Diet and colon cancer: integration of the descriptive, analytic, and metabolic epidemiology.

  • 1 December 1985
    • journal article
    • research article
    • Vol. 69, 223-8
Abstract
Case-control studies of diet and colon cancer have produced inconsistent findings. Data from descriptive and metabolic epidemiologic studies suggest that colon carcinogenesis varies by sex-age status, particularly when considered in relation to anatomic subsite. Thus stratification of data analysis by these criteria may elucidate dietary etiology. Consideration should also be given to the limited and nonlinear variation in colon cancer risk likely to exist in relation to the range of dietary "exposure" within Western study populations. In our community-based case-control study of colon cancer, comprising 220 patients and 440 individually matched controls, the dietary data have been analyzed by both tertile and quintile of individual consumption level. For total dietary energy intake, the tertile approach indicates a weak (men) to moderate (women) dose-response relationship; however, the quintile approach revealed in women a fourfold step-up in risk between the lowest and next-to-lowest quintile. For dietary protein, the quintile approach doubled the step-up in risk between the 2 lowest consuming categories for men compared with the tertile approach. Applying this quintile approach to data analyses stratified by sex, age, and colon subsite, we found: 1) In women, the risk of colon cancer for the upper four quintiles of consumption of calories and other major nutrients was greatest at young ages, whereas for men it was greatest at older ages. 2) The diet-associated risks tend to be maximal in the left (distal) colon for men and in the right (proximal) colon for women.(ABSTRACT TRUNCATED AT 250 WORDS)

This publication has 0 references indexed in Scilit: