Use of electronic nicotine delivery systems by pregnant women II: Hair biomarkers for exposures to nicotine and tobacco-specific nitrosamines.

INTRODUCTION Public awareness of electronic nicotine delivery systems (ENDS) has increased over time, and the perception that ENDS offer a safer alternative to cigarettes may lead some pregnant women to use them to reduce cigarette smoking during pregnancy. No previous studies have used metabolite levels in hair to measure nicotine exposure for ENDS users during pregnancy. We aimed to measure and compare levels of nicotine, cotinine, and tobacco-specific nitrosamines (TSNAs) in hair samples from pregnant women who were current ENDS users, current smokers, and current non-smokers. We also aimed to estimate the association between ENDS use/smoking and smallness for gestational age (SGA). METHODS We used hair specimens from pregnant women who were dual users (ENDS and cigarettes), smokers, and non-smokers from a prospective cohort study to estimate exposure to nicotine, cotinine, and TSNAs. The exposure biomarkers and self-reports of smoking and ENDS use were used in log-binomial regression models to estimate risk ratios (RRs) for SGA among offspring. RESULTS Nicotine concentrations for pregnant dual users were not significantly different from those for smokers (11.0 and 10.6 ng/mg hair, respectively; p= 0.58). Similarly, levels of cotinine, and TSNAs for pregnant dual users were not lower than those for smokers. The RR for SGA was similar for dual users and smokers relative to non-smokers, (RR= 3.5, 95% CI: 0.8-14.8) and (RR= 3.3, 95% CI: 0.911.6), respectively. Using self-reports confirmed by hair nicotine, the RR values for dual ENDS users and smokers were 8.3 (95% CI: 1.0-69.1) and 7.3 (95% CI: 1.0-59.0), respectively. CONCLUSIONS We did not observe lower levels of nicotine, cotinine, and TSNAs for current dual users compared to smokers during pregnancy. The risk of SGA for offspring of pregnant dual users was similar to that for offspring of pregnant smokers. Future studies are needed to further estimate the magnitude of the association between ENDS use and smallness for gestational age.