850

BACKGROUND. The hypothesis that dietary fat is associated with colorectal cancer originated from international comparisons of ecologic data. Experimental data, as well as prospective epidemiologic data among U.S. women support a role specifically for saturated fats in colorectal carcinogenesis. In Singapore, colorectal cancer rates in both men and women have more than doubled between 1970 and 2000. A greater influence of Western diet and lifestyle factors have been observed among Singapore Chinese and may contribute to the population’s rise in colorectal cancer incidence. METHODS. We examined the relationship between individual fatty acids and colorectal cancer in a population-based cohort of 63,257 Singapore Chinese men and women that was established between 1993 and 1998. As of December 31, 2005, 955 incident colorectal cancer cases had been diagnosed. Diet was assessed at baseline with a validated 165-item quantitative food frequency questionnaire. We used Cox regression methods with adjustment for age at baseline, interview year, dialect group, education level, smoking, alcohol use, body mass index, physical activity and family history of colorectal cancer. RESULTS. All relative risk (RR) values presented are for highest versus lowest quartiles of intake. We report no overall association for total fat [RR=1.00, 95% confidence interval (CI)=0.83-1.20) or saturated fat intake (RR=0.97, 95% CI=0.80-1.17) and colorectal cancer. However, among women, there was a statistically significant, dose dependent, positive association for localized disease (Duke A or B) that was stronger in the colon subsite [(RR=2.43, 95% CI=1.40-4.22, p for trend=0.005, for total fat) and (RR=1.78, 95%CI=1.04-3.02, p trend=0.027 for saturated fat)]. No such associations were noted in men. Although total polyunsaturated fatty acids (PUFAs) were unrelated to colorectal cancer, n-3 PUFAs from fish showed a moderate, positive association with colorectal cancer. This association was most evident for advanced disease (Duke C or D) (RR=1.33, 95% CI=1.05-1.70, p trend=0.015), especially among participants with ≥median value intake of n-6 fats (RR=1.81, 95% CI=1.22-2.67, p trend=0.003). The association was somewhat stronger for the colon subsite. CONCLUSIONS. Our findings are consistent with those reported from a U.S. cohort, and suggest that total and saturated fat are acting as early stage risk factors for colorectal cancer. This is one of the first prospective studies of individual fatty acids, assessed using a validated food frequency questionnaire, in relation to colorectal cancer in a nonwhite population. In Singapore, cooking oils used for frying contribute most to fatty acid intake. Our unexpected risk enhancing association observed for n-3 PUFAs from fish suggests that alternative mechanisms may exist for these fatty acids and colorectal cancer that depend on the food source and/or cooking method.

98th AACR Annual Meeting-- Apr 14-18, 2007; Los Angeles, CA