic and Statistical Manual o f Mental Disorders (Fourth (DSM-IV) ed.). Washington, DC: Author. Barlow, D. H., & Campbell, L. A. (2000). Mixed anxiety-depression and its implications for models of mood and anxiety disorders. Comprehensive Psychiatry, 41(2 Suppl 1), 55-60. Barrett, J. E., Williams, J. W., Jr., Oxman, T. E., Katon, W., Frank, E., Hegel, M. T., Sullivan, M., & Schulberg, H. C. (1999). The treatment effectiveness project. A comparison of the effectiveness of paroxetine, problem-solving therapy, and placebo in the treatment of minor depression and dysthymia in primary care patients: background and research plan. General Hospital Psychiatry, 21(4), 260-73. Bebbington, P. (1987). Misery and beyond: The pursuit of disease theories of depression. The International Journal o f Social Psychiatry, 33(1), 13-20. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Beck, A. (1967). Depression: Clinical, experimental, and theoretical aspects. New York: Harper & Row. Beck, A., Rush, A., Shaw, B., & Emery, G. (1979). Cognitive therapy o f depression. New York: Guilford Press. Beck, A., Steer, R., & Garbin, M. (1988). Psychometric properties of the Beck Depression Inventory: Twenty-five years of evaluation. Clinical Psychology Review, 8, 77-100. Beck, A., Ward, C., Mendelsohn, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression. Archives o f General Psychiatry, 4, 561-571. Bedi, R. P. (1999). Depression: an inability to adapt to one's perceived life distress? Journal o f Affective Disorders, 54(1-2), 225-34. Benschop, R., & Schedlowski, M. (1999). Acute psychological stress. InM. Schedlowski & U. Tewes (Eds.), Psychoneuroimmunology: An interdisciplinary introduction (pp. 293-306). New York: Kluwer Academic/Plenum Publishers. Bialik, R. J., Ravindran, A. V., Bakish, D., & Lapierre, Y. D. (1995). A comparison of placebo responders and nonresponders in subgroups of depressive disorder. Journal o f Psychiatry & Neuroscience, 20(4), 265-70. Blumenthal, J. A., Babyak, M. A., Moore, K. A., Craighead, W. E., Herman, S., Khatri, P., Waugh, R., Napolitano, M. A., Forman, L. M., Appelbaum, M., Doraiswamy, P. M., & Krishnan, K. R. (1999). Effects of exercise training on older patients with major depression. Archives o f Internal Medicine, 759(19), 2349-56. Bottiglieri, T. (1996). Folate, vitamin B12, and neuropsychiatric disorders. Nutrition Reviews, 54(12), 382-390. Brody, A. L., Barsom, M. W., Bota, R. G., & Saxena, S. (2001). Prefrontal- subcortical and limbic circuit mediation of major depressive disorder. Seminars in Clinical Neuropsychiatry, 6(2), 102-12. Brody, H. (1980). Placebos and the philosophy o f medicine: Clinical, conceptual, and ethical issues. Chicago: University o f Chicago Press. Brody, H. (1997). The doctor as therapeutic agent: a placebo effect research agenda. In E. Harrington (Ed.), The Placebo Effect: An Interdisciplinary Exploration (pp. 77-92). Cambridge: Harvard University Press. 148 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Brosschot, J., Benschop, R., Godaert, G., de Smet, M., Olff, M., Heijnen, C., & Ballieux, R. (1992). Effects of experimental psychological stress on distribution and function of peripheral blood cells. Psychosomatic Medicine, 54, 396-406. Brown, G., & Bifulco, A. (1985). Social Support, Life Events and Depression. In I. Saronson & B. Saronson (Eds.), Social Support: Theory, Research, and Applications (pp. 349-370). Boston, MA: Martinus Nijhoff. Brown, T. A., Chorpita, B. F., & Barlow, D. H. (1998). Structural relationships among dimensions of the DSM-IV anxiety and mood disorders and dimensions of negative affect, positive affect, and autonomic arousal. Journal o f Abnormal Psychology, 107(2), 179-92. Brown, W., Johnson, M., & Chen, M. (1992). Clinical features of depressed patients who do and do not improve with placebo. Psychiatry Research, 41, 203-214. Brummett, B. H., Babyak, M. A., Barefoot, J. C., Bosworth, H. B., Clapp-Channing, N. E., Siegler, I. C., Williams, R. B., Jr., & Mark, D. B. (1998). Social support and hostility as predictors of depressive symptoms in cardiac patients one month after hospitalization: a prospective study. Psychosomatic Medicine, 60(6), 707-13. Buysse, D., Reynolds, C., Monk, T., Berman, S., & Kupfer, D. (1989). The Pittsburgh Sleep Quality Index. Psychiatry Research, 28, 193-213. Byrne, A., & Byrne, D. (1993). The effect of exercise on depression, anxiety and other mood states: a review. Journal o f Psychosomatic Research, 17(6), 565- 574. Cacioppo, J. (1994). Social neuroscience: Autonomic, neuroendocrine, and immune response to stress. Psychophysiology, 31, 113-128. Cacioppo, J., Poelmann, K., Kiecolt-Glaser, J., Malarkey, W., Burleson, M., Bemston, G., & Glaser, R. (1998). Cellular immune responses to acute stress in female caregivers of dementia patients and matched controls. Health Psychology, 17(2), 182-189. Carney, R. M., Freedland, K. E., Veith, R. C., Cryer, P. E., Skala, J. A., Lynch, T., & Jaffe, A. S. (1999). Major depression, heart rate, and plasma norepinephrine in patients with coronary heart disease. Biological Psychiatry, 45(4), 458-63. 149 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Carson, R. (1997). Costly compromises: a critique of the Diagnostic and Statistical Manual of Mental Disorders. In S. Fisher & R. Greenberg (Eds.), From Placebo to Panacea: Putting Psychiatric Drugs to the Test (pp. 98-114). New York: John Wiley & Sons, Inc. Champion, L. A., & Power, M. J. (1995). Social and cognitive approaches to depression: towards a new synthesis. British Journal o f Clinical Psychology, 34(?X 4), 485-503. Checkley, S. (1996). The neuroendocrinology of depression and chronic stress. Br M ed Bull, 52, 597-617. Chou, C.