Epidemiologic studies of diet and cancer.

Prospective cohort studies conducted within the U.S., including older cohorts established to investigate causes of cardiovascular disease as well as newer cohorts designed to address hypotheses related cancer risk, have already contributed to our understanding of the relation of diet and cancer. During the coming years, however, a much greater wealth of data will emerge from the cohorts that have been initiated within the past decade. The principle strength of these studies is that they will provide an assessment of diet unbiased by the diagnosis of cancer. In addition, the cohort studies provide an assessment of diet closer to the actual onset of cancer. Importantly, validation studies have been conducted of the dietary questionnaires used in several of these studies, thus allowing a quantitative evaluation of the influence of measurement error on the observed measures of association and their confidence intervals.