The Theory and Treatment of Depression

Towards a Dynamic Interactionism Model

Jozef Corveleyn, Patrick Luyten, Sidney J. Blatt, Hilde Lens-Gielis, Jozef Corveleyn, Patrick Luyten, Sidney J. Blatt, Hilde Lens-Gielis

• Publisher: Routledge
• Year: 2013
• ISBN:
  9781134999323
• Edition: 1
• Topic: Medicine
• Subtopic: Psychiatry & Mental Health

Recent research indicates that depression, once believed to be relatively benign, is highly recurrent and does not respond well to treatment. The goal of this book is to facilitate the development of more encompassing theories and more effective treatments for this disabling disorder by fostering dialogue and enhancing the integration of work across the boundaries of separate fields.

Chapter 1

The Lifecycle of Depression

Koen Demyttenaere, Lukas Van Oudenhove, & Jürgen De Fruyt

Mood disorders in general and major depressive disorder (MDD) in particular are common, disabling psychiatric disorders with an estimated lifetime prevalence for MDD of around 17% in the community (Angst, 1997). The Belgian ESEMeD (European Study of Epidemiology of Mental Disorders) results confirm this estimate (lifetime prevalence of MDD 10.4% and 18.9% in males and females, respectively) (Alonso, Angermeyer, Bernert, Bruffaerts, et al., 2004).

Furthermore, the prevalence rates of MDD seem to have increased considerably over the past decades. A substantial increase in the rates in cohorts born after World War II and a decrease in the age of onset have been found in several large epidemiologic and family studies (Klerman & Weissman, 1989). There is some evidence that these changes are at least partly artifactual, though. Identification phenomena, memory effects and differential mortality, among others, can explain why these findings may be to some extent due to artifacts (Klerman & Weissman, 1989).

Over the past fifteen years, there has been increasing interest in the longitudinal course of MDD in the scientific literature. The discovery of psychotropic drugs in general and antidepressants in particular certainly made it possible to treat the acute phase of MDD more appropriately, but did this influence the long-term naturalistic outcome of the disorder? Nowadays, it is becoming more and more clear that MDD should be seen as a disorder that is chronic, progressive and recurrent in nature (Angst, 1997; Mueller et al., 1999; Solomon et al., 2000; Kessing, Andersen, Mortensen, & Bolwig, 1998). In this review, we'll try to clarify some important aspects concerning the longitudinal course of depressive illness, in particular the recent evidence on the differences between first and recurrent episodes.

Finally, there has been a growing amount of research on the validity of the concept of MDD as defined by DSM-IV diagnostic criteria. Until a few years ago, little was known about the significance of subthreshold depressive symptoms or “subsyndromal depression”. Some recent research provides support for the hypothesis that the “cut-off” of five symptoms or two weeks duration as required by DSM-IV is rather arbitrary: conventional limits put on a continuum of depressive symptoms of varying severity and duration. In other words, the key question is: should the diagnosis of MDD be seen in a categorical or a dimensional perspective (Judd et al., 1998a, 1998b; Kendler & Gardner, 1998)?

Increasing Rates of Depression: Fact or Fiction?

In 1989, Klerman and Weissman (1989) commented on the results of several large epidemiologic and family studies, suggesting important temporal changes in the rate of major depression. In the cohorts born after World War II, an increasing rate is found, as well as a decrease in the age of onset (with an increase in the late teenage and early adult years), an increase between 1960 and 1975 in the rates of depression for all ages, a persistent gender effect (with the risk of depression consistently two to three times higher among women than men across all adult ages) and finally a persistent family effect, with the risk about two to three times higher in first-degree relatives as compared with controls.

Undoubtedly, the use of standardized diagnostical instruments in large community-based studies and the application of modern statistical techniques to the large data sets from these studies have made more reliable research on the temporal changes in rates of mental disorders possible. The results from this kind of studies, especially those with a longitudinal design, can provide some important information regarding syndromal validity, natural course and, more indirectly, etiology, pathogenesis and finally treatment of mental disorders. For instance, the combination of familial aggregation and temporal changes suggests that gene-environment interaction plays an important role in the pathogenesis of MDD (Klerman & Weissman, 1989). It is important, though, to mention that the reported studies have a cross-sectional instead of a longitudinal design, which plays a considerable role in some of the issues discussed below.

