That Confounded P-Value

We evaluated the association between chlorination byproducts and colon and rectal cancer risk in a population-based case-control study conducted in Iowa in 1986-1989. Data were gathered from 685 colon cancer cases, 655 rectal cancer cases, and 2,434 controls. We calculated odds ratios for the 560 colon cancer cases, 537 rectal cancer cases, and 1,983 controls for whom water exposure information was available for at least 70% of their lifetime. We estimated exposure to chlorination byproducts with two types of measures: duration of lifetime at residences served by chlorinated water and estimated lifetime trihalomethane exposure. For rectal cancer, we observed an association with duration of chlorinated surface water use, with adjusted odds ratios of 1.1, 1.6, 1.6, and 2.6 for 1-19, 20-39, 40-59, and > or =60 years of exposure, compared with no exposure. Rectal cancer risk was also associated with several different measures of estimated lifetime trihalomethane exposure. For colon cancer and subsites, we detected no important increase in risk associated with duration of chlorinated surface water, nor with trihalomethane estimates. When we evaluated chlorination byproducts jointly with other factors, we found larger relative risk estimates for rectal cancer among subjects with low dietary fiber intake. The risk related to > or =40 years of exposure to a chlorinated surface water source was 2.4 (95% confidence interval = 1.5-4.0) for persons with low fiber intake and 0.9 (95% confidence interval = 0.4-1.8) for persons with high fiber intake, relative to the risk of persons with high-fiber diets and no exposure to chlorinated surface water. We observed a similar risk differential for low and high levels of physical activity.