Depression-Focused Psychotherapies: Treatment Efficacy
Randomized clinical trials have compared the depression-focused psychotherapies with both psychosocial and pharmacological control conditions. This literature, published before 1995, has been reviewed extensively, and the major published controlled studies are summarized in Tables
IPT has been studied as an acute-phase treatment in five randomized clinical trials of outpatients with nonpsychotic major depressive disorder. In the initial study, IPT was superior to a nonscheduled supportive treatment and comparable to treatment with amitriptyline alone during a 12-week clinical trial. Some evidence from this trial suggested that IPT was more effective than amitriptyline in terms of improvements in mood, suicidal ideation, and interest, whereas the pharmacotherapy was superior in terms of resolution of appetite and sleep disturbances. Combined treatment had a modest additive benefit compared with either IPT or amitriptyline as monotherapy.
Shared Features of Depression-Focused Psychotherapies
The general equivalence of IPT and pharmacotherapy was reported in one study of elderly depressed outpatients. In this 16-week trial, IPT and nortriptyline were equally effective for patients who completed the protocol, although IPT was associated with a significantly greater retention of patients in the trial (16 out of 17 vs. 10 out of 18). Among patients who completed the protocol, both treatments resulted in 50%-60% reductions in mean depression symptom scores. By contrast, Reynolds et al. found that IPT plus pill placebo was significantly less effective than pharmacotherapy with nortriptyline and no more effective than pill placebo alone. Patients receiving the combination of IPT and nortriptyline were not significantly more likely to remit compared with those who received nortriptyline alone, although the study was not large enough to detect a moderate-sized additive difference. The study of Reynolds et al. (1999b) was unique in that all patients had bereavement-related depressive syndromes. It is not clear if IPT failed because more intense bereavement states require a different form of treatment or because the minimum duration of therapy (8 weeks for nonresponders) represented an inadequate trial for psychotherapy.
In the multicenter investigation sponsored by the National Institute of Mental Health, the Treatment of Depression Collaborative Research Program (TDCRP), IPT was compared with CT, active imipramine plus clinical management, and inert placebo plus clinical management. The efficacy of IPT was found to be comparable to that of both imipramine and CT by the end of a 16-week acute-phase protocol, although imipramine was more rapidly effective. Unlike imipramine, IPT was as effective for patients with atypical depression as for those with more classic neurovegetative profiles. Within the IPT group, subsequent analyses suggested that patients with less pronounced interpersonal problems did better and patients with personality disorders did worse. Barber and Muenz further examined specific personality traits. They found that IPT (in comparison with CT) was significantly more effective for patients with obsessive-compulsive traits and significantly less effective for those with avoidant traits. Two other groups reported an association between anxiety symptoms and poorer IPT response.
The study of Schulberg et al. tested IPT against both a standardized pharmacotherapy protocol (nortriptyline, titrated against plasma levels) and an unstructured treatment as a usual condition among patients treated in four urban primary care clinics. Results indicated that nortriptyline was significantly more rapidly effective but that IPT “caught up” during the latter half of the 8-month trial; both interventions were significantly more effective than the treatment-as-usual control condition (
Figure 40-1). A subsequent analysis suggested that standardized pharmacotherapy with nortriptyline was more cost-effective.
In the original study of DiMascio and colleagues, IPT was less effective than amitriptyline among patients who met the Research Diagnostic Criteria (RDC) for the endogenous depression subtype, whereas IPT was more effective in situational nonendogenous depression. These relationships were not confirmed in the subsequent TDCRP trial, in which IPT and imipramine were comparably effective in patients who met the RDC for the endogenous or situational subtypes. Moreover, IPT did relatively well in subsequent analyses of the TDCRP focusing on pretreatment symptom severity. IPT response also was not associated with pretreatment symptom severity response in several other studies.
In the original study, IPT also showed a relatively late-emerging advantage in terms of patients’ improved social adjustment. However, there was little evidence of selective or specific effects for IPT at the end of the acute-treatment phase of the TDCRP trial.
