Disparities in chemical exposures among pregnant women and neonates by socioeconomic and demographic characteristics: A nontargeted approach

Abstract

Background: Exposure to environmental chemicals during pregnancy adversely affects maternal and infant health, and identifying socio-demographic differences in exposures can inform contributions to health inequities.

Methods: We recruited 294 demographically diverse pregnant participants in San Francisco from the Mission Bay/Moffit Long (MB/ML) hospitals, which serve a primarily higher income population, and Zuckerberg San Francisco General Hospital (ZSFGH), which serves a lower income population. We collected maternal and cord sera, which we screened for 2420 unique formulas and their isomers using high-resolution mass spectrometry using LC-QTOF/MS. We assessed differences in chemical abundances across socioeconomic and demographic groups using linear regression adjusting for false discovery rate.

Results: Our participants were racially diverse (31% Latinx, 16% Asian/Pacific Islander, 5% Black, 5% other or multi-race, and 43% white). A substantial portion experienced financial strain (28%) and food insecurity (20%) during pregnancy. We observed significant abundance differences in maternal (9 chemicals) and cord sera (39 chemicals) between participants who delivered at the MB/ML hospitals versus ZSFGH. Of the 39 chemical features differentially detected in cord blood, 18 were present in pesticides, one per- or poly-fluoroalkyl substance (PFAS), 21 in plasticizers, 24 in cosmetics, and 17 in pharmaceuticals; 4 chemical features had unknown sources. A chemical feature annotated as 2,4-dichlorophenol had higher abundances among Latinx compared to white participants, those delivering at ZSFGH compared to MB/ML, those with food insecurity, and those with financial strain. Post-hoc QTOF analyses indicated the chemical feature was either 2,4-dichlorophenol or 2,5-dichlorophenol, both of which have potential endocrine-disrupting effects.

Conclusions: Chemical exposures differed between delivery hospitals, likely due to underlying social conditions faced by populations served. Differential exposures to 2,4-dichlorophenol or 2,5-dichlorophenol may contribute to disparities in adverse outcomes.

Figure 1.

Number of chemical features with significant abundance differences by socioeconomic and demographic groups in maternal and cord sera. Note: Abundance differences were considered significant at the false discovery rate (FDR) p-value of 0.1.

Figure 2.

Abundances differences by hospital of delivery (MB/ML vs. ZSFGH) in maternal and cord serum. Note: MB/ML hospitals are the reference group. The dotted line indicates a false discovery rate (FDR) p-value of 0.1. Chemical features with abundance differences significant at the FDR-adjusted level of 0.1 are annotated with their likely names.

Figure 3.

Abundance differences by race/ethnicity in maternal serum. Note: White participants are the reference group. The dotted line indicates a false discovery rate (FDR) p-value of 0.1. Chemical features with abundance differences significant at the FDR-adjusted level of 0.1 are annotated with their likely names.

Figure 4.

Abundance differences by race/ethnicity in cord serum. Note: White participants are the reference group. The dotted line indicates a false discovery rate (FDR) p-value of 0.1. Chemical features with abundance differences significant at the FDR-adjusted level of 0.1 are annotated with their likely names.

Figure 5.

Abundance differences by in maternal serum by maternal age, educational attainment, employment status, nativity, occupation, food insecurity, and financial strain. Note: The dotted line indicates a false discovery rate (FDR) p-value of 0.1. Chemical features with abundance differences significant at the FDR-adjusted level of 0.1 are annotated with their likely names. Born in the US is the nativity reference group. Food secure is the reference group. No financial strain is the reference group.

Figure 6.

Abundance differences in cord serum by maternal age, educational attainment, employment status, nativity, occupation, food insecurity, and financial strain. Note: The dotted line indicates a false discovery rate (FDR) p-value of 0.1. Chemical features with abundance differences significant at the FDR-adjusted level of 0.1 are annotated with their likely names. Born in the US is the nativity reference group. Food secure is the reference group. No financial strain is the reference group.