Although an inverse association between physical activity and risk of colon cancer is well established, a formal estimate of the magnitude of this risk reduction that includes recent studies is not available. This analysis examines the association by sex and study design, restricting analyses to studies where data for colon cancer alone were available. The authors reviewed published studies through June 2008 examining the association between physical activity and risk of colon cancer. Heterogeneity and publication bias were evaluated and random effects models used to estimate relative risks (RR). Differences by sex and study design were evaluated. A total of 52 studies were included. An inverse association between physical activity and colon cancer was found with an overall relative risk (RR) of 0.76 (95% confidence interval ( Moreover, this association is plausibly supported by several biological mechanisms, including decreased inflammation, reduced intestinal transit time, decreased insulin-like growth factor levels, reduced hyperinsulinemia and modulated immune function with physical activity. Although several qualitative reviews of this literature have been conducted, an up-to-date meta-analysis of the studies of physical activity and colon cancer is not available (International Agency for Research on Cancer WHO, 2002;Slattery, 2004). One recent study quantitatively evaluated data on the relation between physical activity and risk of colon cancer, but also included colorectal cancer end points (Samad et al, 2005). No association has been consistently found for physical activity and rectal cancer, and the consensus is that one is unlikely to exist (International Agency for Research on Cancer WHO, 2002). Estimation of the risk reduction associated with physical activity separately for colon cancer is important for public health because the disease burden attributable to physical inactivity is likely underestimated by including rectal cancer end points. The World Health Organization conducted a meta-analysis using data from studies prior to 2000, and estimated that 16% of the global colon cancer disease burden is due to physical inactivity (Bull et al, 2004). However, this meta-analysis was conducted several years ago, and also did not investigate differences by study design. Thus, we conducted a meta-analysis to estimate the summary relative risk of colon cancer associated with physical activity, based on available studies to date.
MATERIALS AND METHODSA search was conducted on PubMed for all publications in English through June 2008. The search terms physical activity, exercise, and colon cancer were used. In addition, previous reviews of the data (Slattery, 2004;Samad et al, 2005;Lee and Oguma, 2006) and the reference lists of included studies were reviewed.We included only case -control or cohort studies with a colon cancer end point. Studies that reported the findings for rectal cancer or for colon and rectal cancers combined (colorectal cancer) were excluded for the reason detailed above. We did not limi...