Markowitz and colleagues recently completed a study of 101 depressed men who were seropositive for human immunodeficiency virus (HIV). This study is not summarized on the evidence tables because the patient group was diagnostically heterogeneous and only about 45% met the criteria for a current episode of major depressive disorder. Patients were randomly assigned to one of four conditions: IPT (n = 24), CBT (n = 27), or supportive psychotherapy with (n = 26) or without (n = 24) imipramine pharmacotherapy. At the end of the 17-week protocol, IPT and the supportive therapy plus active imipramine were equally effective, and on most analyses, both of these treatments were more effective than CBT alone or supportive therapy alone. Markowitz and colleagues speculated that IPT may be a better fit than CBT for the real-world concerns of the depressed HIV-seropositive patient.
Three trials have investigated the efficacy of IPT as a preventive or prophylactic treatment. The first longer-term study of IPT should be considered a continuation-phase treatment trial because patients began immediately after responding to acute-phase pharmacotherapy. Patients were assigned to 8 months of treatment in one of four conditions that included weekly sessions with either a prototypical form of IPT (with or without active amitriptyline) or supportive management (with or without active amitriptyline). Results of this study suggested that IPT did not have a significant preventive effect after the withdrawal of amitriptyline. However, IPT resulted in a significant improvement in social adjustment at the 1-year point of follow-up for those who did not relapse.
In a second study, 128 patients with highly recurrent (unipolar) forms of major depressive disorder were enrolled in a maintenance-phase trial after first completing acute- and continuation-phase therapy of IPT and imipramine hydrochloride (mean dosage = 210 mg/day). All patients had achieved and maintained a clinically significant remission for at least 4 months on combined treatment before they entered the experimental phase of this protocol. Patients were randomly assigned to up to 3 years of care in one of five maintenance conditions: continued monthly IPT in combination with active imipramine, monthly IPT in combination with placebo, monthly IPT alone, imipramine with supportive management, and placebo with supportive management.
Monthly sessions of maintenance IPT produced a significant prolongation of well time compared with the placebo with supportive care condition. However, outcome was superior in both conditions in which patients received active maintenance pharmacotherapy. Beyond a nonsignificant trend during the first 9 months of maintenance therapy, there was no prophylactic advantage for patients receiving the combination of IPT and maintenance pharmacotherapy compared with patients receiving treatment with medication alone.
Subsequent analyses of the study by Frank et al. revealed two interesting correlates of longer-term IPT response. First, unmedicated depressed patients receiving maintenance IPT showed significantly greater risk of recurrent depression if their pretreatment electroencephalographic sleep pattern was characterized by an abnormally low amount of computer-scored delta (slow) waves during the first non-rapid eye movement sleep period. This finding suggested that patients had a more favorable outcome with maintenance IPT if their sleep profile showed a more normal pattern of slow-wave sleep.
Second, an independent assessment of the quality of IPT showed a large effect on treatment outcome. Individual therapists did not differ in their overall performance across cases, but there was a wide variation of performance across therapist-patient dyads. When the dyads were divided according to a median split, patients who participated in above-average dyads showed a substantially lower risk of recurrent depression than did those who participated in below-average dyads. In fact, the below-average IPT dyads showed a rate of recurrence after discontinuance of medication that was almost identical to the rate of recurrence in the placebo and clinical management group. The most important clinical implication of these findings is that IPT may be either a very powerful or a virtually useless maintenance treatment depending on the quality of the therapeutic dyad.
Spanier et al. subsequently evaluated the potential interaction of these two response modifiers. They found that the best outcomes were in the patients with normal slow-wave sleep who participated in higher-quality therapy dyads and the worst outcomes were in the group characterized by decreased slow-wave sleep and membership in a below-average therapy dyad (
Figure 40-2). A higher-quality course of therapy partly offset the increased vulnerability of the patients with diminished slow-wave sleep.
