Document Type



Public Health


Chronic inflammation has been implicated in the pathogenesis of many chronic diseases, including colorectal cancer. Obesity has been identified as a risk factor for colorectal cancer via induction of a chronic state of low-grade bowel inflammation due to excessive release of inflammatory cytokines such as C-Reactive protein (CRP) by adipocytes. Prior studies have produced evidence of positive associations between CRP and colon cancer risk but none to date have assessed if elevated CRP is an independent risk factor when controlling for obesity. Also, controversy exists about which measure of adiposity best predicts risk of colon cancer. METHODS: We examined these questions using the National Health and Nutrition Examination Survey III (NHANES III) database linked to the National Death Index. In addition to CRP level, we studied the following four adiposity measures: body mass index, waist circumference, waist-to-hip ratio, and a new index developed by the NIH that stratifies BMI according to waist circumference. Outcome variables were: all-cause, colorectal cancer, other obesity related cancers, and all other causes. Dichotomous and Polytomous Logistic Regression were performed. RESULTS: CRP levels showed positive but weak correlations with adiposity measures. In the age-adjusted and multivariate Dichotomous Logistic Regression analyses, elevated CRP level was significantly associated with all-cause mortality (OR=1.63 95% CI 1.40-1.91; and OR=1.32, 95% CI 1.08-1.61; respectively.) Using age-adjusted and multivariate Polytomous Logistic Regression, elevated CRP level was associated with mortality from colorectal cancer (OR=2.72, 95% CI 1.30-5.72; OR= 2.44, 95% CI 1.20-4.94; respectively.) OR estimates did not change appreciably when adiposity measures were included in multivariate models. Further, none of the body measures of adiposity were significantly associated with cause-specific death in either age-adjusted or multivariate analyses. CONCLUSIONS: Our findings suggest that CRP may be an independent risk factor for all-cause and, more so, for colorectal cancer mortality. We speculate that the surprising effects related to body measures of adiposity may reflect misclassification related to the single baseline obesity measurement, taken during 1988-94, in light of the obesity epidemic that emerged during the follow-up time period. Further investigation of this relationship is warranted to determine of elevated CRP remains an independent risk factor when longitudinal adiposity history is known.

Included in

Public Health Commons