Research reportUntreated short-term course of major depression: a meta-analysis of outcomes from studies using wait-list control groups
Introduction
Approximately two-thirds of depressed patients initiated on an antidepressant will respond following an adequate trial. Since one-third typically improve with placebo, many of these patients may actually be responding to the non-specific effects of receiving treatment. Several factors have been postulated to account for the high placebo response rates in antidepressant trials: the instillation of hope, positive expectations, the desire to please, and clinician bias, among others (White et al., 1985, Harrington, 1999). Controlled experiments have been performed to manipulate some of these variables in an attempt to identify their relative contributions. However, one factor not accounted for in these models is the occurrence of spontaneous remission, and it is unknown how large a component this is of the placebo, or antidepressant, response.
Although case reports exist documenting instances of spontaneous remission (Endicott and Endicott, 1963, Malen et al., 1975, Lambert, 1976), this has rarely been the focus of systematic inquiry. This should not be surprising. The moment a patient engages in treatment, any remission of symptoms can no longer be considered ‘spontaneous’ (i.e. occurring outside the context of formal treatment). Since major depression is often a self-limiting disorder that tends to resolve over time, many apparent treatment responders are undoubtedly experiencing a natural remission of symptoms which only coincidentally occurs during treatment. As yet, however, we are unable to distinguish between these two types of ‘responses.’
While the placebo pill has long been used by psychopharmacologists to control in part for the occurrence of spontaneous remission, nothing comparable exists for those studying the efficacy of psychotherapy (Parloff, 1986, Strayhorn, 1987). Psychotherapy researchers therefore have been forced to utilize a ‘wait-list control’ (WLC) design. In this format, a certain proportion of subjects is randomly assigned to a WLC group. Following baseline assessments, subjects are told that they qualify for the study, but cannot begin treatment for several weeks because all the slots have been filled. Upon entering the active phase of treatment, they are reassessed, and any change in symptomatology during the waiting period is measured. This serves as the control which accounts for the passage of time when evaluating the efficacy of a psychotherapeutic modality.
Here then is one instance where the course of untreated depression has undergone systematic assessment. Though not identical to the design of antidepressant trials, many similarities exist: subjects are usually recruited through advertisements, they are screened with the same, or similar, instruments, and baseline and post-trial scores are reported. In these studies, of course, the WLC subjects are merely used as a control group, and are not the focus of interest. However, in order to better understand the short-term course of untreated depression, we performed a meta-analysis evaluating the outcomes of these subjects. In particular, we sought to address the following questions: (1) what changes in symptomatology in major depression occur in the short term without treatment, especially during the 4–8 week time frame used in most antidepressant trials? (2) are there differences in rates of spontaneous improvement depending on the severity of the disorder at presentation? and (3) what percent of subjects with major depressive disorder improve without treatment to a degree comparable to what would be considered a positive ‘response’ in an antidepressant trial?
Section snippets
Method
No database exists that readily identifies studies employing WLC groups (we will consider ‘wait-list control’ as synonymous with ‘delayed treatment’). We therefore searched all potentially relevant studies included in seven recent meta-analyses that have evaluated the efficacy of psychotherapy (Steinbrueck et al., 1983, Nietzel et al., 1987, Bowers and Clum, 1988, Dobson, 1989, Robinson et al., 1990, Svartberg and Stiles, 1991, Jorgensen et al., 1998) as well as a standard text on psychotherapy
Results
Table 2 presents the baseline depression rating scores from the 19 studies, number of weeks on the waiting-list, and changes in scores over time. The mean baseline BDI score in 18 studies was 24.3, while the mean baseline Ham-D in 9 studies was 23.4. The mean decrease in BDI scores, weighted by the number of subjects in each study, was 15.7%. The mean decrease in Ham-D scores, weighted by number of subjects in each study, was 11.9%.
In order to estimate the mean decrease in symptomatology over a
Discussion
Our analysis of 19 studies involving 221 depressed subjects randomized to a waiting list for 2–20 weeks found a mean decrease in symptomatology of 10–15%. A sub-analysis of 11 studies that obtained depression rating scores between weeks 4 and 8 — the time frame used in most antidepressant trials — yielded similar results. We therefore would postulate that subjects enrolled in short-term antidepressant trials probably improve on their own by this amount. An analysis of seven studies involving
References (59)
- et al.
The treatment of depression occurring in the context of marital discord
Behav. Ther.
(1986) - et al.
Cognitive-behavioral treatment for depressed adolescents
Behav. Ther.
(1990) Comparison of effectiveness of group interventions for depression in women
Arch. Psychiatr. Nurs.
(1993)An evaluation of a brief group therapy for depression
J. Consult. Clin. Psychol.
(1979)- et al.
Stimulating therapeutic change with interpretations: a comparison of positive and negative connotation
J. Couns. Psychol.
(1982) - et al.
Handbook of Psychotherapy and Behavior Change
(1994) - et al.
Relative contribution of specific and nonspecific treatment effects: Meta-analysis of placebo-controlled behavior therapy research
Psycholog. Bull.
(1988) - et al.
A psychoeducational approach to the treatment of depression: comparison of group, individual, and minimal contact procedures
J. Consult. Clin. Psychol.
(1984) - et al.
Strange bedfellows: reward, punishment, and impersonal distraction strategies in treating dysphoria
Cog. Ther. Res.
(1979) Effects of cognitive and behavioral group treatment on the depressive symptomatology of Puerto Rican women
J. Consult. Clin. Psychol.
(1981)