Association of Polycyclic Aromatic Hydrocarbon Mixtures with Gestational Iron Deficiency: A Cross-Sectional Analysis of the Zunyi Birth Cohort

Background: Maternal iron deficiency is a highly prevalent nutritional disorder during pregnancy and poses substantial risks to both maternal and fetal health. While adequate dietary iron intake and supplementation are critical for maintaining iron balance during pregnancy, emerging evidence suggests that exposure to environmental contaminants, such as polycyclic aromatic hydrocarbons (PAHs), may disrupt iron homeostasis and contribute to iron deficiency. However, epidemiological evidence linking prenatal PAH exposure to gestational iron deficiency remains limited, particularly in populations with low-level exposure. This study aimed to examine the associations of individual and combined PAH exposures during late pregnancy with the odds of iron deficiency. Methods: We conducted a cross-sectional analysis of 1717 pregnant women enrolled in the multicenter Zunyi Birth Cohort and measured ten urinary PAH metabolites. Single-pollutant associations were estimated with weighted quasi-binomial logistic regression using inverse probability weighting (IPW), with restricted cubic spline models used to assess dose–response relationships. Mixture effects were evaluated separately using Bayesian kernel machine regression (BKMR), weighted quantile sum (WQS) regression, and quantile g-computation; these mixture models were fit without IPW. Results: IPW analysis revealed that, compared with pregnant women in the lowest tertile of urinary PAH metabolites, the top three metabolites associated with significantly elevated odds ratios for iron deficiency in those in the highest tertile were 2-OHPhe (OR = 1.96, 95% CI: 1.46–2.63), 3-OHPhe (OR = 1.92, 95% CI: 1.43–2.58), and 1-OHPhe (OR = 1.83, 95% CI: 1.36–2.45). WQS regression found a significant overall mixture index effect (β = 0.28, OR = 1.32, 95% CI: 1.11–1.57, p = 0.002), with 3-OHPhe (48.42%), 1-OHPhe (18.49%), and 1-OHPyr (16.53%) accounting for the largest shares of the total mixture weight. Quantile g-computation also detected a significant positive global mixture association (ψ = 0.13, 95% CI: 0.01–0.26, p = 0.040), with 3-OHPhe (30.39%), 2-OHPhe (29.56%), and 1-OHPyr (24.02%) as the primary positive contributors; these percentages are based on within-direction normalization and are not directly comparable to the WQS weight percentages. BKMR indicated that the estimated mixture association increased monotonically from slightly negative values at lower quantiles to positive values at higher quantiles. Conclusions: These findings provide novel evidence that exposure to specific PAHs, both individually and as a mixture, particularly phenanthrene derivatives like 3-OHPhe, is associated with increased odds of gestational iron deficiency. Given the cross-sectional design, these findings should be regarded as hypothesis-generating and warrant confirmation in prospective studies.