To evaluate the strength of the evidence provided by the epidemiological literature on the association between alcohol consumption and the risk of 18 neoplasms, we performed a search of the epidemiological literature from 1966 to 2000 using several bibliographic databases. Meta-regression models were fitted considering linear and non-linear effects of alcohol intake. The effects of characteristics of the studies, of selected covariates (tobacco) and of the gender of individuals included in the studies, were also investigated as putative sources of heterogeneity of the estimates. A total of 235 studies including over 117 000 cases were considered. Strong trends in risk were observed for cancers of the oral cavity and pharynx, oesophagus and larynx. Less strong direct relations were observed for cancers of the stomach, colon and rectum, liver, breast and ovary. For all these diseases, significant increased risks were found also for ethanol intake of 25 g per day. No significant nor consistent relation was observed for cancers of the pancreas, lung, prostate or bladder. Allowance for tobacco appreciably modified the relations with laryngeal, lung and bladder cancers, but not those with oral, oesophageal or colorectal cancers. This meta-analysis showed no evidence of a threshold effect for most alcohol-related neoplasms. The inference is limited by absence of distinction between lifelong abstainers and former drinkers in several studies, and the possible selective inclusion of relevant sites only in cohort studies.

(c) 2001 Cancer Research Campaign