Reynolds et al. recently completed a study of preventive therapy with IPT and pharmacotherapy in a group of 107 elderly patients with recurrent depression (average age, 67.6 years). Patients were first stabilized on the combination of IPT and pharmacotherapy (nortriptyline; 80-120 ng/mL). After 4 months of sustained remission, patients were randomized to one of four maintenance treatment conditions: nortriptyline and clinical management, combination therapy, IPT alone, and placebo and clinical management. Both groups receiving IPT attended monthly therapy sessions. Results indicated that the combined condition had the best preventive effect, with both of the monotherapies being superior to the placebo plus clinical management condition (
Figure 40-3). The advantage of the combined condition over the monotherapies was particularly evident in the patients age 70 years and older. Thus, when compared to the study of Frank et al. (1990), the late-life group showed a relatively poorer protective effect with antidepressant monotherapy and hence a relatively greater long-term benefit for the combination of IPT and pharmacotherapy.
In summary, there is now reasonably broad empirical support for the utility of IPT for treatment of major depressive disorder. Overall, these effects are similar in magnitude to acute-phase antidepressant pharmacotherapy, although the time course of symptom reduction appears to be slower. IPT may be less effective for patients with prominent anxiety, certain types of personality pathology, and more pronounced interpersonal difficulties. Such patients may do better if they are treated with the combination of IPT and pharmacotherapy. The combination of IPT and pharmacotherapy is an effective longer-term treatment for prevention of recurrent depression. For both midlife and late-life patients, monthly sessions of IPT have some preventive value after withdrawal of antidepressants, particularly for those in higher-quality therapy dyads. The value of longer-term therapy for patients who respond to IPT alone during the acute phase has not been established. Frank and colleagues are currently conducting a study comparing three schedules of maintenance IPT (weekly, every other week, and monthly) in a sample of women with recurrent depression who responded to psychotherapy alone.
A relatively large number of older studies have examined the efficacy of the various forms of behavior therapy in relation to either waiting-list control conditions or tricyclic antidepressants. When these studies are considered together, the weight of evidence is rather impressive (see Tables
Table 40-4). The AHCPR meta-analysis showed unequivocal efficacy for behavior therapy in relation to the low-contact and waiting-list control conditions. Moreover, behavior therapy appeared to have a small advantage in relation to dynamic psychotherapies (9.1% ± 19.9%; 6 studies). An apparently large effect was observed in comparison with pharmacotherapy (23.9% ± 11.6%), although this conclusion was based on only two older studies. However, it should be noted that the fidelity of both the dynamic psychotherapy and the pharmacotherapy comparison conditions in these studies has been questioned.
When each particular model of behavior therapy is considered separately, the strength of the evidence is much less robust. For example, only one controlled study has evaluated the efficacy of social skills training, and only one has evaluated social learning therapy. In a controlled trial of 120 female outpatients with major depressive disorder, Hersen et al. found social skills training (plus placebo) to be no more or less effective than either amitriptyline hydrochloride or short-term dynamic psychotherapy (plus placebo). Skills training did provide a significant advantage on several behavioral indices of assertiveness, but the real-world benefits of these differences have not been established. McLean and Hakstian reported the results of a study of 178 depressed outpatients. Patients were randomly assigned to 10 weeks of acute-phase treatment with a behavior therapy program or one of three comparison conditions: amitriptyline pharmacotherapy, relaxation training, or nondirective psychotherapy. The pattern of results generally favored behavior therapy over the other treatments, with the behavioral control condition, relaxation training, making a surprisingly strong intermediate showing. However, the advantage for the behavior therapy condition dissipated somewhat by 3-month follow-up.
Results of randomized controlled clinical trials generally suggest that behavior therapy produces symptomatic improvements comparable to those seen with other active psychological treatments, including Beck’s model of CT and nondirective or psychodynamically oriented therapies. When differences are observed, results have favored behavior therapy.
