Consistency of perception of tobacco use: Evidence from a biomarker

Public awareness of the adverse effects of tobacco use, tobacco regulations, and alternative forms of tobacco consumption have all increased over time. While these advances may influence perception of tobacco consumption, its temporal consistency remains unclear. We define this temporal consistency of tobacco use perception as the stability of biomarker thresholds for self-identification as a tobacco user. We use data from ten waves of the United States National Health and Nutrition Examination Survey (NHANES), covering 1999-2000 through 2017-2018 (N = 36,624). We compare survey responses regarding tobacco consumption with cotinine, a biomarker of nicotine exposure, to document the evolution of cotinine levels at which an individual self-identifies as a tobacco user over time. The data-driven threshold for classifying tobacco consumers based on cotinine levels fluctuates significantly. We observe an increasing gap between self-reported and cotinine-based measures of tobacco use over time. Significant correlations between socioeconomic covariates and measurement errors from inaccuracies in self-reported tobacco consumption measure may suggest potential biases when relying solely on self-reported tobacco use measure.