Environmental and dietary factors associated with urinary OH-PAHs in mid-pregnancy in a large multi-site study

Background

PAH exposure is associated with adverse health outcomes, but exposure sources in pregnancy are not well-understood.

Objectives

We examined associations between urinary OH-PAHs during pregnancy and environmental tobacco smoke (ETS) and short-term ambient air pollution exposure. Participants included 1603 pregnant non-smokers in three cohorts from 7 sites across the USA. We also examined associations with intake of foods typically high in PAHs in one cohort with dietary assessment data (n = 801).

Methods

Urinary OH-PAHs were measured using LC-MS/MS; urinary cotinine was measured using SPE/UPLC-MS/MS. To accommodate different detection limits by cohort, ETS exposure was represented by modified cotinine quartiles; these combined concentrations below the highest detection limit in the first category (0–0.017 ng/mL), with the rest divided evenly into three categories (0.0171–0.2 ng/mL, 0.21–1.191 ng/mL, 1.192–1465 ng/mL). Air pollution exposure was represented by quartiles of same-day ambient PM_2.5 in residential census tracts estimated from EPA's Downscaler Model. We fitted separate Tobit regression models for log-OH-PAH concentrations in association with cotinine or ambient PM_2.5 quartile adjusted for specific gravity, site, batch, household income, education, employment status, neighborhood deprivation index, season, and year. For the food model, PAH dietary intakes were estimated using food frequency questionnaire data and standard portion weights from a national database.

Results

In adjusted models, the highest modified cotinine quartile vs. the lowest was associated with 48% (95% CI: 13%, 94%) higher urinary 1-hydroxynaphthalene, 36% (15%, 61%) higher 2-hydroxynaphthalene, 41% (23%, 63%) higher 3-hydroxyphenanthrene, and 70% (28%, 127%) higher 1-hydroxypyrene. Second and third quartile cotinine concentrations were associated with higher OH-PAHs, although not consistently. Same-day ambient PM_2.5 was not associated with any OH-PAH, nor was self-reported dietary intake.

Conclusions

ETS is a major source of PAH exposure for pregnant people in the USA while ambient PM_2.5 and diet measured via usual intakes appear less influential. Our findings underscore the importance of policies/actions to reduce environmental tobacco smoke exposure among pregnant people.