The U.S. Food and Drug Administration (FDA) is currently considering whether to ban the use of menthol as an additive to cigarettes. Mentholated cigarettes account for roughly 30 percent of cigarettes sold in the United States, and they are favored by African American smokers by roughly a threefold margin compared to white smokers. They are also favored by women smokers. A major aspect of the FDA charge is to determine whether use of mentholated cigarettes is more harmful than use of nonmentholated cigarettes.
Menthol became a focus of epidemiologists and tobacco-control scientists following the observation by a colleague of mine and me in a 1988 letter to the American Journal of Public Health pointing out that death rates for esophageal cancer among non-Whites (mainly African-Americans) had risen strikingly in parallel with the increase in menthol’s market share and that African American smokers tended to favor mentholated cigarettes.
Many studies since then have addressed whether smoking mentholated cigarettes is associated with increased risk, and these studies have consistently failed to show that menthol has any effect beyond that of smoking per se. However, most studies focused on lung cancer—the most common smoking-related cancer—and few examined the association with esophageal cancer or cancers of the mouth and larynx.
Because our ecologic observation from 24 years ago initially drew attention to menthol, we decided to examine ecologically the changes in the rates of four smoking-related cancers by race and sex from 1973 to 2007 in relation to the prevalence of smoking of mentholated cigarettes. The cancers were those of the lung, throat, mouth and larynx. Information on cancer incidence came from the National Cancer Institute’s SEER program, and information on the prevalence of use of mentholated cigarettes came from representative surveys of the U.S. population.
Ecologic studies of this sort have well-known limitations: above all, they make use of aggregate information on the exposure groups and rates of disease rather than information on individuals. Second, in an ecologic analysis, one cannot control for confounding factors as one can in an analytic epidemiologic study.
However, because there is such a strong contrast in the preference for menthol by race and sex and because there are large differences in the rates of these cancers also by race and sex, we felt that this was a case in which an ecologic approach could be informative.
In spite of dramatic differences in the rates of the four cancers by race and sex, as well as dramatic changes over the 35-year time period, there was little evidence of a correlation between prevalence of mentholated cigarette use and cancer rates. Thus, our study indicates that smoking mentholated cigarettes does not appear to influence risk over and above the effect of smoking per se.
The takeaway for me as a researcher is—as stated in our new paper—that we cannot explain the very large differences in the rates of esophageal cancer between African Americans and whites. For smokers, the bottom line is that smoking is smoking, and menthol is probably unimportant.