Association between secondhand smoke exposure and diabetes mellitus in 131 724 Korean never smokers using self‐reported questionnaires and cotinine levels: Gender differences

Abstract

The aim of this study is to compare the gender-different associations between secondhand smoke (SHS) exposure and diabetes in self-reported never smokers verified by cotinine. Self-reported never smokers verified by cotinine levels (<50 ng/mL) enrolled in the Kangbuk Samsung Health Study between April 2011 and December 2016 were included for this study. SHS exposure was defined as current exposure to passive smoke indoors at home or workplace. The gender-specific association between SHS exposure and diabetes was assessed using multivariable regression analyses. Of the total 131 724 individuals (mean age 35 years, SD 7.1 years), 66.8% were female, and the prevalence of SHS exposure in the entire population was 22.9%. The prevalence of diabetes in the group exposed to SHS was higher than that in the group unexposed to SHS only in females (1.8% vs 1.2%, P < .001 for females; 2.2% vs 2.2%, P = .956 for males). A significant gender interaction existed for the relationships between SHS exposure and diabetes (P for interaction <.001). The multivariate regression model showed that SHS exposure was significantly associated with diabetes only in females (odds ratio [95% CI], 1.40 [1.20, 1.65] for females; 1.00 [0.85, 1.19] for males). In particular, females with SHS exposure ≥1 hour/day, ≥3 times/week, and ≥10 years showed an increased risk of diabetes by more than 50% compared to those without SHS exposure. SHS exposure was significantly associated with diabetes in female never smokers with dose-dependent relationship. However, further longitudinal studies are needed to elucidate the gender difference in the incidence of diabetes associated with SHS exposure.