Alcohol and beer consumption in relation to cancers of bowel and lung: an extended correlation analysis.

Some extensions to simple international correlation analyses are examined with reference to beer consumption and bowel and lung cancers in 29 countries. Simple correlation of cross-sectional data for beer consumption and rectal cancer yields a coefficient of 0.77 in males and 0.75 in females. Lagged correlation analyses show a different temporal relationship between beer and colon cancer (little change over time) and rectal cancer (maximal correlation for contemporaneous, i.e. cross-sectional, data sets). Lagged correlation analysis of beer consumption and lung cancer shows a similar pattern to that for rectal cancer and beer, and markedly different from that for lung cancer and cigarettes, the former showing maximal correlation for contemporaneous data, the latter peaking 15 yr prior to date of mortality. Changes in beer consumption over time correlate with subsequent changes in rectal cancer, particularly amongst younger age groups. The sex-ratio of rectal cancer varies from around 1 in low beer consumption countries to about 1.75 in high consumption countries. The utility of these extensions to the usual simple correlation analyses is considered. The biologic plausibility of beer-bowel cancer relationships, and possible mechanisms, are discussed in the light of the findings.