Colorectal cancer and beer drinking.

Evidence is presented of a significant statistical association between beer drinking and colorectal cancer, particularly rectal cancer. This finding is based on correlations between consumption and cancer mortality and between changes in consumption and changes in cancer mortality for 47 states in the United States of America. Also various secular trends, an urban-rural gradient, socioeconomic gradients and sex ratios in the United States are shown to be generally consistent with a relationship between beer consumption and colorectal cancer, particularly rectal cancer. The limitations on drawing sound aetiological inferences from such data are acknowledged. In particular, several other variables are shown to be associated with both beer drinking and colorectal cancer. Also, a discussion of previous epidemiological studies is given, and it appears there is only a limited amount of direct evidence in humans to support the statistical demographic relationships.