-P., Bentler, P., & Pentz, M. (1998). Comparisons of Two Statistical Approaches to Study Growth Curves: The multilevel model and the latent curve analysis. Structural Equation Modeling, 5(3), 247-266. Christensen, L., & Somers, S. (1996). Comparison of nutrient intake among depressed and nondepressed individuals. International Journal o f Eating Disorders, 20(1), 105-9. Clarke, D. M. (1998). Psychological factors in illness and recovery. New Zealand Medical Journal, 777(1076), 410-2. Cloninger, C., Svrakic, D., & Przybeck, T. (1993). A psychobiological model of temperament and character. Archives o f General Psychiatry, 50, 975-990. Cohen, S., Frank, E., Doyle, W. J., Skoner, D. P., Rabin, B. S., & Gwaltney, J. M., Jr. (1998). Types of stressors that increase susceptibility to the common cold in healthy adults. Health Psychology, 77(3), 214-23. Cohen, S., & Herbert, T. (1996). Health Psychology: Psychological factors and physical disease from the perspective of human psychoneuroimmunology. Annual Reviews in Psychology, 47, 113-142. Cohen, S., Mermelstein, R., Kamarck, T., & Hoberman, H. (1985). Measuring the functional components of social support. In I. Saronson & B. Saronson (Eds), Social Support: Theory, Research, and Applications (pp. 73-94). Boston, MA: Martinus Nijhoff. Cohen, S. I. (1986). Measurement of human adaptation to stressful environments— revisiting a 1958 presentation to the American Rocket Society— revising the stress model to a biopsychosocial systems model. American Journal o f Social Psychiatry, 6(3), 175-182. 150 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cohen-Cole, S., & Levinson, R. (1990). The biopsychosocial model in medical practice. In A. Stoudemire (Ed.), Human behavior: An introduction for medical students (pp. 22-63). Philadelphia, PA: J. B. Lippincott Company. Connor, T. J., & Leonard, B. E. (1998). Depression, stress and immunological activation: The role of cytokines in depressive disorders. Life Sciences, 62(1), 583-606. Conte, H., Plutchik, R., Wild, K , & Karasu, T. (1986). Combined psychotherapy and pharmacotherapy for depression: A systematic analysis of the evidence. Archives o f General Psychiatry, 43, 471-479. Cook, I., Leuchter, A., Witte, E., Stubbeman, W., Abrams, M., & Rosenberg, S. (in press). Early changes in preffontal activity characterize clinical responders to antidepressants. . Coyne, J. C., & Whiffen, V. E. (1995). Issues in personality as diathesis for depression: the case of sociotropy-dependency and autonomy-self-criticism. Psychological Bulletin, 118(3), 358-78. Culbertson, F. (1997). Depression and gender. An international review. American Psychologist, 52(1), 25-31. Deckersbach, T., Gershuny, B. S., & Otto, M. W. (2000). Cognitive-behavioral therapy for depression. Applications and outcome. Psychiatric Clinics o f North America, 25(4), 795-809, VII. Delehanty, S. G., Dimsdale, J. E., & Mills, P. (1991). Psychosocial correlates of reactivity in black and white men. Journal o f Psychosomatic Research, 35(4- 5), 451-60. Derogatis, L. (1977). Administration, Scoring and Procedures Manual fo r the SCL- 90-R. Baltimore: Clinical Psychometrics Research. Dimsdale, J. (1988). The effect of depression on cardiovascular reactivity. In T. Field, P. McCabe, & e. al (Eds.), Stress and coping across development (pp. 215-225). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Dobson, K. (1989). A meta-analysis of the efficacy of cognitive therapy for depression. Journal o f Consulting and Clinical Psychology, 57, 414-419. Drevets, W. (2000). Functional anatomical abnormalities in limbic and prefrontal cortical structures in major depression. Progress in Brain Research, 126, 413-431. 151 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Drevets, W. C. (2001). Neuroimaging and neuropathological studies of depression: implications for the cognitive-emotional features of mood disorders. Current Opinion in Neurobiology, 11(2), 240-9. Drossman, D. A. (1998). Gastrointestinal illness and the biopsychosocial model. Psychosomatic Medicine, 60(3), 258-267. Eaker, E. D. (1998). Psychosocial risk factors for coronary heart disease in women. Cardiology Clinics, 16( 1), 103-11. Engel, G. (1977). The need for a new medical model: A challenge for biomedicine. Science, 196(4286), 129-136. Engel, G. (1980). The clinical application of the biopsychosocial model. American Journal o f Psychiatry, 137(5), 535-544. Evans, F. (1985). Expectancy, therapeutic instructions, and the placebo response. In L. White, B. Tursky, & G. Schwartz (Eds.), Placebo: Theory, Research, and Mechanisms (pp. 215-228). New York: Guilford Press. Fava, M., & Kendler, K. S. (2000). Major depressive disorder. Neuron, 28(2), 335- 41. Fisher, S., & Greenberg, R. (1997). The curse of the placebo: fanciful pursuit of a pure biological therapy. In S. Fisher & R. Greenberg (Eds.), From Placebo to Panacea: Putting Psychiatric Drugs to the Test (pp. 3-56). New York: John Wiley & Sons, Inc. Flett, G., Vredenburg, K., & Krames, L. (1997). The continuity of depression in clinical and nonclinical samples. Psychological Bulletin, 121(3), 395-416. Fombonne, E. (1994). Increased rates of depression: update of epidemiological findings and analytical problems. Acta Psychiatrica Scandinavica, 90(3), 145-56. Friedman, H. (Ed.). (1990). Personality and Disease. New York: John Wiley and Sons. Futterman, A., Kemeny, M., Shapiro, D., & Fahey, J. (1994). Immunological and physiological changes associated with induced positive and negative mood. Psychosomatic Medicine, 56, 499-511. Gabbard, G. (1994). Mind and brain in psychiatric treatment. Bulletin o f the Menninger Clinic, 58(4), 427-446. 