Explanation of the design of the total exposure study

Abstract

This letter is in response to the comments by Lee et al. regarding the design of the study reported in this journal by Roethig et al. (2009). The Total Exposure Study (TES) being commented upon was a carefully designed study, particularly to obtain reasonable population estimates of the biomarkers of exposure (BOE) to smoke constituents. The study was a stratified, cross-sectional multicenter study design (39 sites in 31 states) that included 3,585 adult cigarette smokers and 1,077 nonsmokers. Population estimates for the BOE were generated by weighting sample data with weights from Behavioral Risk Factor Surveillance System (BRFSS; Center for Disease Control and Prevention, 2000a, 2000b). BRFSS is the world’s largest ongoing telephone health survey system, tracking health conditions and risk behaviors in the United States yearly since 1984. The enrollment into the TES was stratified by age (three categories), gender (two categories), and Body Mass Index (BMI, two categories) for both smokers and nonsmokers and by tar delivery (four categories) for the smokers. Weighting of age, gender, BMI, and smoker status was based on the population proportions from BRFSS. The weighted sample mean and sample SE were calculated based on the approaches described by Lohr (1999).