Physical activity and risk of colon adenoma: a meta-analysis

Abstract

Background: Little evidence is available on the relation of physical activity with colon adenomas, a colon cancer precursor.

Methods: We conducted a systematic literature review and meta-analysis of published studies (in English) through April 2010, examining physical activity or exercise and risk or prevalence of colon adenoma or polyp. Random effects models were used to estimate relative risks (RRs) and corresponding confidence intervals (CIs). A total of 20 studies were identified that examined the association and provided RRs and corresponding 95% CIs.

Results: A significant inverse association between physical activity and colon adenomas was found with an overall RR of 0.84 (CI: 0.77-0.92). The association was similar in men (RR=0.81, CI: 0.67-0.98) and women (RR=0.87, CI: 0.74-1.02). The association appeared slightly stronger in large/advanced polyps (RR=0.70, CI: 0.56-0.88).

Conclusion: This study confirms previous reports of a significant inverse association of physical activity and colon adenoma, and suggests that physical activity can have an important role in colon cancer prevention.

Figure 1

Meta-analysis of physical activity and colon adenoma. Study physical activity level comparisons are as follows: Kono et al, 1991: ⩾120 vs 0 min per week; Little et al, 1993: ⩾30 min vs none; Shinchi et al, 1994: daily vs none; Giovannucci et al, 1995: highest vs lowest quintile; Sandler, 1995: highest vs lowest quartile; Giovannucci et al, 1996: highest vs lowest quintile; Neugut et al, 1996: any vs none; Lubin et al, 1997: >5.5 h per day vs <4 h per day; Enger et al, 1997: highest vs lowest quartile; Kahn et al, 1998: high vs low; Kono et al, 1999: ⩾36 MET h per wk vs <4 MET h per wk; Boutron-Ruault et al, 2001: high vs low; Colbert et al, 2002: high vs low quartile; Lieberman et al, 2003: per 5 unit change in physical activity index; Tiemersma et al, 2003: not specified; Hauret et al, 2004: >40 MET h per wk vs <17.1 MET h per wk; Wallace et al, 2005: high vs low tertile; Larsen et al, 2006: high vs low quartile; Rosenberg et al, 2006: ⩾40 MET h per wk vs none; Hermann et al, 2009: active vs inactive. ES=effect size; MET=metabolic equivalent.