152 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Gaynes, B. N., Magruder, K. M., Burns, B. J., Wagner, H. R., Yarnall, K. S., & Broadhead, W. E. (1999). Does a coexisting anxiety disorder predict persistence of depressive illness in primary care patients with major depression? General Hospital Psychiatry, 21(3), 158-67. Gillett, G. (1990). Neuropsychology and meaning in psychiatry. Journal o f Medicine and Philosophy, 75(1), 21-40. Godin, G., & Shepard, R. (1985). A simple method to assess exercise behavior in the community. Canadian Journal o f Applied Sport Science, 10(3), 141-146. Goldberger, A. (1999). Nonlinear dynamics, fractals, and chaos theory: Implications for neuroautonomic heart rate control in health and disease. In C. Bolis & J. Licinio (Eds.), The Autonomic Nervous System (pp. 135-152). Geneva: World Health Organization. Goodwin, F. (1993). Predictors of antidepressant response. Bulletin o f theMenninger Clinic, 57(2), 146-160. Gorman, J., & Sloan, R. (2000). Heart rate variability in depressive and anxiety disorders. American Heart Journal, 140(4), S77-S83. Greenberg, R., & Fisher, S. (1997). Mood-mending medicines: probing drug, psychotherapy, and placebo solutions. In S. Fisher & R. Greenberg (Eds ), From Placebo to Panacea: Putting Psychiatric Drugs to the Test (pp. 115- 172). New York: John Wiley & Sons, Inc. Gruen, R. (1993). Stress and Depression: Toward the development of integrative models. In L. Goldberger & S. Breznitz (Eds.), Handbook o f Stress: Theoretical and Clinical Aspects . New York: The Free Press. Guinjoan, S., Bernabo, J., & Cardinali, D. (1995). Cardiovascular tests of autonomic function and sympathetic skin responses in patients with major depression. Journal o f Neurology, Neurosurgery, and Psychiatry, 58, 299-302. Hamilton, M. (1959). The assessment of anxiety states by rating. Brit. J. o f Med. Psychol., 32, 50-59. Hamilton, M. (1967). Development of a rating scale for primary depressive illness. British Journal o f Social and Clinical Psychology, 6, 278-296. Harrington, E. (Ed.). (1997). The Placebo Effect: An Interdisciplinary Exploration. Cambridge: Harvard University Press. 153 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Heim, C., Owens, M., Plotsky, P., & Nemeroff, C. (1997). The role of early adverse life events in the etiology of depression and posttraumatic stress disorder: Focus on corticotropin-releasing factor. Annals New York Academy o f Sciences, 821, 194-207. Herbert, T., & Cohen, S. (1993). Depression and immunity: a meta-analytic review. Psychological Bulletin, 113, 472-86. Holden, R., Pakula, I., & Mooney, P. (1998). An immunological model connecting the pathogenesis of stress, depression and carcinoma. Medical Hypotheses, 57(4), 309-14. Hollon, S., Shelton, R., & Loosen, P. (1991). Cognitive therapy and pharmacotherapy for depression. Journal o f Consulting and Clinical Psychology, 59, 88-99. Irwin, M. (1995). Psychoneuroimmunology of depression. In F. Bloom & D. Kupfer (Eds.), Psychopharmacology: The Fourth Generation o f Progress (pp. 983- 998). New York: Raven Press, Ltd. Irwin, M., Caldwell, C., Smith, T., Brown, S., Schuckit, M., & Gillin, C. (1990). Major depressive disorder, alcoholism, and reduced natural killer cell cytotoxicity. Archives o f General Psychiatry, 47, 713-719. Irwin, M., & Friedman, E. (1999). Does Psychological Depression Cause Immune Suppression in Humans? In M. Schedlowski & U. Tewes (Eds.), Psychoneuroimmunology: An interdisciplinary introduction (pp. 327-340). New York: Kluwer Academic/Plenum Publishers. Jaccard, J., & Wan, C. (1993). Statistical analysis of temporal data with many observations: issues for behavioral medicine data. Annals o f Behavioral Medicine, 75(1), 41-50. Jemerin, J. M., & Boyce, W. T. (1990). Psychobiological differences in childhood stress response. II. Cardiovascular markers of vulnerability. Journal o f Developmental & Behavioral Pediatrics, 77(3), 140-50. Joffe, R., Sokolov, S., & Streiner, D. (1996). Antidepressant treatment of depression: A metaanalysis. Canadian Journal o f Psychiatry, 41, 613-616. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Johnson, J. G., Han, Y. S., Douglas, C. J., Johannet, C. M., & Russell, T. (1998). Attributions for positive life events predict recovery from depression among psychiatric inpatients: an investigation of the Needles and Abramson model of recovery from depression. Journal o f Consulting & Clinical Psychology, 66(2), 369-76. Jonas, B., Franks, P., & Ingram, D. (1997). Are symptoms of anxiety and depression risk factors for hypertension?: longitudinal evidence from the National Health and Nutrition Examination Survey I Epidemiologic Follow-up Study. Archives o f Family Medicine, 6(1), 43-49. Jones, E. (1998). Depression: Intervention as assessment. In J. Barron (Ed.), Making Diagnosis Meaningful: Enhancing evaluation and treatment o f psychological disorders (pp. 267-297). Washington, DC: American Psychological Association. Joyce, P., & Paykel, E. (1989). Predictors of drug response in depression. Archives o f General Psychiatry, 46, 88-89. Joyce, P. R., Mulder, R. T., & Cloninger, C. R. (1994). Temperament predicts clomipramine and desipramine response in major depression [see comments]. Journal o f Affective Disorders, 30(1), 35-46. Judd, L., AkiskaJ, H., Zeller, P., Paulus, M., Leon, A., Maser, J., Endicott, J., Coryell, W., Kunovac, J., Mueller, T., Rice, J., & Keller, M. (2000). Psychosocial disability during the long-term course of unipolar major depressive disorder. Archives o f General Psychiatry, 57(4), 375-80. Kagan, J. (1994). Galen's Prophecy: Temperament in Human Nature. New York:: Basic Books. Kendler, K., Kessler, R., Neale, M., Heath, A., & Eaves, L. (1993). The prediction of major depression in women: Toward an integrated etiologic model. American Journal o f Psychiatry, 150, 1139-1148. Kendler, K., Thornton, L., & Gardner, C. (2000). Stressful life events and previous episodes in the etiology of major depression in women: an evaluation of the "kindling" hypothesis. American Journal o f Psychiatry, 757(8), 1243-1251. Kessler, R. C., Zhao, S., Blazer, D. G., & Swartz, M. (1997). Prevalence, correlates, and course of minor depression and major depression in the National Comorbidity Survey. Journal o f Affective Disorders, -75(1-2), 19-30. