Twenty-nine RCTs were included in the review (n=3,000 patients). Thirteen trials included participants (n=1,139) with depressive disorders. Sixteen trials included participants (n=1,861) with depressive symptoms. Trial quality was reported to be highly varied; the Amsterdam-Maastricht score ranged from 8 to 16.
Depressive disorder studies: Cognitive-behavioural therapy was associated with significantly greater improvement in depressive symptoms (SMD -0.83, 95% CI -1.36 to 0.31; 13 RCTs; n=1,037 participants) compared with control conditions. There was evidence of high heterogeneity (I2=93.0%). This was largely explained by three trials regarded as 'outliers'; effect sizes were smaller (but still statistically significant) without the 'outliers'.
Depressive symptom studies: Cognitive-behavioural therapy was associated with significantly greater improvement in depressive symptoms (SMD -0.16, 95% CI -0.27 to -0.06; 16 RCTs; n=1,548 participants). There was no evidence of high heterogeneity (I2=28.0%).
Subgroup analyses: For trials of participants with depressive disorders, no significant differences were found in trials using other psychotherapy as a control group. For both trials of participants with depressive disorders and depressive symptoms, only individual treatment delivery had significant effects for cognitive-behavioural therapy.