Forever Chemicals Show Up in Every 12-Year-Old Tested, With Lipid Changes Tied to Puberty

The findings from the PELAGIE cohort add to a growing body of evidence suggesting that per- and polyfluoroalkyl substances may interfere with lipid metabolism during childhood and adolescence. What makes this study particularly informative is its use of mixture-based statistical approaches, namely quantile g-computation, which allows researchers to estimate the combined effect of multiple PFAS compounds acting together rather than examining each chemical in isolation. This matters because humans are never exposed to a single PFAS at a time. Serum measurements in the general population typically reveal a cocktail of several compounds simultaneously, including legacy substances like perfluorooctane sulfonate and perfluorooctanoic acid alongside longer-chain compounds such as perfluoroundecanoic acid and perfluorodecanoic acid. By modeling the mixture, the researchers could capture the joint influence of co-occurring exposures, which is closer to real-world conditions than single-pollutant regression models alone.

The choice of metabolic biomarkers also deserves attention. Rather than relying solely on conventional lipid measures, the study included leptin and adiponectin, two adipokines that reflect adipose tissue function and insulin sensitivity. Leptin is produced proportionally to fat mass and signals satiety to the hypothalamus, but in states of excess adiposity, circulating leptin rises while cellular responsiveness diminishes, a phenomenon known as leptin resistance. Adiponectin, by contrast, is an insulin-sensitizing hormone whose levels tend to fall as visceral fat accumulates. The ratio between these two hormones has been proposed as a sensitive indicator of metabolic dysfunction in children, sometimes detecting disturbances earlier than changes in body mass index or standard lipid panels become apparent. Fructosamine, meanwhile, offers a practical alternative to fasting glucose measurements in studies where blood is drawn without regard to meal timing. Because fructosamine reflects the glycation of serum proteins over the preceding two to three weeks, it provides an integrated picture of recent glycemic control without requiring participants to fast, which is a considerable logistical advantage in pediatric research.

The observation that associations differed by pubertal stage is scientifically meaningful for several reasons. Puberty is characterized by profound hormonal shifts, including rising concentrations of sex steroids and growth hormone, both of which influence lipid metabolism. Growth hormone, for instance, tends to raise LDL cholesterol and lower insulin sensitivity during mid-puberty, while estrogen in girls generally improves lipid profiles by raising HDL cholesterol. These physiological changes create a moving baseline against which any environmental effect must be measured. If a chemical perturbs lipid metabolism, its apparent impact may be amplified, attenuated, or masked depending on the pubertal context in which exposure is assessed. The finding that perfluorononanoic acid and perfluorodecanoic acid were linked to higher triglycerides in peripubertal boys, while perfluorooctane sulfonate showed an inverse association with triglycerides in postpubertal boys, illustrates this complexity. Such direction reversals across developmental stages have been reported in other cohorts and may reflect genuine biological differences in how PFAS interact with maturing endocrine systems, or alternatively may arise from changes in PFAS body burden during puberty, since these chemicals bind to serum proteins and their apparent concentrations can be influenced by shifts in blood volume and body composition as children grow.

The pharmacokinetics of PFAS in children are relevant to interpreting these results. Long-chain PFAS compounds have elimination half-lives in humans that range from several years for perfluorooctane sulfonate to roughly four to five years for some perfluorinated carboxylic acids, and even longer for certain longer-chain analogues. This means that serum levels measured at age 12 partly reflect exposures accumulated over years, and in the case of the longest-lived compounds, may still carry a contribution from prenatal and early-life exposure transferred through the placenta and through breastfeeding. Because the PELAGIE cohort collected cord blood and childhood samples at earlier follow-ups, future analyses from this same population can help disentangle whether metabolic associations at age 12 are driven by concurrent exposure, by earlier life exposure, or by the cumulative burden of both. This is precisely the kind of question that cross-sectional data alone cannot resolve, and it underscores the authors’ call for longitudinal follow-up.

The detection of PFAS in every serum sample examined is consistent with biomonitoring surveys conducted across Europe and North America, which consistently find near-universal presence of perfluorooctane sulfonate, perfluorooctanoic acid, and perfluorohexane sulfonate in blood serum. These three compounds were among the most heavily produced PFAS historically and remain widespread despite phase-outs and restrictions, owing to their extreme persistence, their continued presence in the environment, and in some cases ongoing production in parts of the world where regulations are less stringent. Perfluorooctane sulfonate has been listed under the Stockholm Convention on Persistent Organic Pollutants, and perfluorooctanoic acid has been the subject of international restrictions as well, yet both remain detectable in human serum decades after their peak production. The longer-chain perfluorinated carboxylic acids examined in this study, such as perfluoroundecanoic acid and perfluorodecanoic acid, are less well characterized in the epidemiological literature, partly because they were produced in smaller volumes and measured less frequently in earlier studies. Their emergence as the compounds most consistently associated with lipid markers in this cohort highlights the value of measuring a broader panel of PFAS rather than focusing only on the legacy compounds.