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Khan, A., Warner, H., & Brown, W. (2000). Symptom reduction and suicide risk in patients treated with placebo in antidepressant clinical trials: An analysis of the Food and Drug Administration database. Archives o f General Psychiatry, 57, 311-317. Kirsch, I., & Sapirstein, G. (1998). Listening to Prozac and hearing placebo: a meta analysis of antidepressant medications. Prevention and Treatment, ./(article 2a), 1-17. Klaassen, T., Riedel, W.-J., van-Someren, A., Deutz, N.-E. P., Honig, A., & van- Praag, H.-M. (1999). Mood effects of 24-hour tryptophan depletion in healthy first-degree relatives of patients with affective disorders. Biological Psychiatry, 46(4), 489-497. Kocsis, J. (1990). New issues in the prediction of antidepressant. Psychopharmacology Bulletin, 26(1), 49-53. Kohn, Y., Zislin, J., Agid, O., Hanin, B., Troudart, T., Shapira, B., Bloch, M., Gur, E., Ritsner, M., & Lerer, B. (2001). Increased prevalence of negative life events in subtypes of major depressive disorder. Comprehensive Psychiatry, 42(1), 57-63. Kop, W. J. (1997). Acute and chronic psychological risk factors for coronary syndromes: Moderating effects of coronary artery disease severity. Journal o f Psychosomatic Research, 43(2), 167-181. Kraemer, H. C. (1991). To increase power in randomized clinical trials without increasing sample size. Psychopharmacology Bulletin, 27(3), 217-24. Kraemer, H. C., Gullion, C. M., Rush, A. J., Frank, E., & Kupfer, D. J. (1994). Can state and trait variables be disentangled? A methodological framework for psychiatric disorders. Psychiatry Research, 52(1), 55-69. Kraemer, H. C., & Telch, C. F. (1992). Selection and utilization of outcome measures in psychiatric clinical trials. Report on the 1988 MacArthur Foundation Network I Methodology Institute. Neuropsychopharmacology, 7(2), 85-94. Kraemer, H. C., & Thiemann, S. (1989). A strategy to use soft data effectively in randomized controlled clinical trials. Journal o f Consulting & Clinical Psychology, 57(1), 148-54. Krishnan, K., Gadde, K., & Kim, Y. (1998). Psychoneuroendocrinology and brain imaging in depression. Psychoneuroendocrinology, 21(2), 465-472. 156 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Kupfer, D. (1995). Sleep research in depressive illness: clinical implications— a tasting menu. Biological Psychiatry, 38(6), 391-403. Lane, D., Carroll, D., & Lip, G. Y. (1999). Psychology in coronary care. Qjm, 92(8), 425-31. Lara, M. E., & Klein, D. N. (1999). Psychosocial processes underlying the maintenance and persistence of depression: implications for understanding chronic depression. Clinical Psychology Review, 19(5), 553-70. Lavoie, K., & Fleet, R. (2000). The impact of depression on the course and outcome of coronary artery disease: review for cardiologists. Canadian Journal o f Cardiology, 16(5), 653-662. Lawlor, D., & Hopker, S. (2001). The effectiveness of exercise as an intervention in the management of depression: Systematic review and meta-regression analysis of randomised controlled trials. British Medical Journal, 322(7289), 763-766. Legato, M. (1997). Gender-specific physiology: how real is it? How important is it? International Journal o f Fertility & Womens Medicine, 42(1), 19-29. Lehtinen, V., & Joukamaa, M. (1994). Epidemiology of depression: Prevalence, risk factors and treatment situation. Acta Psychiatr Scand, Suppl 377, 7-10. Leigh, H., & Reiser, M. (1985). The Patient: Biological, psychological, and social dimensions o f medical practice. New York: Plenum Press. Lepore, S. (1997). Measurement of Chronic Stressors. In S. Cohen, R. Kessler, & L. Gordon (Eds.), Measuring Stress: A Guide fo r Health and Social Scientists (pp. 102-120). Oxford: Oxford University Press. Leserman, J., Jackson, E., Petitto, J., Golden, R., Silva, S., Perkins, D., Cai, J., Folds, J., & Evans, D. (1999). Progression to AIDS: The effects of stress, depressive symptoms, and social support. Psychosomatic Medicine, 61(3), 397-406. Leuchter, A., Cook, I., Uijtdehaage, S., Dunkin, J., Lufkin, R., Anderson-Hanley, C., Abrams, M., Rosenberg-Thompson, S., O'Hara, R , Simon, S., Osato, S., & Babaie, A. (1997). Brain structure and function and the outcomes of treatment for depression. Journal o f Clinical Psychiatry, 38(16), 22-31. Leuchter, A., Cook, I., Witte, E., & Abrams, M. Changes in brain function of depressed subjects during treatment with placebo, American Journal o f Psychiatry, 159(1), 122-129. 157 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Levin, D., & Solomon, G. (1990). The discursive formation of the body in the history of medicine. Journal ofMedicine and Philosophy, 15, 515-537. Light, K. C., Dolan, C. A., Davis, M. R., & Sherwood, A. (1992). Cardiovascular responses to an active coping challenge as predictors of blood pressure patterns 10 to 15 years later. Psychosomatic Medicine, 54(2), 217-30. Manuck, S., Cohen, S., Rabin, B., Muldoon, M., & Bachen, E. (1991). Individual differences in cellular immune response to stress. Psychological Science, 2(2), 111-115. Marusic, A., Gudjonsson, G. H., Eysenck, H. J., & Stare, R. (1999). Biological and psychosocial risk factors in ischaemic heart disease: Empirical findings and a biopsychosocial model. Personality and Individual Differences, 26(2), 285- 304. Maughan, B., & McCarthy, G. (1997). Childhood adversities and psychosocial disorders. British Medical Bulletin, 53(1), 