Response error in self-reported current smoking frequency by black and white established smokers

Abstract

As compared with white smokers, black smokers, although they report using fewer cigarettes per day, are at higher risk for most smoking-related diseases. Among black smokers serum cotinine levels are also higher in proportion to cigarettes per day; this observation has led to suggestions of bias in self-reporting. The purpose of this study was to evaluate and compare the extent of errors in self-reported smoking patterns among black and white established smokers. Ninety-seven white and 66 black smokers participated in structured telephone interviews, filled out two self-administered questionnaires one week apart, and collected all of their cigarette butts for a week. Group differences in the validity of self-reported smoking patterns were assessed by comparison with cigarette butt counts and the measured butt lengths. Both black and white smokers significantly overestimated smoking on our measure of smoking frequency (both P < 0.001); the group difference in bias was not significant (P = 0.13). There was no evidence that underreporting was more common among blacks than among whites (P = 0.67). Test-retest reliability was not significantly different in the two groups (P = 0.09). Both groups performed poorly when asked to categorize their smoking frequency according to the cutpoints of the Fagerström Test for Nicotine Dependence. Black smokers smoked more of each cigarette and smoked longer cigarettes, but they smoked fewer total millimeters of cigarettes per day (all P < 0.001). Contrary to an earlier report, the disproportionately high cotinine levels could not be attributed to reporting error.