The best-studied behavioral treatments are Rehm’s self-control therapy and Lewinsohn’s psychoeducational model of treatment. Unfortunately, these treatments have not been compared with pharmacotherapy or alternative forms of psychotherapy delivered by professionally allegiant clinicians (i.e., those adhering to the model).
Most recently, N. S. Jacobson and colleagues conducted a study comparing a relatively simple treatment, behavioral activation, with both the full model of CT and a cognitive-behavioral intervention without the focus on core schemas. The same therapists conducted all three treatments. The study group consisted of 38 men and 113 women with mild to moderate major depressive disorder; about 80% of the patients were referred from a health maintenance organization. The acute phase included 20 sessions, and the patients received 6 months of additional follow-up. The investigators found that the three treatments were equally effective, with response rates ranging from 58% to 68% (
Figure 40-4). The severity of pretreatment symptoms did not predict differential response, and relapse rates were similar for all treatments across the 6 months of follow-up. The most important aspect of this study is its finding that behavioral activation appears to be well suited for use by relatively junior, nondoctoral therapists. Before his untimely death, Jacobson initiated a study contrasting behavior activation with both the full model of CT and pharmacotherapy. Results of this important study should be available in several years.
Another fairly recent study compared a brief model of problem-solving therapy, consisting of only 6 sessions, with clinical management and administration of either amitriptyline (mean dosage = 139 mg/day) or a pill placebo. The study was conducted in primary care clinics. The problem-solving intervention was so brief that the amount of clinical contact did not differ significantly across conditions. Both active treatments were significantly more effective than the pill placebo condition after 6 and 12 weeks of treatment. Remission rates at week 12 were as follows: problem-solving therapy, 60% (18 of 30); amitriptyline, 52% (16 of 31); and placebo, 27% (8 of 30).
Meta-analysis indicates that group (51%) and individual (58%) behavior therapy are roughly comparable in efficacy in the treatment of depressed outpatients. This conclusion is supported by the results of one randomized clinical trial directly comparing group and individual modalities. Behavioral marital therapy also has shown promise in two controlled trials.
There have been relatively few studies of the long-term efficacy of behavior therapy. In the single small published trial of a continuation therapy, Hersen et al. (1984) reported comparable outcomes in patients receiving six monthly sessions of social skills training, dynamic psychotherapy, or continuation pharmacotherapy. The apparent comparability of these treatments is compromised by the fact that the patients in the continuation phase of the trial were no longer necessarily representative of the larger sample (i.e., they first had to respond to their respective modalities).
In naturalistic follow-up studies, evidence in favor of an enduring beneficial effect of behavior therapy (relative to control conditions) has been reported by some investigators, but evidence of differential prophylaxis is not unanimous. Relapse rates as high as 50% or 87% after the end of acute-phase behavioral treatment have been reported. Methodological differences between studies preclude more definitive conclusions.
In summary, an older literature review suggests that behavior therapy is a credible treatment for depressed outpatients, although the most compelling evidence of efficacy comes from studies using waiting-list control groups. After a period of relative quiescence, there is some renewed interest in simpler models of behavioral treatment as a result of the studies of N. S. Jacobson et al. and Mynors-Wallis et al.. These studies, perhaps particularly timely within the context of cost-effectiveness, illustrate the potential utility of relatively simple, focused interventions aimed at helping depressed people to become more active and engaged in problem solving. Although it seems likely that these less sophisticated approaches will be less beneficial for patients with more severe, complicated depressions, the incremental benefit of more fully developed models of CBT now warrants serious reexamination.
Beck’s model of CT is the best-studied psychological treatment of major depression. CT has been extensively studied in comparison with both waiting-list control conditions and other forms of psychotherapy (see
Table 40-3). However, despite such intensive study, only two published trials of CT have included a placebo plus clinical management condition.