156-169. McDaniel, J., Musselman, D., Porter, M., Reed, D., & Nemeroff, C. (1995). Depression in patients with cancer: Diagnosis, biology, and treatment. Archives o f General Psychiatry, 52, 89-99. McDaniel, S. (1995). Collaboration between psychologists and family physicians: Implementing the biopsychosocial model. Professional Psychology: Research and Practice, 26(2), 117-122. McDowell, I., & Newell, C. (1996). Measuring Health: A guide to rating scales and questionnaires (2nd ed.). New York: Oxford University Press. McEwen, B., & Stellar, E. (1993). Stress and the individual: Mechanisms leading to disease. Archives o f Internal Medicine, 153, 2093-2101. McHugh, S., & Vallis, M. (1986). Illness behaviour: Operationalization of the biopsychosocial model. In S. McHugh & T. Vallis (Eds ), Illness behavior: A multidisciplinary model (pp. 1-31). New York, NY: Plenum Press. McLaren, N. (1998). A critical review of the biopsychosocial model. Australian and New Zealand Journal o f Psychiatry, 32(1), 86-92. Miller, A.-H. (1998). Neuroendocrine and immune system interactions in stress and depression. Psychiatric Clinics o f North America, 21(2), 443-463. 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Mills, P. J., & Dimsdale, J. E. (1991). Cardiovascular reactivity to psychosocial stressors. A review of the effects of beta-blockade. Psychosomatics, 32(2), 209-20. Miser, W. F. (2000). Exercise as an effective treatment option for major depression in older adults. Journal o f Family Practice, 49(2), 109-10. Moerman, D. E., & Jonas, W. B. (2000). Toward a research agenda on placebo. Advances in Mind-Body Medicine, 16(1), 33-46. Moller, H. J., Blank, R., & Steinmeyer, E. M. (1989). Single-case evaluation of sleep-deprivation effects by means of nonparametric time-series analysis (according to the HTAKA model). European Archives o f Psychiatry & Neurological Sciences, 239(2), 133-9, Montgomery, S., Asberg, M., Traskman, L., & Montgomery, D. (1978). Cross cultural studies on the use of the CPRS in English and Swedish Depressed Patients. ActaPsychiat. Scand, Suppl. 271, 33-37. Moreno, F., Gelenberg, A., Heninger, G., Potter, R., McKnight, K., Allen, J., Phillips, A., & Delgado, P. (1999). Tryptophan depletion and depressive vulnerability. Biological Psychiatry, 46, 498-505. Mullen, P. E. (1998). "A Critical Review of the Biopsychosocial Model": Comment. Australian and New Zealand Journal o f Psychiatry, 32(1), 95-96. Murray, P., Rosenthal, K., Kobayashi, G., & Pfaller, M. (1998). Medical Microbiology (3rd ed.). St. Louis: Mosby. Musselman, D. L., Evans, D. L., & Nemeroff, C. B. (1998). The relationship of depression to cardiovascular disease: Epidemiology, biology, and treatment. Archives o f General Psychiatry, 55(1), 580-592. Mutrie, N. (2000). The relationship between physical activity and clinically defined depression. In S. Biddle, K. Fox, & S. Boutcher (Eds ), Physical Activity and Psychological Well-being (pp. 46-62). New York: Routledge. Naisberg, Y. (1996). Homeostatic disruption and depression. Medical Hypotheses, 47(6), 415-22. Naliboflf, B., Benton, D., Solomon, G., Morley, J., Fahey, J., Bloom, E., Makinodan, T., & Gilmore, S. (1991), Immunological changes in young and old adults during brief laboratory stress. Psychosomatic Medicine, 53, 121-132. 159 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Naliboff, B., Solomon, G., Gilmore, S., Fahey, J., Benton, D., & Pine, J. (1995). Rapid changes in cellular immunity following a confrontational role-play stressor. Brain, Behavior, and Immunity, 9, 207-219. Nelson, E., & Cloninger, C. R. (1997). Exploring the TPQ as a possible predictor of antidepressant response to nefazodone in a large multi-site study. Journal o f Affective Disorders, 44(2-3), 197-200. Nemeroff, C., Heim, C., Owens, M., Newport, D., Miller, A., & Plotsky, P. (1999, ). Neurochemical mechanisms underlying depression and anxiety disorders: The influence of early trauma New Developments in Understanding Depression and Its Treatment, pp. 3-6. Newman, J. D., McGarvey, S. T., & Steele, M. S. (1999). Longitudinal association of cardiovascular reactivity and blood pressure in Samoan adolescents. Psychosomatic Medicine, 67(2), 243-9. Nierenberg, A. A., & Wright, E. C. (1999). Evolution of remission as the new standard in the treatment of depression. Journal o f Clinical Psychiatry, 60(Suppl 22), 7-11. Organization, W. H. (1992). The ICD classification o f mental and behavioural disorders: Clinical descriptions and diagnostic guidelines. Geneva, Switzerland: Author. Page-Shafer, K., Delorenze, G., Satariano, W., & Winkelstein, W. J. (1996). Comorbidity and survivial in HIV-infected men in the San Francisco Men's Health Survey. Ann Epidemiol, 6, 420-430. Paris, J. (1993). Personality disorders: A biopsychosocial model. Journal o f Personality Disorders, 7(3), 255-264. Patton, G. (1993). Eating problems. In C. Costello (Ed.), Symptoms o f Depression (pp. 227-242). New York: John Wiley & Sons, Inc. Paykel, E. (1983). Methodological aspects of life events research. Journal o f Psychosomatic Research, 27(5), 341-352. Paykel, E., Hollyman, J., Freeling, P., & Sedgwick, P. (1988). Predictors o f therapeutic benefit from amitriptyline in mild depression: a general practice placebo-controlled trial. Journal o f Affective Disorders, 14, 83-95. Paykel, E. S. (1994). Life events, social support and depression. Acta Psychiatrica Scandinavica, Supplementum, 377, 50-8. 