Risk factors for esophageal cancer in lifelong nonsmokers.

Information on the etiology of esophageal cancer in lifelong nonsmokers is of interest to understand and quantify risk factors for the disease in the absence of the residual confounding by tobacco. Of a total of 316 cases with histologically confirmed incident cancers of the esophagus, 46 (17 males and 29 females) who described themselves as lifelong nonsmokers were selected to assess esophageal cancer risk in the absence of potential confounding and interactive effects of smoking. These patients were compared to 230 lifelong nonsmoker controls (85 males and 145 females) admitted to hospital for acute, nonneoplastic, non-alcohol-related conditions. The major risk factor for cancer of the esophagus in lifelong nonsmokers was elevated alcohol consumption: compared to drinkers of fewer than 4 drinks per day the relative risk (RR) was 2.7 (95% confidence interval, 1.1-6.8) for 4 to fewer than 8 drinks, and 5.4 (95% confidence interval, 1.4-21.0) for 8 or more drinks, with a significant trend in risk. Among selected indicator foods considered, significant protective effects were observed for fish (RR = 0.5 for the highest consumption tertile), green vegetables (RR = 0.6), and fresh fruit intake (RR = 0.3). Consequently, there was a significant inverse relationship with an estimate of beta-carotene intake (RR = 0.5 and 0.4, respectively, for the middle and highest tertiles of intake versus the lowest), with a significant trend in risk. The estimated RR for the highest alcohol consumption and lowest beta-carotene intake category was 8.6, and these two factors together explained over 45% of cases. Gastrectomy and family history of cancer of the esophagus were also associated with increased risk (RR = 4.6 and 4.3, respectively).