There is no doubt about the efficacy of CT as an acute-phase treatment compared with waiting-list control conditions. In the AHCPR’s meta-analysis, CT had an overall efficacy rate of 46.6% (±6.9%) and an advantage of 30% (±22%) when compared with waiting-list control conditions (seeTable 40-4). Although CT was not consistently superior to the placebo plus clinical management condition in the TDCRP study, a more recent subanalysis did suggest efficacy in the subset of patients with features of atypical depression. Moreover, Jarrett et al. (1999) found CT to be equal to phenelzine and superior to pill placebo in a well-controlled 10-week double-blind study of 108 patients with atypical depression (
Figure 40-5). Both active treatments yielded 58% intent-to-treat response rates, compared with 28% in the pill placebo group. Differential attrition from the pill placebo condition hampered interpretation of these data somewhat, although significant differences were apparent at week 4 (i.e., before there was a pronounced difference in dropout rates).
With respect to comparisons with pharmacotherapy, results of four studies indicate that CT is superior to treatment-as-usual medication control conditions, whether CT is administered alone or in combination with treatment as usual. In each of these studies, a primary care provider prescribed the pharmacotherapy. Results from studies utilizing more rigorous pharmacotherapy conditions have yielded more consistent evidence of parity (seeTable 40-3). Exceptions to such parity include an initial study by Beck’s group, in which results favored CT over imipramine hydrochloride, Murphy et al.‘s small study of depressed outpatients (in which both CT and relaxation training were significantly more effective than desipramine), Markowitz et al.‘s study of mildly depressed HIV-seropositive men, in which imipramine was more effective than CT, and the TDCRP study. In the TDCRP study, pharmacotherapy was more effective than CT in a more severely depressed subgroup of patients and with respect to rapidity of improvement. The AHCPR meta-analysis—which did not include the studies of Hollon et al., McKnight et al., Murphy et al., Blackburn and Moore, and Jarrett et al. —revealed a modest 15.3% (±26.1%) advantage for CT over pharmacotherapy on the basis of three suitable studies (see
Table 40-4). Inclusion of the five more recent studies probably would eliminate the difference between CT and pharmacotherapy.
Recently, M. B. Keller et al. studied a modified form of CT and the antidepressant nefazodone singly and in combination, in a large (N = 681) multicenter trial of patients with chronic major or “double” depressive disorders. The modification of CT, developed by McCullough, emphasized the use of situational analysis of interpersonal interchanges to help chronically depressed patients learn more specific, goal-directed approaches to improving relationships. The two monotherapies were comparably effective at week 12, although nefazodone was more rapidly effective. The combination of CT and pharmacotherapy was markedly more effective than both of the monotherapies (e.g., intent-to-treat response rates: CT alone, 48%; nefazodone alone, 48%; in combination, 73%). Analyses of the rates of change in symptom measures suggested that the combined condition benefited by having both the early symptom effects of nefazodone and the later emerging symptom effects of psychotherapy.
CT has generally been found to have an efficacy comparable to that of other active psychotherapies, including behavioral marital therapy, behavior therapy, IPT, brief dynamic therapy, pastoral counseling, and nondirective group therapy. In the study by Murphy et al., CT was not significantly more effective than relaxation training (e.g., patients receiving a final Beck Depression Inventory score of less than 9: CT, 82%; relaxation training, 73%), although the latter therapy was intended to serve as a nonspecific control group. As discussed previously, CT was less effective than IPT in Markowitz and colleagues’ (1998) study of mildly depressed HIV-seropositive men. In N. S. Jacobson et al.‘s first study, individual CT appeared to be more effective than behavioral marital therapy in a subset of patients with satisfactory marriages, whereas the latter treatment produced greater gains on measures of marital satisfaction in maritally distressed couples. Interestingly, a combination of individual CT and behavioral marital therapy was no more effective than the individual modalities.
