Tobacco Use and Prevalence of Human Papillomavirus in Self-Collected Cervicovaginal Swabs Between 2009 and 2014.

Objective: To evaluate relationships between an objective biomarker of current tobacco exposure and high-risk genital human papillomavirus (HPV) prevalence among adult women in the United States.

Methods: We performed a retrospective analysis of adult women (aged 18-59 years) using three consecutive 2-year cycles (2009-2014) from the cross-sectional National Health and Nutrition Examination Surveys. Women who provided self-collected cervicovaginal swabs and serum were included. Human papillomavirus genotyping was conducted on cervicovaginal samples with a Linear Array HPV assay. Cotinine, a major metabolite of nicotine, was assayed from serum to provide a biomarker of recent tobacco exposure. Participants were stratified into three levels of tobacco exposure (nonsmokers, secondhand smoke exposure, and smokers) based on serum cotinine concentration levels using previously published ethnic-specific cut points. Weighted percentages are provided to account for unequal selection probabilities among participants and adjustments for nonresponse.

Results: Among the 5,158 women analyzed, 2,778 were classified as nonsmokers (57.1%, 95% CI 54.5-59.6%), 1,109 classified as having secondhand smoke exposure (18.4%, 95% CI 16.5-20.3%), and 1,271 classified as smokers (24.6%, 95% CI 22.8-26.5%) using serum cotinine concentration levels. Prevalence of HPV infection differed between nicotine exposure groups (P<.001): 441 smokers (32.1%, 95% CI 29.6-34.7%), 322 women with secondhand smoke exposure (26.1%, 95% CI 22.7-29.7%), and 451 nonsmokers (15.1%, 95% CI 13.3-17.1%) had a high-risk genital HPV infection. Controlling for demographics and number of lifetime sexual partners, the risks compared with nonsmokers for infection with a high-risk HPV genotype for smokers (adjusted odds ratio [OR] 1.7, 95% CI 1.4-22) and secondhand smokers (adjusted OR 1.4, 95% CI 1.1-1.8) are similarly increased (P<.001).

Conclusion: In this large cross-sectional, population-based study, we show a relationship between an objective biomarker of current tobacco use and genital HPV infection. Cigarette smoking and exposure to secondhand smoke are associated with increased odds of infection with high-risk genital HPV independent of lifetime number of sexual partners.