Several mechanisms have been proposed to explain how PFAS might influence lipid metabolism. These compounds activate peroxisome proliferator-activated receptors, a family of nuclear receptors that regulate fatty acid oxidation, adipogenesis, and lipid storage. Experimental studies in rodents and in human cell models have shown that PFAS exposure can alter expression of genes involved in cholesterol synthesis and transport, mitochondrial fatty acid metabolism, and bile acid homeostasis. Some PFAS also interact with other nuclear receptors involved in endocrine signaling, raising the possibility that their metabolic effects are mediated partly through interference with thyroid hormone action, since thyroid status strongly influences lipid profiles in growing children. Whether these mechanisms operate similarly in preadolescents as in adults remains an open question, and the sex-specific patterns observed in this study may reflect differences in hormonal milieu, body composition, or PFAS clearance between boys and girls.

The non-fasting design of the blood collection warrants brief consideration when interpreting the lipid findings. Triglyceride concentrations in particular are sensitive to recent food intake, and non-fasting measurements can introduce variability that fasting protocols would avoid. However, growing consensus in pediatric lipidology supports the use of non-fasting lipid profiles for routine screening, since postprandial changes in total cholesterol, HDL, and LDL are generally modest, and non-fasting sampling reflects the metabolic state in which children spend most of their time. The use of the Friedewald equation to derive LDL cholesterol is standard practice, though it becomes less accurate at very high triglyceride concentrations, a situation unlikely to be common in a healthy cohort of twelve-year-olds.

It is also worth noting the strengths and limitations inherent in the study design. The PELAGIE cohort offers the advantage of detailed prospective data collection from early pregnancy onward, allowing careful adjustment for factors such as maternal characteristics, socioeconomic variables, and child diet and physical activity, all of which can confound associations between chemical exposure and metabolic health. The sample size of 451 children is respectable for a single-region birth cohort, though stratification by both sex and pubertal stage reduces the number of participants in each subgroup analysis, which can limit statistical precision and increase the possibility that some findings reflect chance. Cross-sectional designs, by their nature, cannot establish temporal ordering between exposure and outcome, and reverse causation remains a theoretical concern, since body size and body composition can influence PFAS serum concentrations through dilution and volume of distribution effects. Children with higher adiposity may show lower serum PFAS concentrations per unit of exposure simply because the same body burden is distributed across a larger volume, a phenomenon that has been documented in prior studies and that could bias observed associations in either direction.

From a public health perspective, the study reinforces the rationale for reducing PFAS exposure during childhood, a period when the developing cardiovascular and endocrine systems may be particularly susceptible to chemical perturbation. Regulatory actions in the European Union and elsewhere have increasingly targeted entire classes of PFAS rather than individual compounds, reflecting recognition that the thousands of substances in this chemical family share structural features and persistence. Dietary interventions, such as reducing consumption of food packaged in greaseproof materials, and household measures, such as filtering drinking water where contamination is known, represent practical steps families can take, though the ubiquity of these chemicals means that individual-level avoidance has limits. The findings from this French cohort, together with converging evidence from other populations, support continued biomonitoring and longitudinal research to clarify whether the metabolic associations observed at age twelve translate into measurable cardiovascular risk as these children mature into adolescence and adulthood.

Subject of Research: Association between PFAS exposure and metabolic biomarkers in 12-year-old children

Article Title: PFAS exposure in 12-year-old children and its association with metabolic markers

Article References: Lejeune, N., Rouxel, E., Monfort, C., Tillaut, H., Goxe, C., Rouget, F., Giton, F., Gaudreau, É., Lainé, F., Garlantézec, R., Cordier, S., Chevrier, C., & Warembourg, C. (2026). PFAS exposure in 12-year-old children and its association with metabolic markers. Pediatric Research. https://doi.org/10.1038/s41390-026-05396-4

Image Credits: AI Generated

DOI: 10.1038/s41390-026-05396-4

Keywords: PFAS, children, metabolic markers, cholesterol, puberty, PELAGIE cohort, lipids, environmental health, endocrine disruptors, triglycerides, exposure, year-old