[Epidemiological prospective studies on physical activities and the risk of colon cancer: a meta-analysis].

OBJECTIVE: To explore the relationship between physical activity (PA) and the risk of colon cancer. METHODS: Cohort studies on physical activity and risk of colon cancer were identified by searching MEDLINE, EMBASE, Chinese Bio-medicine and Chinese Wanfang databases from January 1979 to December 2009. Results from the individual studies were synthetically combined in our study. Inverse variance weighting was used in fixed effects model and the random effects estimate was based on the DerSimonian-Laird method. Variance-weighted least squares method was used for trend test of summarized dose-response data. RESULTS: A total of 28 studies were included in our analysis. An inverse association between physical activities and the risk of colon cancer was observed with the relative risks (RR) as 0.75 [95% confidence interval (CI): 0.66 - 0.86] in males and 0.85 (95%CI: 0.76 - 0.95) in females, respectively. However, the findings from those documents with high quality showed significant and borderline significant associations between PA and colon cancer in both males (RR = 0.74, 95%CI: 0.61 - 0.90) and females (RR = 0.99, 95%CI: 0.95 - 1.02). Meanwhile, the dose-response trend was not observed either in males (P = 0.142) or in females (P = 0.417). For men, the pooled RRs differed by subsites were 0.62 (95%CI: 0.45 - 0.85) and 0.74 (95%CI: 0.56 - 0.99) for highest level PA, compared with lowest level PA in proximal colon and distal colon cancer, respectively. For women, the pooled RRs were 0.84 (95%CI: 0.69 - 1.01) in proximal colon and 0.75 (95%CI: 0.53 - 1.05) in distal colon cancer, respectively. CONCLUSION: These results added to the evidence for the protective effects in colon cancer among men and women.