Prenatal Lead Exposure is Negatively Associated with the Gut Microbiome in Childhood Original paper

What was studied?

This study investigated prenatal lead exposure and gut microbiome relationships in late childhood, testing whether maternal blood lead measured in the second and third trimesters predicts gut microbial composition, diversity, and functions in children aged 9–11 years. Maternal whole-blood lead was quantified by ICP-MS, and child stool underwent shotgun metagenomic sequencing (MetaPhlAn2/StrainPhlAn for taxonomy; HUMAnN2 for pathways). Analyses included Shannon alpha diversity, Bray–Curtis beta diversity via PERMANOVA, weighted quantile sum (WQS) regression with repeated holdouts to model “mixture” effects across taxa, taxa-wide association (TWAS) for individual species, and pathway summaries. Models adjusted for child sex/age, maternal age, BMI, socioeconomic status, and sequencing batch.

Who was studied?

Participants were 123 children from the Mexico City PROGRESS prospective birth cohort (49 female, 74 male) who provided stool at 9–11 years. Mean prenatal maternal blood lead was 33.6 µg/L (second trimester) and 34.9 µg/L (third trimester). Families were generally low-to-middle SES; SES skewed toward higher lead quartiles, and batch indicators were included to mitigate processing heterogeneity. Stool samples were collected at home/clinic, rapidly refrigerated, processed with the FAST protocol, and stored at −70 °C before sequencing.

Most important findings

Across methods, associations were directionally consistent and predominantly negative. Alpha diversity showed small, non-significant inverse associations with prenatal lead in adjusted linear models, whereas beta diversity was near-null for second-trimester lead but trended for third-trimester lead (adjusted R²≈1.1%, p=0.066). In WQS mixture analysis, both trimesters showed similar negative overall effects on the gut microbiome (β≈−0.17 for second and third trimesters; 88–89% of repeated holdouts below zero), with the negative-direction model strongly favored. Taxa repeatedly carrying high weights in the negative mixture included Ruminococcus gnavus, Bifidobacterium longum, Bifidobacterium bifidum, Alistipes indistinctus, and Bacteroides caccae.

TWAS supported trimester-specific patterns: for the second-trimester lead, inverse associations were observed for Alistipes putredinis, R. gnavus, B. caccae, B. intestinalis, Coprococcus catus, and A. indistinctus, while B. coprocola was positively associated; for the third trimester, B. bifidum, B. longum, and A. indistinctus were inversely associated, with positive associations for B. coprocola, Eubacterium eligens, and B. finegoldii. Pathway analysis of highly weighted taxa indicated that shared (“core”) pathways were enriched for nucleic acid biosynthesis/essential functions, while trimester-specific sets more often involved amino-acid biosynthesis and specialized metabolism, suggesting potential functional reprogramming windows. Collectively, results point to a modest but consistent adverse shift in community structure and key taxa linked to mucin utilization (R. gnavus), early-life colonization and carbohydrate metabolism (Bifidobacterium spp.), fiber degradation (B. caccae), and variable health signals within Alistipes.

Key implications

For clinicians, these pilot data support that prenatal lead may imprint the child’s gut microbiome years later, with the second trimester emerging as a plausible sensitive window. Microbiome-signature curation for metal exposure should therefore record timing (2T vs 3T), exposure magnitude, and co-determinants (SES, diet, antibiotics) when interpreting taxa such as Bifidobacterium (often beneficial early colonizers), R. gnavus (mucin degrader linked to dysbiosis contexts), B. caccae (fiber degrader with mixed health effects), and A. indistinctus (reports of both protective and adverse roles). Important caveats include small sample size, two sequencing batches (addressed analytically), WQS role-reversal of outcome/exposures (association-seeking), and maternal blood as a proxy for fetal lead. Nonetheless, the convergence of negative directions across diversity, mixture, and taxa-level analyses suggests clinically meaningful, exposure-timing-aware microbiome alterations that could inform risk stratification, surveillance of developmental cohorts, and future interventional studies.