Re: Multicenter Case–Control Study of Exposure to Environmental Tobacco Smoke

Boffetta et al. (1) in their multicenter study on exposure to environmental to-bacco smoke (ETS) and the risk of lung cancer showed that the odds ratios for spousal and workplace ETS exposures were 1.16 and 1.17, respectively, with no clear dose–response relationships. Although they allowed for intake of fruits and vegetables, they did not allow for intake of saturated and unsaturated fats. A recent meta-analysis of 51 studies showed that smokers have a statistically significantly (Pł.0001) higher intake of saturated fat than nonsmokers (2). ETS-exposed nonsmokers have a statistically significantly (Pł.001) higher intake of fat compared with unexposed nonsmok-ers (3). Many studies have shown an in-creased risk of lung cancer with increas-ing consumption of fat, for example Wynder et al. (4). Alavanja et al. (5), in a case–control study of 429 nonsmoking women with lung cancer (comparable in size to the study of Boffetta et al.), showed that the relative risk (RR) for all histologic types of lung cancer compar-ing high quintiles with low quintiles of saturated fat intake was 6.14 (95 % con-fidence interval [CI] 4 2.63–14.4; P for trend 4.0001). The RR for adenocar-cinoma comparing high quintiles with low quintiles of saturated fat intake was 11.38 (95 % CI 4 3.47–34.4; P for trend 4.0001). In the study by Boffetta et al., 78 % of the case patients were female and 51% of the tumor histologic types were ad-enocarcinoma. In light of the evidence given by Alavanja et al., how can Bof-fetta et al. attribute this very small in-creased risk to ETS exposure