160 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Pearlstein, T. (1995). Hormones and depression: What are the facts about premenstrual syndrome, menopause, and hormone replacement therapy? Am J Obstet Gynecol, 173, 646-653. Penninx, B. W., Geerlings, S. W., Deeg, D. J., van Eijk, J. T., van Tilburg, W., & Beekman, A. T. (1999). Minor and major depression and the risk of death in older persons. Archives o f General Psychiatry, 56(10), 889-95. Pike, J., Smith, T., Hauger, R., Nicassio, P., Patterson, T., McClintick, J., Costlow, C., & Irwin, M. (1997). Chronic life stress alters sympathetic, neuroendocrine, and immune responsivity to an acute psychological stressor in humans. Psychosomatic Medicine, 59, 447-457. Quitkin, F., McGraph, P., Rabkin, J., Stewart, J., Harrison, W., Ross, D., Tricamo, E., Gleiss, J., Markowitz, J., & Klein, D. (1991). Different types of placebo response in patients receiving antidepressants. American Journal o f Psychiatry, 148, 197-203. Quitkin, F., McGraph, P., Stewart, J., Ocepek-Welikson, K., Taylor, B., Nunes, E., Delivannides, D., Agosti, V., Donovan, S., Ross, D., Petkova, E., & Klein, D. (1998). Placebo run-in period in studies of depressive disorders. Clinical, heuristic and research implications. British Journal o f Psychiatry, 173(9), 242-248. Quitkin, F. M. (1999). Placebos, drug effects, and study design: a clinician's guide. American Journal o f Psychiatry, 156(6), 829-36. Ravindran, A., Griffeths, J., Merali, Z., & Anisman, H. (1996). Variations in lymphocyte subsets associated with stress in depressive populations. Psychoneuroendocrinology, 21, 659-671. Ravindran, A.-V., Griffiths, J., Merali, Z., & Anisman, H. (1998). Circulating lymphocyte subsets in major depression and dysthymia with typical or atypical features. Psychosomatic Medicine, 60(3), 283-289. Robinson, L., Berman, J., & Neimeyer, R. (1990). Psychotherapy for the treatment of depression: A comprehensive review of controlled outcome research. Psychological Bulletin, 108, 30-49. Roy, M.-P., Steptoe, A., & Kirschbaum, C. (1998). Life events and social support as moderators of individual differences in cardiovascular and cortisol reactivity. Journal o f Personality and Social Psychology, 75(5), 1273-1281. 161 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Rozanski, A., Blumenthal, J. A., & Kaplan, J. (1999). Impact of psychological factors on the pathogenesis of cardiovascular disease and implications for therapy. Circulation, 99(16), 2192-217. Sadler, J. Z., & Hulgus, Y. F. (1990). Knowing, valuing, acting: Clues to revising the biopsychosocial model. Comprehensive Psychiatry, 37(3), 185-195. Savard, J., Miller, S., Mills, M., O'Leary, A., Harding, H., Douglas, S., Mangan, C., Belch, R., & Winokur, A. (1999). Association between subjective sleep quality and depression on immunocompetence in low-income women at risk for cervical cancer. Psychosomatic Medicine, 61(4), 496-507. Schedlowski, M., Jacobs, R., Stratmann, G., Richter, S., Hadicke, A., Tewes, U., Wagner, T., & Schmidt, R. (1993). Changes of natural killer cells during acute psychological stress. Journal o f Clinical Immunology, 13(2), 119-126. Schoenfeld Smith, K., Petroski, G. F., Hewett, J. E., Johnson, J. C., Wright, G. E., Smarr, K. L., Walker, S. E., & Parker, J. C. (1996). A biopsychosocial model of disability in rheumatoid arthritis. Arthritis Care and Research, 9(5), 368- 375. Schulberg, H. C., Katon, W., Simon, G. E., & Rush, A. J. (1998). Treating major depression in primary care practice: An update of the Agency for Health Care Policy and Research practice guidelines. Archives o f General Psychiatry, 55(12), 1121-1127. Schwartz, G. (1982). Testing the biopsychosocial model: The ultimate challenge facing behavioral medicine? Journal o f Consulting and Clinical Psychology, 50, 1040-1053. Schwartzman, J., & Glaus, K. (2000). Depression and coronary heart disease in women: Implications for clinical practice and research. Professional Psychology: Research and Practice, 37(1), 48-57. Segerstrom, S.-C. (1997). Optimism and coping with enduring stress: Psychological and immune effects. Dissertation Abstracts International: Section B: The Sciences cmdEngineering, 5#(6-B), 3355. Seligman, M. (1975). Helplessness. San Francisco: Freeman. Shapiro, A., & Shapiro, E. (1997). The placebo: is it much ado about nothing? In E. Harrington (Ed.), The Placebo Effect: An Interdisciplinary Exploration (pp. 12-36). Cambridge: Harvard University Press. 162 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Shapiro, A., Struening, E., Barten, H., & Shapiro, E. (1975). Correlates of placebo reaction in an outpatient population. Psychological Medicine, 5, 389-396. Shapiro, D., Jamner, L., Lane, J., Light, K., Myrtek, M., Sawada, Y., & Steptoe, A. (1996). Blood pressure publication guidelines. Psychophysiology, 33, 1-12. Sheffield, D., Krittayaphong, R., Cascio, W., Light, K., Golden, R., Finkel, J., Glekas, G., Koch, G., & Sheps, D. (1998). Heart rate variability at rest and during mental stress in patients with coronary artery disease: Differences in patients with high and low depression scores. International Journal o f Behavioral Medicine, 5(1), 31-47. Sherwood, A., & Turner, J. (1992). A conceptual and methodological overview of cardiovascular reactivity research. In J. Turner, A. Sherwood, & K. Light (Eds.), Individual differences in cardiovascular responses to stress (pp. 3- 32). New York: Plenum Press. Simon, G., Vonkorff, M., Ustun, T., Gater, R., Gureje, O., & Sartorius, N. (1995). Is the lifetime risk of depression actually increasing? Journal o f Clinical Epidemiology, 48(9), 1109-1118. Solomon, G. (1997). Stress, hormones, immunity, the complexity of intertwined systems, and the simplicity of humanism. Journal o f Intensive Care Medicine, 12, 219-222. Spiegel, D. (1996). Cancer and depression. British Journal ofPsychiatiy - Supplement(30), 109-16. Spring, B., Chiodo, J., & Bowden, D. (1987). Carbohydrates, tryptophan, and behavior: a methodological review. Psychol Bull, 102, 234-56. Spruijt-Metz, D. (1999). Adolescence, affect and health. London: Psychology