Effects of alcohol and tobacco on aerodigestive cancer risks: a meta-regression analysis

Objective: Meta-analysis was used to summarize the published evidence on the associations between alcohol and tobacco consumption and cancers of the oropharynx, pharynx, larynx, and esophagus. The objective was to produce summary risk estimates with uniform methods and on uniform exposure scales so that the magnitudes of the risks could be compared across tumor sites. Methods: Epidemiologic studies that estimated the effects of alcohol and tobacco consumption on the risk of cancers of the upper aero-digestive tract were identified from the MEDLINE database, 1966-2001. Alcohol and tobacco data were converted into common units (grams/week). For all studies meeting eligibility criteria, effect parameters (slopes) were estimated for both exposures. The exposure-risk slopes for each study were combined, site by site, using random effects meta-regression methods. Results: Fourteen studies met the final selection criteria. The carcinogenic effects of alcohol and tobacco were found to be multiplicative on the relative risk scale. Tobacco appeared to have a much stronger effect on the larynx than on any of the other aerodigestive sites, while alcohol's effect was strongest on the pharynx. The weakest association was that of alcohol and adenocarcinoma of the esophagus - an order of magnitude weaker than that for tobacco and laryngeal cancer. Conclusions: Meta-analysis was used to combine the results from all available studies, providing a comprehensive summary of the combined effects of alcohol and tobacco on the upper aerodigestive cancers.