Some of the findings mentioned above, like the higher rate among women and first-degree relatives of depressed patients, are consistently reported in the literature (Klerman & Weissman. 1989). Others, especially the increasing rate in recent birth cohorts are more controversial and merit some further clarification. First, the cohort effect in MDD is not universally reported. For example, community-based epidemiological studies in Puerto Rican, Mexican-American and Korean populations do not report these temporal changes, whereas comparable American, German and Canadian studies consistently find a cohort effect (Klerman & Weissman, 1989). Furthermore, there are some possible explanations why these findings of temporal changes may be artifactual rather than reflecting real changes in the rate of MDD. We will discuss three of the most interesting ones in detail.

Identification Phenomenon

An identification phenomenon can be an important confounding factor, as the studies mentioned above relied mainly on subjects' retrospective reports. Hasin and Link (1988) set up a study to test the hypothesis that there is an interaction between age and recognition. In other words, they expected that older individuals are less likely to recognize depression as a mental disorder, compared to younger adults, who might have a more “psychological” view to their experiences. The subjects in the study, 152 randomly selected citizens living in a suburban area of New York, were asked to imagine themselves in the situation described by the following vignette, describing a DSM-III major depressive episode with a duration of a month. A statement reflecting impairment in functioning was inserted after the vignette. Furthermore, subjects were requested to fill in a questionnaire. The questions asked whether the subjects regarded the situation described by the vignette as emotional/psychological in nature, and whether they considered the situation to be a problem or not:

“During the last month, even though nothing has gone wrong that you can think of, you have been feeling depressed. You have not been able to get enough sleep at night, you haven't had much appetite, and you have lost over five pounds. You have felt tired much of the time, and haven't been able to concentrate as well as usual. You have also noticed that you aren't enjoying things the way you would normally… “(Hasin & Link, 1988).

The results after analysis of the answers were quite clear. Age had a highly significant effect that did not change after controlling for vignette types, education and gender. The chance of recognizing the vignette as major depression was 3.78 times higher for a 25-year old than for a 60-year old. Thus, older people are less likely to remember depressive episodes or report them when asked for in a questionnaire regarding mental health. The authors conclude that there is a cohort effect in the perception rather than in the real rate of depression (Hasin & Link, 1988).

Memory Effect: Diminished Recall

Memory effects can provide a second possible explanation for a cohort effect in rates of major depression as found in cross-sectional studies: older adults are more likely to forget previous depressive episodes compared to younger ones. Giuffra and Risch (1994) set up a simulation study to examine the effect of forgetting on differences between cohorts. The results suggest that small, constant annual rates of forgetting applied to successive birth cohorts generate cohort-like effects. These findings confirm the hypothesis that diminished recall in older cohorts is one of the possible confounding factors involved in the reported cohort effect in major depression rates. More generally, we have to take into account the somewhat limited value of cross-sectional studies when doing research on temporal changes in prevalence.

Moreover, work by Rice et al. indicates that a lifetime diagnosis of major depression is not stable over time, although the error decreases as severity increases. They assessed a population of 2,226 first-degree relatives of 612 probands who participated in another study at two time points, with an interval of 6 years. A substantial proportion of the assessed subjects were considered as having a lifetime diagnosis of major depression at one of the two time points, but not at the other. This could be due to suboptimal repeatability of the diagnostic instrument (reliability), diminished recall or limited validity of the underlying diagnostic construct, however (Rice, Rochberg, Endicott, Lavori, & Miller, 1992). Anyway, these findings provide some further evidence for the role of memory effects as a confounding factor in cohort effects found in cross-sectional studies.

Differential Mortality

Finally, the findings of temporal changes could be at least partly explained by a difference in mortality between healthy subjects and depressive subjects. If depression is associated with a higher mortality, the older birth cohorts would have lost more depressed subjects by death compared to younger birth cohorts. This could decrease the rate of depression in interviewed older populations (Klerman & Weissman, 1989).

The relationship between MDD and mortality is complex and there have been some conflicting studies concerning this issue. In a recent systematic review regarding this topic, Wulsin, Vaillant and Wells (1999) found 57 studies. Twenty-nine of these (51%) found a positive association, 13 (23%) were negative and the results of the remaining 15 studies (26%) were mixed. In general, most of the studies on the link between depression and early death varied widely in methodology (sample selection, comparison groups, control for confounding variables, etc.), making comparisons between the different studies rather difficult. Furthermore, publication bias certainly is a problem. Some of the mentioned studies were well designed however. For these reasons, “the evidence that depression increases mortality may not be strong enough to answer the question definitively” (Wulsin et al., 1999).

However, some conclusions can be drawn from this review. 16% to 19% of the mortality in psychiatric samples of depression is due to suicide, which is consistent with the often reported suicide rate of 15% in MDD. By contrast, in most studies based on depressive samples recruited from the community or medical settings, no more than 1 % of deaths are caused by suicide. Apart from suicide and other non-natural causes of death, a quite strong link between depression and cardiovascular mortality was found (Wulsin et al., 1999).