The Sheffield project compared CBT and dynamic-interpersonal therapy, each provided for either 8 or 16 weeks. Although neither treatment was identical to the models of CT and IPT described previously, they are sufficiently similar to be relevant to this review. At the end of acute treatment, the two therapies were comparably effective. There was an interaction between pretreatment symptom severity and duration of therapy, with more severely symptomatic patients benefiting more from longer courses of treatment. Trends favoring the longer course of CT also were evident after 1 year of follow-up.
In the AHCPR meta-analysis, individual CT (50.1%) appeared somewhat more effective than group CT (39.2%). Published comparative studies have similarly documented a slight advantage for individual CT compared with group CT. When CT has been compared with other group therapies, both positive and negative findings have been reported. Ravindran et al. (1999) recently found that group CT added little symptomatic benefit when combined with sertraline in a study of 97 patients with primary dysthymic disorder.
CT is the only form of individual psychological treatment to be specifically modified for treatment of hospitalized depressed patients. Initial “open” treatment experiences were described by Shaw and J. Scott. In a small but controlled study of 30 patients with antidepressant-resistant, “neurotic” major depression, deJong et al. reported inpatient CT to be significantly more effective than either the low-contact inpatient or the outpatient comparison condition. (In the low-contact condition, inpatients received only a single weekly cognitive restructuring group. The outpatients were on a waiting list for hospitalization.) In an “open” study of 16 unmedicated patients with features of endogenous depression, Thase et al. (1991a) reported a response rate of 81% (13 patients) after up to 4 weeks of CT. A subsequent report expanded the series to 30 patients, with 9 of the next 14 patients also responding. Poorer outcomes were observed among patients with significant comorbidity and/or a history of nonresponse to antidepressant medication. In two controlled studies of combined treatment (Bowers 1990; Miller et al. 1989), some evidence favored CT plus antidepressants over antidepressants alone. In Bowers’s study, however, the advantage of adding CT to standard inpatient treatment was significantly greater than that observed in the attention-control group that received relaxation training. In the study by Miller and associates, the additive benefit was largely limited to a subgroup of patients with high levels of dysfunctional attitudes.
During naturalistic follow-up, patients who were treated with CT generally fared better over 1- or 2-year follow-ups than patients who were treated with treatment-as-usual interventions or those who were withdrawn from antidepressant pharmacotherapy. However, in the TDCRP study, no appreciable difference in relapse rates was found among patients who responded to CT, IPT, imipramine, or placebo. It should be noted that acute-phase therapy with CT was not significantly more effective than placebo in this trial and that, in the absence of significant acute-phase efficacy, the assumption of prophylaxis may not hold true. Comparability in survival rates among several forms of active psychotherapy has been observed in two other follow-up studies.
In the naturalistic studies of Blackburn et al. and Simons et al., relapse after CT or pharmacotherapy was heralded by high levels of residual symptoms and/or high residual scores on a measure of dysfunctional attitudes. These findings were replicated by Thase et al. (1992) during a controlled 1-year follow-up of 50 CT responders. In this latter study, the relapse rate at 1-year follow-up was only 10% among patients who terminated 16 weeks of individual CT after having achieved at least 2 months of symptom remission (i.e., Hamilton Rating Scale for Depression (HAM-D) score of 6 or less for 4 weeks). By contrast, patients with less complete or sustained remission had a relapse rate of more than 50%. The magnitude of this difference suggests that termination of time-limited therapy should be determined by the quality of each particular patient’s response and should not occur solely on the basis of completion of a predetermined number of sessions.
Several studies have evaluated CT as a continuation- or maintenance-phase therapy. In one controlled trial, sessions of monthly CT and continuation pharmacotherapy had comparable prophylactic effects over 6 months. Two other early studies, employing relatively small samples and a type of CBT not entirely consistent with Beck’s model of treatment, failed to find any added prophylactic value for continued “booster” therapy sessions after termination of acute treatment. More recently, Jarrett and Kraft (1997) observed that an 8-month course of continuation-phase CT (10 sessions) reduced relapse rates by about 50% compared with a historical control group that received only acute-phase therapy. The data of Jarrett and Kraft (1997) suggest that as the vulnerability of a particular patient for relapse/recurrence increases, the cost-effectiveness of longer-term models of CT increases.
