Validity of cigarette smoking habits in three epidemiologic studies in Utah.

Utah has lower incidence and mortality for many smoking-related forms of cancer and heart disease. It is an important epidemiologic question to assess whether the population attributable risk associated with cigarette smoking in this low-risk population is biased from under-reporting because of societal pressures not to smoke. To answer this question, we compared reported cigarette use to serum cotinine values in three different epidemiologic study designs. Included in these analyses were data from men interviewed for a cross-sectional study of dietary intake and hormones, women interviewed as a part of a case-control study of cervical cancer, and men interviewed in conjunction with a cardiovascular disease and hypertension family follow-up study. Cross-sectional study participants reported accurate cigarette usage 93.8% of the time; case-control participants accurately reported cigarette use 98.5% of the time; participants interviewed in the family cohort study correctly reported usage 82.8% of the time. Most inaccurate reporting of smoking was by exsmokers being followed for a disease known to be linked to smoking. The low attributable risk of smoking related to diseases in Utah is not from underreporting of cigarette smoking, and makes Utah an ideal population to examine other risk factors for diseases where smoking increases risk.