The nature of the depressive experience in analogue and clinically depressed samples

Abstract

The use of university students as analogue subjects in depression research remains a controversial one, because it is not clear if the depressive experience of analogue and clinical samples differs qualitatively or quantitatively. Accordingly, this study directly compared 101 adult outpatients with major depressive disorder to 175 analogue depressed university students on self-report severity ratings of DSM-IV symptoms of major depressive episode. Several goodness-of-fit indices revealed that the covariance matrices of the depression symptoms were very similar in the two samples. These and other results from the study supported a phenomenological continuity hypothesis. Further, although the two samples differed quantitatively, the nature of the depressive experience did not appear to be one of transient and mild distress for many individuals in the analogue sample.

Introduction

Research on psychological processes in depression often relies on analogue studies of university students. A typical methodology involves administering the Beck Depression Inventory (BDI; Beck et al., 1961, Beck et al., 1979) to a large sample of students. Subsamples of `depressed' and `non-depressed' subjects are then identified and selected for further study based on a somewhat arbitrary cutoff on the BDI. This controversial research practice is based on the assumption that analogue and clinical forms of depression are part of the same continuum (i.e., differences are viewed as quantitative rather than qualitative).

Despite early cautions and criticisms on the use of the BDI for such purposes (Hammen, 1980; Sacco, 1981; Coyne and Gotlib, 1983; Gotlib, 1984), it continues to be a popular approach. Some investigators, sensitive to this issue, have utilized terms such as `dysphoric' (e.g., Clark et al., 1992) and `depressed mood' (e.g., Nolen-Hoeksema et al., 1993) in describing their analogue college samples. Importantly, however, the results obtained in these types of studies are still discussed in the context of psychological models of clinical depression and the presumed analogue–clinical continuum is retained.

The analogue–clinical issue continues to be a contentious one as reflected in several recent exchanges in prominent psychology journals (Vredenburg et al., 1993; Coyne, 1994; Kendall and Flannery-Schroeder, 1995; Tennen et al., 1995aTennen et al., 1995b; Weary et al., 1995; Flett et al., 1997). Following a critical and extensive review of the available evidence, Vredenburg et al. concluded that the use of analogue depressed college students was justified in depression research; yet these authors also underscored the glaring paucity of research that has directly compared analogue and clinically depressed samples. In contrast to the conclusion of Vredenburg et al., the review of Coyne (1994)concluded that the `self-reported distress' in college samples was mild, transient, and qualitatively distinct from a major depressive episode.

This important debate will not begin to be resolved without relevant empirical investigations designed to address this issue, and a call for such research has been a consistent theme in the literature. Flett et al. (1997)recently observed that most of the debate has focused on what they refer to as `phenomenological continuity'. This phenomenological continuity hypothesis states that the nature of the depressive experience differs in intensity (quantitatively), but not in kind (qualitatively), in analogue and clinical samples.

The present study therefore sought to directly compare analogue and clinically depressed samples on the phenomenology of the depressive experience. The most widely accepted definition of `clinical depression' in the literature, at least in North America, is the major depressive episode as described in recent editions of the Diagnostic and Statistical Manual of Mental Disorders (DSM). The current DSM-IV (APA, 1994) provides a detailed description of a major depressive episode including various groups of symptoms (affective, self-perception, cognitive, somatic, suicidal ideation), as well as their required duration (at least two weeks) and consistency (symptoms must be present most of the day, nearly every day). All of this information was used to develop a self-report assessment of the depressive experience with the purpose of gathering detailed information about the severity of all of the symptoms listed for major depressive episode. This measure was used to test the continuity hypothesis. Specifically, symptom patterns will emerge from analogue and clinical samples in this assessment. For example, if college students are more likely to experience distress and self-esteem issues whereas individuals with clinical depression are more likely to experience serious somatic symptoms such as weight loss, then this `qualitative difference' will be captured by the types of symptoms individuals report with the greatest severity. The covariance matrices for these major depression symptom ratings can be compared across samples using several goodness-of-fit indicators. If the two matrices do not provide a close match, this would indicate a qualitatively different nature of depressive experience in analogue and clinical samples, and the phenomenological continuity hypothesis would not be supported. Conversely, if the two matrices do provide a close match, this would provide support for the continuity position.