Background: In the period 1990 to 2001, a case-control study on oral cancer and maté consumption was conducted at the Cancer Institute of Uruguay.
Methods: The study included 696 newly diagnosed cases with squamous cell carcinoma and 696 controls afflicted with nonneoplastic conditions not related to tobacco smoking and alcohol drinking. The participants were matched on age and residence and the study was restricted to men.
Results: In order to control confounding for tobacco and alcohol, we fitted 2 models. According to model 1, the odds ratio (OR) for maté consumption was 1.15 (95% confidence interval [CI], 0.76-1.73), whereas the results for model 2 showed an OR of 3.47 (95% CI, 1.60-7.52).
Conclusions: The inclusion of a term for the interaction between maté and smoking (or drinking) was rewarding and the ORs were even higher than those observed with the crude estimates.
Keywords: benzo[a]pyrene; confounding; interaction; maté consumption; oral cancer.
Copyright © 2012 Wiley Periodicals, Inc.