Abstract
In this paper the effects of physical activity on depression and anxiety are examined. The scientific strength of studies of physical activity, depression, and anxiety are examined against the standards of science accepted in epidemiology, with a focus on the independence, consistency, dose–response gradient, and biological plausibility of the evidence. There are too few population-based studies of physical activity and mental health to draw conclusions. Nonetheless, the studies generally showed a moderate dose-independent reduction in the risk for depression among physically active people, with limited evidence for the independence and consistency of the associations observed. Experimental studies with small groups generally showed moderate reductions in depression, anxiety, and some related physiological signs. However, published studies usually had weak experimental control, sampling people who did not have clinical conditions. Hence, the internal validity and generalizability of the experimental studies are weak. A major shortcoming of the physical activity literature reviewed is the absence of evidence about plausible, explanatory mechanisms that are coherent with biological or cognitive theory about the etiology and treatment of depression and anxiety.