Smoking, alcohol and thyroid cancer occurrence: systematic review and meta-analysis of case-control studies

English language case-control studies reporting on the association of thyroid cancer occurrence with smoking and alcohol drinking are summarized. Two independent researchers performed study selection and data extraction. Random effects model was applied and external adjustment was performed to control for important confounders. Twenty reports published between 1987 and 2007 were included in quantitative synthesis. For smoking, mean association was inverse (odds ratio [OR]: 0.785; 95% CI: 0.701-0.879) remaining after adjustment for alcohol drinking; heterogeneity was moderate. For alcohol drinking, mean association was inverse (OR: 0.795; 95% CI: 0.660-0.958) (remaining after adjustment for smoking, OR: 0.832; 95% CI: 0.688-1.007); heterogeneity was large becoming moderate after adjustment. Data from case-control studies identified showed inverse mean association between thyroid cancer occurrence and ever-smoking or ever-drinking alcohol.