Metallomics Reviews
Perinatal Tobacco Exposure and the Child Gut Microbiota
Clinical Overview
This birth-cohort study from Sabadell (Spain) evaluated whether perinatal and childhood exposure to tobacco smoke and mercury shapes the gut microbiota of 151 healthy 7-year-old children. Total mercury was quantified in cord blood (μg/L) and 4-year hair (μg/g), tobacco via maternal/child urinary cotinine (μg/g creatinine) and questionnaires, and the fecal microbiota via 16S rRNA V3–V4 sequencing. Despite high mercury exposure (65% of children with cord blood >5.8 μg/L) and substantial tobacco exposure in utero (46.6%), only prenatal active maternal smoking showed a robust association: reduced Akkermansia and increased Dorea relative abundances at age 7 (q≈0.004–0.01).
What was reviewed and who was studied
The paper reports an exposome-wide microbiome analysis within the INMA-Sabadell birth cohort, focusing on tobacco smoke and mercury exposures from pregnancy through age 7 and their relation to fecal microbiota in 151 children. Exposures were assessed by biomonitoring (cord blood and hair mercury, urinary cotinine) and questionnaires, while fecal samples collected at 7 years were profiled at genus level using Illumina MiSeq 16S rRNA sequencing and compositional statistics (Aitchison distance, MaAsLin2).
Major findings
The main findings are summarized below.
| Finding | Details |
|---|---|
| Exposure burden | 46.6% of children were exposed to tobacco smoke in utero, 30.3% during childhood, and 17.9% had sustained exposure across periods. Mean urinary cotinine was 224.1 μg/g creatinine in mothers and 17.5 μg/g in children; 65% had cord blood mercury >5.8 μg/L and 38.4% had hair mercury >1 μg/g. |
| Core microbiota structure | At 7 years, fecal communities were dominated by Bacteroidota (44.65%), Firmicutes (39.28%), Verrucomicrobiota (11.02%), Actinobacteria (2.74%), and Proteobacteria (2.29); key genera included Bacteroides (30.9%), Akkermansia (11.2%), Alistipes (11.1%), and Faecalibacterium (10.8%). |
| β-diversity determinants (envfit) | In multi-determinant models using Aitchison distance, active smoking during pregnancy and second-hand smoke at 4 years (urinary cotinine ≥4 ng/mL) each explained ~2–4% of β-diversity variance (p=0.0045 and 0.0322). Having siblings at birth also associated with β-diversity. Genetic variables (GWAS PCs and FUT2) explained <5%. |
| β-diversity in pollutant-specific models | In adjusted single-pollutant PERMANOVA models (Aitchison), none of the 12 tobacco or 2 mercury metrics showed significant β-diversity associations after multiple-testing correction. |
| Genus-level effects of prenatal tobacco | Maternal active smoking at any time in pregnancy was associated with lower Akkermansia abundance at 7 years (coef≈−3.75, q≈0.005). Sustained maternal smoking from first to third trimester was associated with higher Dorea abundance (coef≈2.0, q≈0.006–0.01). |
| Mercury–microbiome relations | Despite high total mercury levels in cord blood and hair, neither prenatal nor 4-year mercury metrics were associated with β-diversity or genus-level abundance after adjustment for fish intake and covariates. |
Implications for Microbial Metallomics
The study links a well-characterized exposure metallome (total mercury) and tobacco-related biomarkers to long-term microbiome composition, highlighting prenatal tobacco smoke rather than mercury as the key determinant of genus-level structure at age 7.
