The effectiveness of exercise as an intervention in the management of depression: systematic review and meta-regression analysis of randomised controlled trials

Professor Biddle points out 'that even though the research designs
included in the analysis were weak, the effect sizes were large.' This
statement is consistent with our conclusion that there is no evidence of
exercise being clinically effective as a treatment for depression. The
characteristics of the trials that identify them as being of poor quality
- lack of concealment, analysis that did not use intention to treat and
lack of blinding - have each been shown to exaggerate results by 20-
40%[1]. 'Established criteria in epidemiological research' are appropriate
for investigating aetiological pathways but this is not the way to
investigate clinical effectiveness. Our meta-analysis clearly shows that
there is currently no evidence to support exercise as a clinically
effective intervention for depression. This is not the same as evidence of
no effect. Though difficult, it would not be impossible to design an
appropriate high quality randomised controlled trial of clinically
diagnosed depressed patients. The outcome should be a dichotomous clinical
diagnosis, assessed by someone blind to treatment allocation, and follow-
up of at least 12 months should be undertaken. A third arm in the trial of
'social contact' and a qualitative element would also be valuable in
trying to determine whether any effect that might be shown was due to the
exercise per se or social contact.

1. Schulz KF, Chalmers I, Hayes RJ, Altman DG. Empirical Evidence of
Bias. Dimensions of Methodological Quality Associated with Estimates of
Treatment Effects in Controlled Trials. JAMA 95;273:408-412.