Coping with Depression

The Adolescent Coping with Depression Course (CWD-A) is the most extensively tested treatment manual for depression in adolescence. CWD-A was adapted from an adult treatment manual (Lewinsohn, Antonuccio, Steinmetz-Brekenridge, & Teri, 1984) and aims at building skills in a didactic, highly structured format delivered within a group setting. Skills taught in treatment include relaxation, behavioral activation, social skills training, problem solving/conflict resolution, and recognizing and challenging distorted thoughts.

In the initial test of this manual in adolescents, 59 youth with moderate depression were recruited through health professionals, school counselors, and the media (Lewinsohn, Clarke, Hops, & Andrews, 1990). Enrolled youth were randomized to CWD-A, CWD-A with additional parent sessions, or a wait-list control group. CWD-A was delivered in 14 2-hour sessions over the course of 7 weeks. Upon termination from treatment, youth randomized to CWD-A and CWD-A with additional parental support had significantly lower rates of depressive diagnoses (57.1% and 52.4%, respectively) relative to youth randomized to the wait-list control group (94.7%). Six- and 24-month data suggested gains made by youth in active treatment were maintained after termination from treatment (Lewinsohn et al., 1990). A second test of the manual was conducted with 124 adolescents with moderate depression (Clarke, Rohde, Lewinsohn, Hops, & Seeley, 1999). Youth were randomized to CWD-A, CWD-A with parent group, or wait-list control. Youth in both active treatment arms were randomized to receipt of booster sessions following the acute phase of treatment. As in the prior trial, youth randomized to either CWD-A group had higher rates of recovery from depressive disorder diagnosis (66.7% for both groups combined) relative to rates of recovery among youth in the wait-list control (48.1%). Two years following treatment, youth in the active treatment arms experienced low rates of depression recurrence with no statistically significant difference associated with the receipt of booster sessions (25%).

The CWD-A manual has since been tested in more clinically complicated samples. Rohde, Clarke, Mace, Jorgensen, and Seeley (2004) recruited a sample of 93 youth from juvenile justice with comorbid major depressive disorder (MDD) and conduct disorder. Youth were randomized to either a modified version of CWD-A or an alternative intervention targeting potential skill deficits associated with their conduct disorder diagnoses, life skills (LS) training. Rohde and colleagues found youth assigned to CWD-A experienced significantly greater rates of recovery from MDD (39%) relative to youth randomized to the LS training (19%) at the posttreatment assessment; however, the groups did not differ significantly at 6-month (46.3% and 40.05% for CWD-A and LS training, respectively) and 12-month follow-up assessments (36.6% and 37.0%) (Rohde et al., 2004). Rates of recovery from conduct disorder diagnoses did not differ significantly between the CWD-A (9%) and LS training (17%) groups. Notably, in this comorbid sample the rate of recovery from MDD (39%) was lower than rates of recovery in the Clarke et al. (1999) trial (67%). Despite the superiority of CBT to control, overall lower rates of recovery in this comorbid sample may suggest the CWD-A manual is more efficacious when treating moderate depression without co-occurring disruptive behavior problems.

In a further test of the robustness of CWD-A, efficacy of the manual has been examined in treatment for the depressed adolescent offspring of depressed parents. Clarke and colleagues (2002) enrolled 88 depressed youth through large health maintenance organizations (HMOs) whose parents had a current or a recent (within the past 12 months) depressive episode. Adolescents were randomized to either 16 sessions of CWD-A delivered in a group format or treatment as usual (TAU) through their HMO. Rates of recovery from depressive illness did not differ significantly between youth randomized to CWD-A (31.6%) and TAU (29.8%) upon termination from treatment, nor were significant group differences found at 12- or 24-month follow-up assessments. Notably, the posttreatment response rate for CWD-A is lower in this trial than in any previous trials of this manual. This result is consistent with the broader negative impact of parental depression on youth mental health and response to treatment seen in other investigations (Garber et al., 2009; Rishel et al., 2006).

Overall, research on the CWD-A manual suggests that this intervention has acute benefits in reducing depressive symptoms in four randomized clinical trials. Youth receiving this intervention maintain gains as far as 2 years posttreatment. However, the efficacy of this intervention in samples with more clinically complicated presentations has been less well supported.

Pittsburgh Cognitive Therapy Study

In contrast with the highly structured, group-based CWD-A manual, the cognitive therapy manual developed in Pittsburgh by Brent and colleagues was developed to be delivered in an individual setting over 12 to 16 sessions (Weersing & Brent, 2003, 2006). The manual is principle based, allowing greater flexibility in the implementation of three core techniques: cognitive restructuring, behavioral activation, and problem solving. Application of these techniques is tailored to the individual patient and guided by cognitive case conceptualization. One randomized clinical trial of this treatment manual has been conducted to date (Brent et al., 1997). Adolescents (N = 107) with moderate to severe levels of depression were recruited through psychiatric referral and advertisement and randomized to one of three conditions: CBT, systemic behavior family therapy (SBFT), or nondirective supportive therapy (NST). Notably, youth in all three treatment arms received extensive psychoeducation. Youth in the study were clinically complicated, with over half the sample meeting criteria for a comorbid diagnosis and approximately one quarter with a history of suicide attempt. Rates of recovery were significantly higher for youth randomized to CBT (60%) relative to SBFT (37.9%) and NST (39.4%). At 2 years posttreatment, 80% of youth across groups had recovered, with no significant difference between groups (Birmaher et al., 2000), although high rates of service utilization among youth over follow-up make interpretation of this finding difficult.

