A meta-analysis of exercise interventions among people treated for cancer

This review applied meta-analytic procedures to integrate primary research findings that tested exercise interventions among people treated for cancer. Extensive literature searching strategies located published and unpublished intervention studies that tested exercise interventions with at least five participants (k = 30). Primary study results were coded. Meta-analytic procedures were conducted. The overall weighted mean effect size (ES) for two-group comparisons was 0.52 (higher mean for treatment than control) for physical function, 0.35 for symptoms other than fatigue, and 0.27 for body composition. More modest positive ESs were documented for mood (0.19), quality of life (0.14), fatigue (0.11), and exercise behavior (0.04). ESs were larger among single-group pre-post design studies. ESs among control group participants were typically negative and not (statistically) significantly different from 0. Exercise interventions resulted in small positive effects on health and well-being outcomes among existing studies. Future research should examine intervention-specific characteristics that result in optimal results, such as dose.