Although the important and interesting discussion on possible mechanisms linking depression and mortality lies beyond the scope of this review, we will mention two hypotheses briefly. First, depression may cause early death in an indirect way, for example due to poor self-care (including unhealthy food habits but also compliance with medical treatment). Second, there may be a more direct, pathophysiological link between depression and mortality. For example, there has been considerable evidence that depression is associated with important changes in endocrinological and immune systems. These changes make depressive individuals more vulnerable to cardiovascular and other diseases (Whooley & Browner, 1998). There is some evidence from well-designed studies that even subsyndromal depressive symptoms are associated with both higher mortality and immunological activation, which suggests a possible link (Glaser, Robles, Sheridan, Malarkey, & Kiecolt-Glaser, 2003; Whooley & Browner, 1998). We will discuss the importance of subsyndromal depressive symptoms more thoroughly further on. Dysfunction of the autonomous nervous system or abnormal platelet aggregation could also cause higher cardiovascular mortality among depressed individuals (Whooley & Browner, 1998; Ösby, Brandt, Correia, Ekbom, & Sparen, 2001).

As already mentioned above, there have been some studies on the association between depression and mortality in all kinds of populations and samples and it is important to take this into account when looking at the result of those studies.

Bruce, Leaf, Rozal, Florio and Hoff (1994) examined the relationship between psychiatric illness and mortality over 9 years in the community sample from the New Haven Epidemiologic Catchment Area Study, consisting of 3,560 subjects. Subjects who reported a recent episode of MDD at the first interview had a 9-year relative mortality risk of 2.01. There was a significant interaction with gender: recently depressed men were 4.22 times more likely to die during the follow-up period compared to men without recent depressive episode; for women, the relative mortality risk was 1.65 (Bruce et al., 1994).

In a recent study by Ösby et al. (2001), the cause of death of all Swedish inpatients with a diagnosis of bipolar (N=15,386) or unipolar mood disorder (N=39,182) between 1973 and 1995 was determined. All patients were followed from their first hospital admission, using the Swedish psychiatric inpatient register. The date and the cause of death were found in the national cause-of-death register. The standardized mortality ratios (SMRs) for suicide in unipolar disorder were 20.9 and 27.0 for males and females, respectively. These rates are particularly high, which can be explained by the population studied, consisting of severely ill inpatients. The risk of suicide was particularly high for younger patients during the first years after the first diagnosis, findings that are important when looking for prevention strategies. The SMRs for all natural causes of death in unipolar disorder were 1.5 and 1.6 for males and females, respectively (Ösby et al., 2001). In a longitudinal study of psychiatric inpatients, Angst et al. found higher suicide rates compared to a control population, as well as a higher cardiovascular mortality rate. Moreover, pharmacological treatment lowered the suicide rate (Angst, Stassen, Clayton, & Angst, 2002).

Bingefors, Isacson, Knorring, Smedby and Wicknertz (1996) conducted a community-based study in a primary and psychiatric ambulatory care setting. They identified all first-incidence antidepressant users in ambulatory care and analyzed their mortality during a nine year follow-up period. They found that in patients aged 65 and older, antidepressant treatment at index was a significant predictor of higher 9-year mortality. A hazard ratio of 1.52 was found,...

Table of contents

• Cover
• Half title
• FIGURES OF THE UNCONSCIOUS 5
• Full title
• Copyright
• About the editors
• List of contributors
• Contents
• Foreword
• Introduction
• Chapter 1 The life cycle of depression
• Chapter 2 Mood and memory: A cognitive psychology perspective on maintenance of depressed mood and vulnerability for relapse
• Chapter 3 The convergence among psychodynamic and cognitive-behavioral theories of depression: Theoretical overview
• Chapter 4 The convergence among psychodynamic and cognitive-behavioral theories of depression: A critical review of empirical research
• Chapter 5 A dialectic model of personality development and psychopathology: Recent contributions to understanding and treating depression
• Chapter 6 ‘Closed doors and landscapes in the mist’ 1. Childhood and adolescent depression in developmental psychopathology
• Chapter 7 ‘Closed doors and landscapes in the mist’ 2. Depression in psychoanalytic developmental psychopathology: From single track models to complex developmental pathways
• Chapter 8 Corticotropin releasing factor (CRF) and major depression: Towards an integration of psychology and neurobiology in depression research
• Epilogue Towards integration in the theory and treatment of depression? The time is now
• Author / Subject index