Brief CBT

A briefer manual for CBT has been developed and tested among youth with a depressive disorder diagnosis within two randomized clinical trials in the United Kingdom. Vostanis, Feehan, Grattan, and Bickerton (1996b) randomized 57 moderately depressed youth to either brief CBT or a control intervention involving nondirective supportive treatment. Brief CBT included emotion recognition and labeling, social skill building, and challenging distorted cognitions. Treatment consisted of nine individual sessions delivered every other week over a maximum of 6 months; this is notably a lower dose of treatment spread over a longer length of time than in previous trials. As in the coping with depression course, sessions were highly structured and didactic. Youth treated with brief CBT did not demonstrate significantly lower rates of depression than youth treated with supportive therapy upon termination from treatment (Vostanis et al., 1996b). Rates of depressive disorders were low both among youth randomized to CBT (13%) and for youth randomized to supportive therapy (25%). This lack of a significant difference between treatment arms was also evident at 9-month and 2-year follow-up periods (Vostanis, Feehan, & Grattan, 1998; Vostanis, Feehan, Grattan, & Bickerton, 1996a). This lack of an effect may be compounded by natural remission rates of depression over time and relativity low doses of CBT.

In a second trial of a similar manual, Wood, Harrington, and Moore (1996) randomized 53 moderately depressed youth to brief CBT or a relaxation training (mean number of sessions = 6.4). In contrast to the results of Vostanis and colleagues, youth randomized to CBT had lower levels of depressive symptoms than youth in relaxation training immediately after treatment (Wood et al., 1996). Clinical remission was defined as a score ≤ 3 on the Clinical Global Impressions Scale (CGI). Youth randomized to CBT experienced greater levels of improvement at posttreatment assessment (54%) relative to youth in relaxation training (21%), but these differences were not significant at 3-month (45% and 25% for CBT and relaxation training, respectively) and 6-month (54% and 38%) follow-up.

Modular CBT

A final, critical manual in the treatment for depression in adolescents is modular CBT. This manual was developed through combining elements of the CWD-A and Pittsburgh cognitive manuals. This treatment takes a modular approach in which all youth receive six sessions building core skills in CBT. These core sessions include psychoeducation, goal-setting, mood monitoring, behavioral activation, social problem solving, and cognitive restructuring. In addition to these core sessions, modular CBT includes a number of additional sessions that can be used as needed for the individual teen in sessions 7 through 12. Additional sessions include a variety of potential skills to learn, including increasing social engagement, improving skills in communication, assertiveness training, and additional work with parents (Treatment for Adolescents with Depression Study [TADS] Team, 2004).

This manual was used in the largest treatment trial for depressed adolescents to date, the TADS. The sample of the TADS study was comprised of youth with moderate to severe depression. Youth enrolled in the TADS study were randomized to CBT, fluoxetine, combination treatment with fluoxetine and CBT, or pill placebo. After acute treatment, youth receiving combination treatment had the highest rates of response (71%) followed by the rates of response in youth receiving fluoxetine alone (61%). Youth receiving CBT experienced lower rates of response (43%), with no significant difference found between this condition and pill placebo (35%). Youth in the CBT condition continued to make gains over follow-up at weeks 18 and 36. Rates of response at week 18 were highest for combination treatment (85%), with rates of response to CBT alone (65%) comparable to rates of response in youth treated with fluoxetine (69%). By week 36, rates of response to CBT were comparable to both combination treatment and fluoxetine alone (TADS Team, 2007).

This finding lies in stark contrast with the above literature showing notably lower acute response rates than found in prior CBT trials. This outcome is particularly surprising given the support for both manuals used to develop the modular CBT manual; each of the contributing manuals had response rates approximately 20 percentage points higher (Brent et al., 1997; Clarke et al., 1999). Moreover, this finding lies in contrast to later tests of the same manual. Brent and colleagues (2008) recruited adolescents with clinically significant depression despite currently taking a selective serotonin reuptake inhibitor (SSRI). Youth were randomized to a change in medication alone or change in medication along with modular CBT. Youth in all treatment arms improved; however, youth receiving CBT in addition to either change in medication improved more than youth receiving medication alone (14% difference in response rate). Though data on use of the various modules in the TADS trial have not been published, one possible explanation for the difference in these two trials may lie in dosage of core components of CBT. Indeed, within the TORDIA sample, superior response was associated with receiving a higher dose of problem-solving skills and social skills training, relative to other components of the modular intervention (Kennard et al., 2009).

Cumulative Evidence for CBT

Across manuals, when core CBT skills (e.g., cognitive restructuring and behavioral activation) are delivered at an optimal dose and intensity (weekly for at least 12 weeks), CBT appears to be efficacious in treating depression in adolescence. The evidence to date supports the acute efficacy of CBT either when implemented in a highly structured group setting or more flexibly implemented in individual treatment. Despite consistent positive effects at posttreatment, the superiority of CBT as compared to other treatments diminishes over follow-up. This pattern of results appears to be driven in large part by catch-up effects in control conditions, as might be expected given the cyclical nature of depressive disorder (e.g., untreated episodes resolve in 9 months, on average [see Weersing & Weisz, 2002, for discussion; Kovacs, Obrosky, Gatsonis, & Richards, 1997]). The efficacy of CBT for clinically complicated youth and the effectiveness of CBT when delivered in real-world service settings are less clear. These issues are covered in greater detail in sections that follow.