Press. Stangl, D. K., & Greenhouse, J. B. (1998). Assessing placebo response using Bayesian hierarchical survival models. Lifetime Data Analysis, 4(1), 5-28. Steinbrueck, S., Mawell, S., & Howard, G. (1983). A meta-analysis of psychotherapy and drug therapy in the treatment of unipolar depression with adults. Journal o f Consulting and Clinical Psychology, 51, 856-863. Stoudemire, A. (Ed.). (1995). Psychological Factors Affecting Medical Conditions. Washington, DC: American Psychiatric Press, Inc. 163 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Swartzman, L. C., & Burkell, J. (1998). Expectations and the placebo effect in clinical drug trials: why we should not turn a blind eye to unblinding, and other cautionary notes. Clinical Pharmacology & Therapeutics, 64( 1), 1-7. Taylor, S. (1999). Health Psychology (4th ed ). Boston. McGraw Hill. Temoshok, L. (1990). On attempting to articulate the biopsychosocial model: Psychological-psychophysiological homeostasis. In H. S. Friedman (Ed.), Personality and disease, (pp. 203-225). New York: John Wiley & Sons. Tewes, U. (1999). Concepts in Psychology. In M. Schedlowski & U. Tewes (Eds.), Psychoneuroimmunology: An interdisciplinary introduction (pp. 93-111). New York: Kluwer Academic/Plenum Publishers. Thase, M., Entsuah, A., & Rudolph, R. (2001). Remission rates during treatment with venlafaxine or selective serotonin reuptake inhibitors. British Journal o f Psychiatry, 178, 234-241. Thase, M. E. (1999). When are psychotherapy and pharmacotherapy combinations the treatment of choice for major depressive disorder? Psychiatric Quarterly, 70(4), 333-46. Tilders, F. J., & Schmidt, E. D. (1999). Cross-sensitization between immune and non-immune stressors. A role in the etiology of depression? Advances in Experimental Medicine & Biology, 461, 179-97. Tueting, P. (1991). Psychophysiological predictors of drug treatment response. Psychiatric Medicine, 9(1), 145-161. Uchino, B., Cacioppo, J., Malarkey, W., & Glaser, R. (1995). Individual differences in cardiac sympathetic control predict endocrine and immune responses to acute psychological stress. Journal o f Personality and Social Psychology, 69, 763-743. Uchino, B.-N., Cacioppo, J.-T., & Keicolt-Glaser, J.-K. (1996). The relationship between social support and physiological processes: A review with emphasis on underlying mechanisms and implications for health. Psychological Bulletin, 119(3), 488-531. Vilhjalmsson, R. (1993). Life stress, social support and clinical depression: a reanalysis of the literature. Social Science &Medicine, 37(3), 331-42. 164 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wade, T. D., & Kendler, K. S. (2000). The relationship between social support and major depression: cross-sectional, longitudinal, and genetic perspectives. Journal o f Nervous & Mental Disease, 188(5), 251-8. Waked, E. G., & Jutai, J. W. (1990). Baseline and reactivity measures of blood pressure and negative affect in borderline hypertension. Physiology & Behavior, 47(2), 265-71. Wakefield, J. (1998). Meaning and melancholia: Why the DSM-IV cannot (entirely) ignore the patients intentional system. In J. Barron (Ed.), Making diagnosis meaningful: Enhancing evaluation and treatment o f psychological disorders (pp. 29-72). Washington, DC: American Psychological Association. Waldstein, S., Bachen, E., & Manuck, S. (1997). Active coping and cardiovascular reactivity: A multiplicity of influences. Psychosomatic Medicine, 59(6). Wallace, E. (1990). Mind-body and the future of psychiatry. Journal o f Medicine and Philosophy, 75(1), 41-74. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: the PANAS scales. Journal o f Personality & Social Psychology, 54(6), 1063-70. Watson, M., Haviland, J., Greer, S., Davidson, J., & Bliss, J. (1999). Influence of psychological response on survival in breast cancer: a population-based cohort study. Lancet, 554(9187), 1331-1336. Weissman, S. (1999). Psychiatric Diagnosis. In S. Weissman, M. Sabshin, & H. Eist (Eds.), Psychiatry in the New Millenium (pp. 57-66). Washington, DC: American Psychiatric Press, Inc. Wexler, B ., & Cicchetti, D. (1992). The outpatient treatment of depression: Implications of outcome research for clinical practice. Journal o f Nervous and Mental Disease, 180, 277-286. Whybrow, P. (1997). A Mood Apart: A thinker's guide to emotion and its disorders. New York: HarperPerennial. Whybrow, P., Akiskal, H., & McKinney, W. J. (1984). Toward a psychobiological integration: affective illness as a final common path to adaptive failure, Mood Disorders: Toward a New Psychobiology (pp. 173-203). New York: Plenum. Wickramaratne, P., & Weissman, M. (1996). Time Trends in Risk of Depression. Journal o f Clinical Epidemiology, 49(9), 1077-1078. 165 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Widiger, T., Frances, A., Pincus, H., Ross, R., First, M., & Davis, W. (Eds.). (1996). DSM-IVSourcebook (Vol. 2). Washington, DC: American Psychiatric Association. Wilhelm, K., Parker, G., & Hadzi-Pavlovic, D. (1997). Fifteen years on: evolving ideas in researching sex differences in depression. Psychological Medicine, 27(4), 875-83. Williams, J., Gibbon, M., First, M., Spitzer, R., Davies, M., Borus, J., Howes, M., Kane, J., Pope, H., Rounsaville, B., & Wittchen, H. (1992). The Structured clinical Interview for DSM-III-R (SCID). II. Multisite test-retest reliability. Archives o f General Psychiatry, 49, 630-636. Willner, P. (1985). Depression: A psychobiological synthesis. New York: John Wiley and Sons. Wilson, D. K., Kliewer, W., Bayer, L., Jones, D., Welleford, A., Heiney, M., & Sica, D. A. (1999). The influence of gender and emotional versus instrumental support on cardiovascular reactivity in African-American adolescents. Annals o f Behavioral Medicine, 21(3), 235-43. Wolkenstein, A., & Butler, D. (1998). The pre-encounter psychosocial review: The anticipatory integration of the biopsychosocial model in primary care. Families, Systems and Health, 16(4), 337-346. Wulsin, L. R., Vaillant, G. E., & Wells, V. E. (1999). A systematic review of the mortality of depression. Psychosomatic Medicine, 67(1), 6-17. Young, S .