Concern has been raised recently about the efficacy and safety of antidepressants for treatment of depression in children. Some clinicians, therefore, have suggested that psychotherapy be started before medication treatment of depression in children. How effective is psychotherapy for the treatment of depression in children and adolescents?
Weisz and associates1 recently conducted a meta-analysis to assess the effects of psychotherapy for the treatment of depression in children and adolescents. This is the largest study sample to date on this topic. Given the significance of the findings, it is important to understand what kinds of studies were included in the analysis.
The authors applied strict inclusion criteria. Patients were required to have elevated levels of depressive symptoms, a diagnosis of a major depressive disorder or dysthymic disorder, or research diagnostic criteria diagnoses of minor or intermittent depression. Patients had to be randomly assigned to at least 1 active treatment group and to 1 nonactive treatment group (untreated, waiting list, minimally treated, or placebo-control group). The study sample had to have a mean age of 19 years or younger. Finally, the treatments used had to target depressive symptoms or disorder. In addition to including published studies, the authors also included non–peer-reviewed studies and doctoral dissertations in order to avoid possible publication bias.
Thirty-five studies met criteria for inclusion. The number of participants in the separate studies ranged from 9 to 439. Many types of psychotherapy were used, including cognitive-behavioral therapy with and without a family component; systemic behavioral family therapy; interpersonal psychotherapy; attachment-based family treatment; social skills training; relaxation training; social competency training; structured learning therapy; self-control; behavioral problem solving; rational emotive therapy; role-playing; and cognitive bibliotherapy. The patients were obtained from a wide variety of settings, such as the community, schools, day-treatment programs, and outpatient settings. The age of the participants included in the sample ranged from 7 to 19 years.
A number of important clinical questions were addressed in the meta-analysis. First, is psychotherapy effective for depression in children and adolescents? The meta-analysis showed that the overall mean effect size for psychotherapy was 0.34, which is a modest effect size. (In general, a small effect size is about 0.20 and a large effect size is about 0.80.)
The investigators also compared the effect size of psychotherapy for depression to psychotherapy for other childhood conditions such as aggression, attention deficit hyperactivity disorder, and fears. The effect size for these nondepression conditions was 0.69, which was significantly higher than the effect of psychotherapy for depression.
Second, does the effect of psychotherapy for depression in children and adolescents last? At follow-up assessments 1 year after treatment, there was practically no treatment effect. In essence, the improvements made during the course of acute treatment were not apparent a year later.
Third, are psychotherapy treatments that focus on changing cognitions more effective than those that do not address changing cognitions? The analysis showed that the mean effect size for cognitive treatments was 0.35 and for noncognitive treatments was 0.47 Ñ not a significant difference. Therefore, there was no advantage in using treatment that emphasized change in cognitions, such as cognitive-behavioral therapy.
Fourth, does psychotherapy have a beneficial effect on suicidality? Six of the studies in the sample included a measure of suicidality and were used for this analysis. The mean effect size for the suicidality measures was 0.18. Therefore, the psychotherapy did not seem to have an effect on improvement in suicidal thinking and behavior.
Overall, the results of the meta-analysis by Weisz and colleagues showed that psychotherapy for depression in children and adolescents had a modest effect and that this effect was not long lasting. These results raise concern as to why the psychotherapy was not more effective for youth.
Possible explanations
Study design issues such as patient characteristics, comorbid conditions, treatment duration, assessment measures, skill of the therapist, and concurrent treatments are some factors that may affect study outcome. These may be related to the finding of weak efficacy for psychotherapy in depressed children and adolescents.
Another consideration, perhaps a more crucial one, is that the leading psychotherapy treatments for depression were developed for adults. Cognitive-behavioral therapy, which focuses on changing a personÕs thinking in order to improve depressive symptoms, was developed for the treatment of depressed adults. Similarly, interpersonal psychotherapy, which focuses on interpersonal behavior, was developed for depressed adults. Although these treatments have been modified for use in depressed adolescents, their origins are in the treatment of adult psychopathology. It may be time to develop new psychotherapy models that are based on specific developmental levels (eg, social, emotional, and cognitive features) specific to children and specific to adolescents in order to find more effective psychotherapy treatments.
Depression in children and adolescents is a serious disorder that impairs a child's overall functioning and family life. Better medication treatments and better psychotherapy treatments are desperately needed for these children. It may be time to substantially increase collaboration among psychopharmacology researchers and psychotherapy researchers in the area of childhood depression, in the pharmaceutical industry, and in the National Institute of Mental Health in order to identify more effective treatments for childhood depression.
Dr Wagner is the Robert L. Stubblefield Professor in the department of psychiatry and behavioral sciences and director of child and adolescent psychiatry at the University of Texas Medical Branch at Galveston.
