Prenatal Acetaminophen, the Infant Gut Microbiome, and Childhood Cognition

What the study examined

This 2022 analysis by Laue and colleagues used the GESTation and Environment (GESTE) cohort, a prospective birth cohort recruited between 2007 and 2009 at the University of Sherbrooke in Quebec, Canada. The study asked whether acetaminophen (paracetamol) and caffeine — two of the most common xenobiotics encountered during pregnancy and early childhood — are associated with the composition of the child gut microbiome and with cognitive development at ages 6 to 7 years.

Prenatal exposure was assessed by quantifying acetaminophen and caffeine directly in meconium (the first neonatal stool, which integrates late-gestation exposure), available for 49 children. Childhood exposure was measured in stool collected during home visits, available for 85 children. Both compounds were extracted from feces and measured by liquid chromatography-tandem mass spectrometry (LC-MS/MS). Acetaminophen was analyzed as detected versus non-detectable because roughly half of meconium and about 90% of childhood stool samples were below the detection limit; caffeine was log2-transformed as a continuous measure.

The gut microbiome was characterized by shotgun metagenomic sequencing of fecal DNA, yielding species-level taxonomy, phylum-level relative abundances, and functional metabolic pathways. Cognition was evaluated with subtests of the Wechsler Intelligence Scale for Children, Fourth Edition (WISC-IV) — including Block Design, Coding, Digit Span, Information, and Vocabulary — plus a summary score, and motor development was assessed by questionnaire.

Key findings

Prenatal acetaminophen exposure was associated with lower gut bacterial alpha diversity in childhood (Shannon index beta approximately -0.17, 95% CI -0.31 to -0.04), and with shifts in phylum-level composition: lower relative abundance of Firmicutes (beta approximately -0.09) and higher relative abundance of Actinobacteria (beta approximately 0.18). These associations were of borderline statistical significance after correction for multiple testing, so the authors described them as suggestive trends rather than definitive effects.

The most notable neurodevelopmental result was an interaction rather than a main effect. Acetaminophen and caffeine showed no consistent direct association with cognitive scores, but the relative abundance of Proteobacteria modified the exposure-cognition relationship. Among children with a higher relative abundance of Proteobacteria, the pattern of association between prenatal acetaminophen/caffeine exposure and WISC-IV subscale scores differed markedly from children with low Proteobacteria — the microbiome appeared to condition whether and how these exposures tracked with cognition.

Prenatal exposures were more strongly associated with the childhood microbiome than concurrent childhood exposures, consistent with a developmental window during which the naive gut is colonized. Prenatal acetaminophen was also linked to enrichment of a succinate-fermentation-to-butanoate metabolic pathway, while childhood caffeine tracked with methionine- and sulfate-related pathways. Prenatal caffeine, by contrast, showed no significant associations with diversity, taxa, or pathways.

The proposed mechanism

The infant gut is assembled during a narrow early-life window: the sterile or near-sterile fetal gut is colonized around and after birth, and the community that establishes then helps train immune, metabolic, and neurologic development. A pharmacologically active compound present during this assembly can act as a selective pressure, favoring some taxa over others and steering the trajectory of the maturing community. Because acetaminophen is metabolized in part by gut bacteria and can alter the redox and metabolic environment of the lumen, exposure during colonization plausibly leaves a durable compositional imprint — which is consistent with the observation that prenatal exposure predicted the microbiome years later better than concurrent exposure.

Proteobacteria (a phylum that includes the Enterobacteriaceae) function here as a signal of community state. A relative expansion of Proteobacteria is a widely used, non-specific marker of gut dysbiosis and low community stability, often reflecting increased luminal oxygen and inflammation. A Proteobacteria-rich gut is therefore a different biochemical environment — one in which microbial metabolites reaching the developing brain via the gut-brain axis (short-chain fatty acids, neuroactive metabolites, immune mediators) may differ from a Firmicutes-dominated, more mature community. That offers a coherent reason why Proteobacteria abundance, rather than exposure alone, was the variable that tracked with cognitive differences.

How it fits the metal-microbiome-disease axis

This study does not involve heavy metals. Acetaminophen and caffeine are organic xenobiotics, not metals, and no metal exposure was measured, so it cannot speak directly to metal toxicity. Its value to the metal-microbiome-disease axis is as a mechanistic template: it demonstrates the general causal shape that the axis depends on — an early-life environmental exposure reshapes the gut microbiome, and the resulting microbiome state is what tracks with a downstream health outcome (here, cognition).

The specific node it illuminates is the one metal exposures also converge on: Proteobacteria expansion as a marker of dysbiosis. Independent toxicology literature reports that developmental exposure to lead, arsenic, and cadmium can likewise expand Enterobacteriaceae/Proteobacteria and reduce microbial diversity. The GESTE finding that a Proteobacteria-dominated gut alters the relationship between an exposure and neurodevelopment is therefore a real, transferable insight: it suggests that the microbiome is not a passive bystander but an effect modifier standing between exposure and disease.

This is also where metallobiology becomes relevant even though metals were not studied. A Proteobacteria-rich, inflamed gut is precisely the setting in which nutritional immunity and microbial metal competition intensify: host calprotectin sequesters zinc and manganese to starve pathogens, while Enterobacteriaceae deploy siderophores to scavenge iron. Luminal iron availability is itself a strong driver of Proteobacteria blooms. So the dysbiosis signature this paper links to cognition is metabolically intertwined with trace-metal handling in the gut — a plausible, mechanistic bridge to the axis, but one that this cohort did not test and that should not be over-read from these data.

Interpretation and limitations

The results should be read as hypothesis-generating. Sample sizes were small (49 with meconium, 85 with childhood stool), acetaminophen had to be dichotomized because most samples were non-detectable, and the population was relatively homogeneous (largely upper-middle-class French-Canadian families), which limits generalizability. Several microbiome associations were only borderline after correction for multiple comparisons, and the key cognitive result was an interaction with Proteobacteria rather than a main effect — a pattern that needs replication before it can be considered robust.

The authors are careful to frame the work as suggestive evidence that early-life gut bacteria may modify how common xenobiotic exposures relate to neurodevelopment, and they call for larger and more diverse cohorts. For the metal-microbiome-disease axis, the appropriate takeaway is structural rather than substantive: the study validates the exposure-to-microbiome-to-outcome logic and highlights Proteobacteria as a shared, measurable pivot point, while leaving the direct metal question for studies that actually quantify metal exposure alongside the microbiome and the outcome.