-N. (1993). The use of diet and dietary components in the study of factors controlling affect in humans: A review. Journal o f Psychiatry and Neuroscience, 18(5), 235-244. Zarate, C., & Tohen, M. (1996). Epidemiology of mood disorders throughout the life cycle. In K. Shulman, M. Tohen, & S. Kutcher (Eds ), Mood Disorders Across the Life Span (pp. 16-33). New York: Wiley-Liss, Inc. Zimmermann, C., & Tansella, M. (1996). Psychosocial factors and physical illness in primary care: Promoting the biopsychosocial model in medical practice. Journal o f Psychosomatic Research, 40(4), 351-358. 166 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appendix A. Individual Predictors of HDRS Change for All Subjects HDRS Change Over Time [ 3 ____________ SE ____________P Gender (Female) 0.13 0.07 .06 Ethnicity (Non-White) 0.16 0.08 .04 Number of episodes -0.03 0.13 .02 HDRS Item 1: Depressed mood 0.12 0.06 .05 HDRS Item 4: Insomnia early 0.10 0.04 .02 HDRS Item 5: Insomnia middle 0.11 0.04 .01 HDRS Item 6: Insomnia late* 0.09 0.04 .02 HDRS Item 7: Work and activities 0.11 0.06 .08 HDRS Item 11: Anxiety somatic 0.06 0.03 .06 HDRS Item 12: Somatic symptoms- GI 0.08 0.05 .09 HDRS Item 15: Hypochondriasis* 0.06 0.03 .06 SCID total 0.04 0.02 .05 SCID item: Fatigue 0.14 0.07 .06 BDI baseline 0.01 0.004 .003 BDI wash-in difference -0.01 0.006 .09 MADRS baseline 0.02 0.007 .02 MADRS wash-in difference* -0.02 0.006 .003 HAM-A baseline 0.01 0.005 .01 SCL-90-R Somatization 0.11 0.05 .03 SCL-90-R Depression 0.18 0.06 .002 SCL-90-R Anxiety 0.13 0.05 .01 SCL-90-R Hostility 0.15 0.04 .002 SCL-90-R Psychoticism 0.21 0.06 .001 SCL-90-R General Severity Index 0.25 0.07 .0004 SCL-90-R Positive Symptom Distress 0.15 0.08 .06 SCL-90-R Positive Symptom Total 0.01 0.002 .005 Growth curve models with random intercept were used to determine which one-variable models were the best predictors of change in HDRS after adjusting for treatment. Results are shown for all variables with p< 10. 51 (25 drug, 26 placebo) subjects were used in the analysis. Due to missing data, only 43 (21 drug, 22 placebo) subjects were used in the analysis for SCL-90-R items. * A borderline or significant interaction with treatment Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Comparisons of metabolic factors among gestational diabetes mellitus probands, siblings and cousins
PDF
A pilot survey of medical abortion knowledge and practices among obstetrician/gynecologists and family practitioners in Los Angeles County
PDF
An intervention and program evaluation to determine the effectiveness of public health reforms on primary prevention practices by chiropractic interns
PDF
Extent, prevalence and progression of coronary calcium in four ethnic groups
PDF
Association between latchkey status and smoking behavior in middle school children
PDF
Development and evaluation of standardized stroke outcome measures in a population of stroke patients in rural China
PDF
Cumulative pregnancy rates after oocyte donation
PDF
A case/parental/sibling control study of Ewing's sarcoma/peripheral primitive neuroectodermal tumor (pPNET)
PDF
A descriptive analysis of medication use by asthmatics in the Children's Health Study, 1993
PDF
Determinants of mammographic density in African-American, non-Hispanic white and Hispanic white women before and after the diagnosis with breast cancer
PDF
A linear model for measurement errors in oligonucleotide microarray experiment
PDF
Breast cancer in the multiethnic cohort study: Genetic (prolactin pathway genes) and environmental (hormone therapy) factors
PDF
Dietary fiber intake and atherosclerosis progression: The Los Angeles Atherosclerosis Study
PDF
Evaluation of the accuracy and reliability of self-reported breast, cervical, and ovarian cancer incidence in a large population-based cohort of native California twins
PDF
Descriptive epidemiology of thyroid cancer in Los Angeles County, 1972-1995
PDF
BRCA1 mutations and polymorphisms in African American women with a family history of breast cancer identified through high throughput sequencing
PDF
A large-scale genetic association study of prostate cancer in a multi-ethnic population
PDF
Emotional intelligence and smoking risk factors in early adolescents
PDF
Family history, hormone replacement therapy and breast cancer risk on Hispanic and non-Hispanic women, The New Mexico Women's Health Study
PDF
Interaction of dietary fiber and serum cholesterol on early atherosclerosis
Asset Metadata
Creator
Marie, Ariane
(author)
Core Title
Biopsychosocial factors in major depressive disorder
School
Graduate School
Degree
Doctor of Philosophy
Degree Program
Epidemiology
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
biology, biostatistics,health sciences, mental health,OAI-PMH Harvest,psychology, psychobiology
Language
English
Contributor
Digitized by ProQuest
(provenance)
Advisor
Azen, Stanley (
committee chair
), Leuchter, Andrew (
committee chair
), Chou, Chih-Ping (
committee member
), Gatz, Margaret (
committee member
), Spruitj-Metz, Donna (
committee member
)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c16-240305
Unique identifier
UC11338975
Identifier
3093419.pdf (filename),usctheses-c16-240305 (legacy record id)
Legacy Identifier
3093419.pdf
Dmrecord
240305
Document Type
Dissertation
Rights
Marie, Ariane
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
biology, biostatistics
health sciences, mental health
psychology, psychobiology