The review included 35 studies (n=2,095 participants, range nine to 439).
The mean weighted least square effect size for depression measures (35 studies) was 0.34 (standard deviation 0.40; p<0.01) and the mean unweighted least square effect size for depression measures (35 studies) was 0.40 (p<0.01).
Investigating the maintenance of treatment gains over life revealed a significant negative correlation (p=0.03) between follow-up time lag and follow-up effect size, with follow-up assessments near the end of treatment exhibiting relatively large effects (for example, two to three months), whereas follow-ups with lags of one year or more did not demonstrate treatment effect. There were significantly higher effect size values for the non-depression outcome studies (p=0.01) when controlling for factors that might influence effect size and type of control group.
Comparing psychotherapy with passive control groups (20 studies) yielded a mean psychotherapy effect size of 0.41 (p<0.01), whereas comparison with active control groups (15 studies) produced a mean effect size of 0.24 (p=0.03). Studies that used active control groups included clinically referred youths more often than studies that used passive control groups (p=0.002).
The 31 treatments that emphasise changing cognitions were compared with the 13 that did not, with a mean effect size for cognitive treatments 0.35 (p<0.01) and a mean effect size for non-cognitive treatments 0.47 (p<0.01); the effect size difference between the two was not significant. The mean effect size for studies that used primarily recruited participants (n=29 participants) was 0.34 (p<0.01) and that for studies that used clinically referred participants (n=six participants) was 0.32 (p<0.05).