Pattern of Prenatal Secondhand Smoke Exposure among Rural Women in Main Tobacco Production Region, China

Background: Prenatal secondhand smoke (SHS) exposure is the most modifiable risk factor associated with adverse birth outcomes. However, few researches were interested in finding the heterogeneity change tendency throughout pregnancy. The objectives of this study were to explore the heterogeneity by classifying subgroups based on the trajectories of urine cotinine (UC) throughout pregnancy in main tobacco production region, so as to provide individual intervention.Methods: This study used a prospective prenatal cohort consisting of 385 pregnant women who were UC validated as passive smoking at first (T1), second (T2) and third trimester (T3) time points. Urine cotinine were analyzed by high performance liquid chromatography-tandem mass spectrometry. The limit of detection for cotinine was 0.08ng/mL. Latent Class Growth Curve Modeling (LCGCM) was used to classify pregnant women into the unobserved subgroup (latent classes) with similar trajectories of UC.Results: The three-class patterns were determined prefer based on the information criteria. The general trends of UC in three patterns were upward with different trajectories from T1. The first class was steep increase at T2 and followed by sharp decrease at T3 (n=7, 1.8%); the second class was flat increase throughout pregnancy (n=98, 24.8%); the third class was moderately rising during pregnancy (n=290, 73.4%). The minority group and residence in tobacco production areas were positively associated with the UC increase trend.Conclusions: Patterns of prenatal SHS exposure among rural women generally go up with different trajectories during pregnancy. The lifestyle of minority and living environment of tobacco production may increase UC level. LCGCM may assist to better understanding the pattern of UC change over time. Identification of three subgroups and their distinctive patterns of change have the potential help for individual intervention.