for research on Cancer, 2002). Over time, the designs of these studies have become progressively more sophisticated in terms of the measurement of physical activity and accounting for potential confounding factors. The studies summarized below were able to remove people with preexisting disease from the analysis and then examine the effect of physical activity over time in healthy populations.
Among both men and women, high levels of physical activity may decrease the risk of colon cancer by as much as 50 percent (Colditz et al., 1997; Marrett et al., 2000). Although studies have not consistently used a standard measure of activity or similarly defined a “high” level of activity, a dose-response relationship between physical activity and colon cancer has consistently been observed across studies with a variety of designs and in many different populations (Garabrant et al., 1984; Slattery et al., 1988; Gerhardsson de Verdier et al., 1990; Whittemore et al., 1990; Giovannucci et al., 1995a; Martinez et al., 1997). In addition, physical activity appears to lower the risk of large adenomatous polyps, which suggests that it may act early in the adenoma-carcinoma sequence (Giovannucci et al., 1995a; Giovannucci et al., 1996; Kahn et al., 1998). The relation is seen across levels of obesity, suggesting a benefit of physical activity for reduction in cancer incidence in addition to weight reduction.
Maintaining high levels of physical activity throughout life appears to impart the greatest protection (Lee et al., 1989; Kune et al., 1990). In a study of Harvard alumni, men who were at least moderately active at two assessments were 48 percent less likely to develop colon cancer than men who were inactive at both assessments (Lee et al., 1991). However, this does not mean that those who have been sedentary in the past cannot reap the benefits if they become active. The same study showed that among men who were sedentary at the initial assessment, those who increased their activity during the 11- to 15-year follow-up period were 13 percent less likely to develop colon cancer than those who remained sedentary during the same time period (Lee et al., 1991). Data from at least two prospective studies indicated that both men and women can lower their risk of colon cancer simply by engaging in moderate physical activity such as brisk walking or stair climbing for an hour a day (Giovannucci et al., 1995a; Martinez et al., 1997). In those studies the risk decreased with increasing walking pace among those whose only recreational activity was walking.
Several mechanisms have been proposed to explain this. Physical activity may decrease gastrointestinal transit time, thereby minimizing contact of the gastrointestinal tract with potential carcinogens in the stool (McTiernan et al., 1998), and it may also reduce circulating levels of insulin, a growth factor for colonic epithelial cells (McKeown-Eyssen, 1994; Gio-