Dose–response relationship in cognitive behavioral therapy for depression: A nonlinear metaregression analysis.

Abstract

Objective: Evidence on the optimal “dose” of cognitive behavioral therapy (CBT) for treating major depressive disorder is sparse. This analysis aimed to evaluate the dose–response curve in CBT using a nonlinear approach, whereby “dose” was defined as number of treatment sessions. The dose–response curve of CBT was compared to other psychotherapies and pharmacological treatments for depression. Method: A systematic review and metaregression analysis of randomized controlled trials (RCTs) examining the efficacy of CBT in adults with acute depression was conducted. Treatment arms examining other psychosocial or pharmacological interventions were also analyzed. Cubic spline metaregression techniques were used to model nonlinear dose–response curves. Results: Seventy-two studies and 7,377 participants were included. Modeling the dose–response curve between change of depression symptom severity and the number of CBT sessions resulted in a nonlinear curve characterized by a strong improvement in symptom severity from baseline within the first eight sessions. Symptom reduction continues in the further course of the treatment, but at a slower pace. A similar pattern of symptom development was found for other therapies as well, although the prominence of early improvement and overall effect sizes vary across treatment arms. Conclusion: Results imply a general tendency for the strongest alleviation of depressive symptom severity in early stages of CBT treatment, thus, if aiming at symptom alleviation, speak for short CBT interventions. However, these findings have to be discussed in the light of the limited data regarding the sustainability of treatment effects in short-term therapies and effects beyond symptomatic changes.