Stress Generation and Recurrent Depression: The Role of Differential Treatment Response

The National Institute of Mental Health (NIMH) Treatment of Depression Collaborative Research Programme (TDCRP) (Elkin et al., 1989; Sotsky et al., 1991) compared three forms of treatment for depression -- imipramine plus clinical management (IMI-CM), cognitive behaviour therapy alone (CBT), interpersonal therapy alone (IPT) -- against a placebo control plus clinical management (PLA-CM) condition. These three treatments were found equally effective in the treatment of the index episode of depression when compared to the placebo control (Elkin et al., 1989). The results from the TDCRP study also indicated that patient characteristics, irrespective of treatment modality, were predictive of treatment effects. Six patient dimensions -- social dysfunction, cognitive dysfunction, expectation for improvement, endogenous features, double depression and duration of current episode -- were all found to be significant predictors of outcome (Sotsky et al., 1991). Patient characteristics were also found to be associated with differential outcome depending on treatment modality. Elevated social dysfunction, for example, interfered with successful outcome in IPT, whereas cognitive dysfunction hindered successful outcome with CBT. Cognitive dysfunction also predicted poor treatment response in the IMI-CM condition. Cognitive vulnerability would be expected to mediate response to treatment in CBT, as the presumed mechanism of change is dysfunctional depressogenic cognitions (e.g., Beck et al., 1979; Whisman, 1993). The finding that cognitive vulnerability was also implicated in treatment response to a pharmacological intervention is without theoretical explanation or specific causal agency. The purpose of the proposed research is to further examine the relationship between treatment outcome and patient characteristics associated depression. In particular, the relationship between treatment outcome and two personality/cognitive characteristics implicated as vulnerability factors for depression - self-criticism and dependency - will be explored. The first aim of the proposed investigation is to examine the relationship between change in self-criticism and dependency and the relationship of this change to treatment response and reduction in severity of depression. Mode-specific outcome measures will be administered to outpatients with major depression randomly assigned and treated with CBT, IPT or PHT. The second aim of this study is to examine the issue of whether successful targeting of either cognitive dysfunction in CBT or interpersonal functioning in IPT will result in a reduced risk for relapse and/or recurrence of a major depressive episode. We believe that the dependent type, who centres their definition of self entirely around issues of affiliation with others, would be more likely to show changes related to such themes in a therapeutic intervention specifically designed to address interpersonal problems. In contrast, the self-critic who defines self primarily around issues related to achievement and engages in excessively harsh and unrealistic self-appraisal would respond best to an intervention that targeted these punitive cognitions. HYPOTHESES/RESEARCH QUESTIONS Prediction of Treatment Outcome (Objective 1): Two sets of hypotheses are proposed. In all analyses the DEQ will be used to assess self-criticism and dependency. The first set of hypotheses involves mode specific treatment outcomes and the second set of hypotheses address differences in the mechanisms of change across the treatments. The first set of hypotheses are: (a) all treatments will be equally effective in the treatment of the index episode, (b) baseline self-criticism and dependency scores will predict outcome in all treatments, with higher self-criticism and dependency scores related to poor outcome, (c) CBT will demonstrate greater specificity for targeting self-criticism than will either PHT or IPT, (d) IPT will demonstrate greater specificity for treating interpersonal functioning than will either PHT or CBT, (e) PHT will demonstrate greater specificity for treating endogenous symptoms than will either CBT or IPT. The second set of hypotheses are: (a) change in self-criticism scores and dysfunctional cognitions will mediate a positive treatment response in CBT but not in IPT or PHT, (b) change in dependency scores and interpersonal deficits will mediate positive treatment response in IPT but not in CBT or PHT, (c) change in endogeneity will mediate positive treatment response in PHT but not in CBT or IPT. Prediction of Relapse and Recurrence (Objective 2): It is hypothesized that: (a) CBT and IPT will produce a lower rate of relapse and recurrence than PHT because of the greater reduction in stable dysfunctional cognitions related to either self-critical and/or interpersonal vulnerabilities; (b) in cases where interpersonal vulnerabilities are predominant, IPT will produce lower rates of relapse and recurrence than either CBT and PHT, in cases where self-critical vulnerabilities are predominant, CBT will produce lower rates of relapse and recurrence than either IPT or PHT.