Depression, Bereavement, and “Understandable” Intense Sadness: Should the DSM-IV Approach Be Revised?

The issue of the differentiation between depression and “understandable” intense sadness (representing a “normal” response to an adverse life event) has significant clinical, scientific, political, and ethical implications, which have become particularly visible in the past few decades, in parallel with the escalation of the prevalence rates of depression in the community, of the estimated social costs of depression, of the number of people on treatment for depression, and of the prescriptions of antidepressant medications (1) . Psychiatry has been accused of inappropriately medicalizing the ordinary experience of sadness in order to expand the range of its jurisdiction (2), and the high prevalence rates of depression reported by community studies have been regarded as unbelievable even by some prominent psychiatrists, who have emphasized the risk to misdiagnose “normal reactions to a difficult environment” as a mental disorder (3) .

According to DSM-IV, periods of sadness are inherent aspects of the human experience, which should not be diagnosed as a major depressive episode unless criteria are met for severity, duration, and clinically significant distress or impairment. The implication of this statement is that “understandable” intense sadness following an adverse life event does qualify for the diagnosis of major depression if the severity, duration, and impairment criteria are fulfilled. In other terms, the context in which the depressive symptoms occur is not relevant to the diagnostic decision; what counts is the clinical picture.

The only exception to this rule is represented by bereavement. If the symptoms begin within 2 months of the death of a loved one and do not persist beyond these 2 months, the diagnosis of major depression should not be made, unless the symptoms are associated with marked functional impairment or include morbid preoccupation with worthlessness, suicidal ideation, psychotic symptoms, or psychomotor retardation. The rationale for this exclusion criterion can be found in the Introduction to DSM-IV: in order to represent a mental disorder, a condition “must not be merely an expectable and culturally sanctioned response to a particular event, for example, the death of a loved one” (p. xxi). A depressive state is an expectable and culturally sanctioned response to the death of a loved one and, as such, does not represent a mental disorder. Some empirical support to this position is provided by the high prevalence of a major depressive syndrome, ranging from 35% to 58%, 1 month after the death of a loved one (4, 5) and by the fact that most bereaved people describe themselves as being what they would have expected to be given the circumstances, whereas depressed patients experience their condition as a “change,” “not usual self” (4) .

The asymmetry introduced by DSM-III/IV between the death of a loved one and other major adverse life events has, not surprisingly, attracted the attention of researchers. In this issue, Kendler et al. (6) report that the similarities between bereavement-related and other life stressor-related depression far outweigh their differences, arguing against the continued use of the bereavement exclusion criterion in DSM-V. In another recent paper, Wakefield et al. (7), on the basis of similar findings, proposed to exclude both bereavement-triggered and other loss-triggered uncomplicated intense sadness from the DSM-V diagnosis of major depression (i.e., to introduce a “contextual” criterion excluding intense sadness that appears “proportionate” to a loss).

These opposite proposals based on the same research evidence are likely to divide our field and the public opinion for several years. Of course, both proposals have significant treatment implications. On the one hand, the risk is to medicalize an adaptive response, thus disrupting the individual’s coping processes. On the other, the risk is to deprive a person with a full depressive syndrome of a treatment that may be needed. Furthermore, the impact of the implementation of either proposal on the prevalence rates of major depression is likely to be substantial: for instance, in the community sample studied by Wakefield et al. (7), uncomplicated other loss-triggered cases accounted for 20.1% of all cases of major depression, and uncomplicated bereavement-triggered cases for 6.5%.

The issue is even more complex than it may appear to an unsophisticated thinker. The death of a loved one is a clear-cut, easily ascertainable life event that is usually out of the person’s control, but other adverse life events may be quite different in these respects. The presence itself of a depressive state may impair the accuracy of a person’s report of recent life events or may expose a person to adverse events. The experimental induction of depressed mood has been found to produce an increase in reports of past stressful events (8), and the relationship between depression and “dependent” events (i.e., events that can be explained by the depressive state, such as being fired from a job) has been found to be stronger than the relationship between depression and “independent” events (9) .

Furthermore, whether an adverse life event has been really decisive in triggering a depressive state may be difficult to establish in several cases. This is well known since the 1960s, when Aubrey Lewis, testing a set of criteria aimed to distinguish between “contextual” and “endogenous” depression, concluded that most depressive cases were “examples of the interaction of organism and environment,” so that “it was impossible to say which of the factors was decidedly preponderant” (10) .

Further research is clearly needed to explore the applicability and reliability of a “contextual” criterion in the diagnosis of major depression and the clinical utility of such a criterion for the prediction of treatment response and clinical outcome. The limited available research evidence suggests that definite “situational” major depression does not differ from definite “nonsituational” major depression on many clinical and psychosocial variables (11) and that response to antidepressant medications is unrelated to whether or not major depression is preceded by a life event (12) . At the present state of knowledge, it may be therefore unwise to disallow the diagnosis of major depression in a person meeting the severity, duration, and impairment criteria for that diagnosis just because the depressive state occurs in the context of a significant life event.

On the other hand, the removal of the bereavement exclusion criterion from the DSM-V diagnosis of major depression—a move that may be perceived as a further step in psychiatry’s attempt to pathologize normal human processes—requires strong and unequivocal research evidence. Some differences between bereavement-related and other life stressor-related depression found by Kendler et al. (6) (the lower percentage of bereaved individuals who sought treatment; the lower levels of neuroticism in those people) and by Wakefield et al. (7) (the lower proportion of bereaved people who reported that their condition interfered with life a lot) seem to point in the DSM-IV direction and deserve further investigation. Moreover, bereavement may be a quite different experience after the death, for instance, of a son or a friend (these events were included in the same category in the study by Kendler et al.) or in the elderly compared to younger people (the mean age at index episode, in Kendler et al.’s sample, was 35 years). Finally, if we expect DSM-V to be more widely used in various cultural contexts than DSM-IV, some cross-cultural validation of Kendler et al.’s findings is probably warranted.