More recent studies have taken two forms: the use of CT both sequentially to treat residual symptoms and facilitate withdrawal of antidepressants and as a more long-term preventive treatment. Fava et al. (1994) randomized 40 patients with residual depressive symptoms despite adequate pharmacotherapy to either 10 sessions of CT (in addition to ongoing pharmacotherapy) or continued pharmacotherapy alone. CT had a significant effect in reducing residual symptoms. During the first 4 years of follow-up, this brief intervention significantly reduced the risk of relapse in comparison to the pharmacotherapy-alone group.
Fava et al. (1998) next studied 40 antidepressant responders with highly recurrent depression; all patients had been in remission for at least 10 weeks before entering the study. During the following weeks, patients were randomly assigned to receive either 10 sessions of CT or standard clinical management while antidepressants were withdrawn slowly. Fava et al. again observed highly significant effects favoring CT, in terms of both symptom reductions and protection against recurrent depressive episodes (
The results of a large, independent replication study of this approach have recently been published. In this two-center study, 158 partially remitted patients taking antidepressants were randomized to a 16-session, 20-week course of either CT in combination with pharmacotherapy or pharmacotherapy alone. Over the next 68 weeks, 45% of the control group relapsed despite maintenance pharmacotherapy, compared with only 29% of the group that received CT (P = 0.02). A similar difference favored the CT group in terms of proportion of patients achieving full remission.
Blackburn and Moore (1997) compared continuation and maintenance treatment strategies in a sample of 60 patients with recurrent major depression; 37 patients had responded to acute-phase antidepressant therapy and 23 had responded to CT. The former group was randomly allocated to receive either 24 months of maintenance pharmacotherapy or a switch to maintenance CT (three visits in month 1, two visits in month 2, and monthly visits thereafter). All CT responders received maintenance psychotherapy. Results over the next 12 months demonstrated comparable benefits on symptom measures and relapse risks. For example, 31% (4 of 13) relapsed despite maintenance pharmacotherapy, 36% (5 of 14) relapsed after switching to CT, and 24% (4 of 17) relapsed despite maintenance CT. Although this design does not permit definitive conclusions (the risk of recurrent depression without active treatment in this sample was not determined), the results observed in the switch group are similar to those reported by Fava et al. and Paykel et al.
One advantage of the relatively large number of studies of CT is that evidence about correlates or predictors of response has emerged (see Whisman 1993 for a detailed review of earlier studies). Specifically, single or unmarried status, high pretreatment levels of dysfunctional attitudes, chronicity, and increased initial symptom severity have been associated with poorer outcomes after 10-16 weeks of treatment with CT . Men and women appear to be equally responsive to CT, although more severely depressed women may have somewhat poorer outcomes. Patients with certain comorbid personality disorders appear to respond as well to CT as do patients with no personality disorders, although Barber and Muenz (1996) found that those with obsessive-compulsive traits were less responsive to CT than to IPT. It should be noted that research trials of major depressive disorder typically have excluded patients with more severe Axis II psychopathology, such as those with antisocial and/or borderline personality disorders. Such patients were more likely to drop out of CT in a private-practice study.
Positive correlates of CT outcome may include high pretreatment levels of learned resourcefulness or self-efficacy, optimism, motivation, and homework compliance. The relationship between learned resourcefulness and CT outcome has not been replicated by all groups and may be relevant to only more severely depressed patients. High levels of therapist core skills have been associated with favorable outcomes in one study but not in others. Years of therapeutic experience and technical competence in structuring therapy have also been associated with better outcomes. Nevertheless, this line of research warrants further attention, particularly in trying to understand the results of studies such as the TDCRP, in which the performance of CT varied across sites.
Revision date: July 5, 2011
Last revised: by Janet A. Staessen, MD, PhD