Statistics from Altmetric.com
Multicentre, stratified, cluster randomised controlled trial.
Follow up period:
Thirty four schools in Queensland and New South Wales, Australia; time period not stated.
2479 young people in Year 9 (mean age 14.3 years), including 521 young people at high risk of depression (combined score for the Children’s Depression Inventory (CDI) and the Center for Epidemiological Studies-Depression Scale (CES-D) in the highest 20%).
Universal cognitive behavioural intervention (eight weekly 45–50 minute sessions), targeted cognitive behavioural intervention (eight weekly 90 minute sessions delivered only to young people at high risk of depression), universal plus targeted intervention, or no intervention.
Depressive symptoms (CDI; CES-D); depressive disorders (Anxiety Disorders Interview Schedule for Children, Longitudinal Interval Follow-Up Evaluation).
Patient follow up:
85% of high risk young people; overall follow up not reported.
There was no significant difference between universal and no intervention groups in change of depression scores for the full population of students (p>0.05). In young people at high risk of depression, there was no significant difference between treatment groups in change in CES-D depression score (p>0.05), but a significant difference was found in change between groups using the CDI score (p = 0.03); pairwise comparisons found no significant differences. In young people at high risk of depression, there was no significant difference between treatments in depression diagnoses after 18 months (18% with universal intervention alone v 21% with targeted intervention alone v 18% with universal plus targeted v 20% with no intervention, reported as not significant).
Universal and targeted cognitive behavioural approaches alone or in combination do not reduce depression diagnoses more than no intervention for young people at risk of depression.
Studies of cognitive behavioural therapy (CBT) for youth depression have yielded encouraging findings in some prevention and treatment trials,1 but recent reviews show marked variability across studies.2,4 Sheffield et al found that universal, indicated, and combined prevention programs failed to outperform an assessment-only control group among adolescents. The findings raise key questions.
What degree of training, supervision, and experience is sufficient for CBT? Group leaders for the intervention were school personnel and professional clinicians, most of whom may not have had much previous exposure to CBT. Leader training lasted one day, and leader supervision after training is not noted. It is possible that delivering effective CBT requires more training, supervision, and experience than was provided in this study.
What mechanisms of change are necessary for CBT to be successful? Sheffield et al wisely measured the skills targeted in their CBT intervention (negative thinking, ineffective problem solving). Their findings suggest that the youths may not have learned those skills. CBT may not work so well if its mechanisms of change are not activated. The intervention included some traditional CBT skills but omitted others (for example, scheduling pleasant events, relaxation, behavioural activation). Perhaps a richer array of CBT skill building would have made a difference.
Does self-monitoring have protective effects? Perhaps the assessment-only condition had unanticipated beneficial effects by promoting self-monitoring. Youths in the CBT group did improve, but youths who experienced only assessments showed equivalent improvements in symptoms and wellbeing. This might explain the absence of group differences in outcome.
What is the appropriate index of success? Because depression is episodic, and symptom reduction over time is common, perhaps symptom reduction is a less ideal outcome measure than, say, duration of post-treatment episodes or functional impairment during episodes.
Perhaps the most important question for practice is whether CBT is an effective intervention for adolescent depression. No single study can answer this question. The marked variability across studies in level of CBT benefit suggests that effects may depend on how the CBT is done, who delivers it, and how benefit is measured. In our view, it would be premature to dismiss CBT as ineffective; but the time is certainly right to ask hard questions about the conditions under which it does and does not work well.
For correspondence: Jeanie K Sheffield, School of Psychology, University of Queensland, Brisbane, Queensland 4072, Australia;
Sources of funding: not stated.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.