| Concept | Implication |
|---|---|
| High prenatal mercury burden without microbiome signal | In this cohort, cord blood and hair total mercury, even above reference thresholds, did not translate into detectable genus-level or β-diversity changes at 7 years, suggesting that long-term gut compositional shifts may be less sensitive to mercury than to tobacco, at least at this taxonomic resolution. |
| Prenatal tobacco exposure and Akkermansia depletion | The sustained reduction of Akkermansia in children of mothers who smoked during pregnancy supports the idea that in utero exposure can durably alter mucosa-associated taxa, providing a candidate microbial mediator for tobacco-related developmental effects. |
| Prenatal tobacco exposure and Dorea enrichment | Increased Dorea with sustained maternal smoking suggests that early-life tobacco exposure may favor expansion of genera often framed as potentially pathogenic, aligning with a pro-inflammatory configuration that may be relevant for later disease phenotypes. |
| Multi-determinant compositional modeling | Use of Aitchison distance and envfit within a 69-variable exposome framework shows that pollutant–microbiome effects can be disentangled from diet, BMI, and social environment, offering a methodological template for future metallomic–microbiome analyses. |
| Lack of childhood tobacco effect at 7 years | The absence of strong associations for 4- and 7-year tobacco metrics implies that prenatal exposures may be more critical than later exposures for setting microbiome trajectories, guiding timing of exposure reduction strategies. |
| Mercury metrics and matrix choice | Reliance on cord blood and hair total mercury, with hair only partly reflecting methylmercury and not correlating well with blood or toxicity, underscores the importance of carefully selecting matrices and speciation endpoints for microbiome-metallomics studies. |
Limitations
Key limitations include modest microbiome sample size (n=151) relative to the large determinant set, which restricts power, and storage of fecal samples at 4°C for up to 49 hours before freezing, potentially influencing community profiles. Antibiotic use in the month before sampling was not captured, and other unmeasured exposures (e.g., additional metals, endocrine disruptors, air pollution) may confound associations. Observational design and exposure data taken at 4 years, prior to microbiome sampling at 7 years, limit causal inference and temporal precision.
Future perspectives
Logical next steps include repeating this analysis with species- and strain-level resolution via shotgun metagenomics to clarify which Akkermansia and Dorea lineages are impacted by prenatal tobacco exposure and whether other taxa masked at genus level respond to mercury. Earlier and more granular mercury biomarkers, especially during early gestation, could clarify critical windows of susceptibility. Pooling INMA with similarly profiled cohorts would increase power to detect modest mercury–microbiome effects and allow joint modeling of multiple pollutants. Longitudinal follow-up of these children for metabolic, respiratory, and inflammatory outcomes could test whether the observed microbial signatures mediate later health risks.
Key takeaways for Researchers and Clinicians
This Spanish birth-cohort analysis links detailed tobacco and mercury exposure histories from pregnancy through early childhood to the fecal microbiota of 151 7-year-old children. Total mercury in cord blood (μg/L) and hair (μg/g), despite frequently exceeding reference levels, showed no clear association with gut β-diversity or genus-level composition. In contrast, prenatal active maternal smoking, but not later childhood tobacco exposure, was consistently related to altered microbiota structure, including lower Akkermansia and higher Dorea abundances at school age, after adjustment for maternal education, BMI, siblings at birth, and fish intake where relevant.
Methodologically, the study illustrates the value of compositional approaches (Aitchison distance, clr transformation), multi-imputation for covariates, and multivariable MaAsLin2 modeling in exposome–microbiome research. Clinically, these findings support reinforcing tobacco cessation before and during pregnancy, not only for direct fetal toxicity but also for potential long-term imprinting of the child gut microbiota. A concise translational hook is that prenatal tobacco smoke appears more influential than a high early-life mercury burden in shaping the 7-year microbiome, positioning specific genera such as Akkermansia and Dorea as candidate microbial sentinels of in utero exposure.
Citation
Pérez-Castro S, D’Auria G, Llambrich M, Fernández-Barrés S, Lopez-Espinosa M-J, Llop S, Regueiro B, Bustamante M, Francino MP, Vrijheid M and Maitre L (2024) Influence of perinatal and childhood exposure to tobacco and mercury in children’s gut microbiota. Front. Microbiol. 14:1258988. doi: 10.3389/fmicb.2023.1258988