Maternal paraben exposure triggers childhood overweight development

Childhood overweight and obesity have reached epidemic levels globally, affecting up to one-third of children in Europe and North America. Lifestyle and genetics do not fully explain the rapid rise, focusing attention on environmental contributors, especially during critical developmental windows such as fetal life. Endocrine-disrupting chemicals (EDCs), widely present in consumer products, can cross the placenta and may program metabolic outcomes. Parabens, alkyl esters of p-hydroxybenzoic acid used as preservatives, are common EDCs detected in urine, blood, and breast milk; exposure is often higher in urine due to rapid metabolism and excretion. Although parabens have been linked to outcomes like breast cancer and allergy, evidence regarding obesity is scarce, with only indirect associations via a non-specific metabolite (3,4-DHB). No prior study has evaluated maternal paraben exposure during pregnancy as a risk factor for childhood overweight. The study aims to determine whether prenatal exposure to parabens, particularly butyl paraben, is associated with childhood overweight and to explore underlying biological mechanisms across epidemiological, in vitro, and in vivo models.

Prior research documents global increases in pediatric obesity and implicates perinatal environmental exposures as potential contributors. EDCs present in numerous consumer products can interfere with endocrine and metabolic signaling and cross the placental barrier. Parabens are ubiquitous preservatives; human biomonitoring studies detect them commonly in urine and serum, with higher urinary concentrations due to rapid metabolism. Epidemiological evidence suggests associations of parabens with certain health outcomes (e.g., breast cancer, allergies), but links to obesity are limited and confounded when using non-specific metabolites like 3,4-dihydroxybenzoic acid. Experimental studies in mouse cell lines have reported paraben-induced adipocyte differentiation and PPARγ activation, with stronger effects for longer-chain parabens (e.g., nBuP), though human-relevant data are limited. There is a gap in evaluating prenatal maternal exposure to specific parabens and subsequent childhood overweight, as well as mechanistic pathways that could mediate such effects.

Study design combined epidemiology with in vitro and in vivo experiments in a translational framework.

• Human cohort (LINA): Prospective German birth cohort (622 mothers, 629 children) recruited at 34 weeks gestation (2006–2008). Maternal urine at 34 weeks was analyzed for eight parabens (MeP, EtP, iPrP, nPrP, sBuP, iBuP, nBuP, BzP) using validated LC-MS/MS methods (LOQ: 0.5 µg/L for MeP; 0.1 µg/L for others). Parabens with >70% <LOQ (iPrP, sBuP, BzP) were excluded. Cosmetic product use during pregnancy (daily use, up to 6 items) was collected via questionnaires, categorized as leave-on vs. rinse-off, and cross-referenced with an EDC database (TOXFOX) in 2016 to classify products as paraben-containing or not. Anthropometrics: weight/height annually; BMI classified using IOTF cut-offs from age 2–8 (overweight vs. non-overweight); macrosomia defined as birth weight ≥4000 g. Subsets: N=496 for macrosomia analysis; N=223 for ever-overweight (2–8 years) logistic models; N≈392–396 for longitudinal BMI analyses (1–8 years). Statistical analysis: Mann–Whitney U to compare urinary paraben concentrations by cosmetic use; logistic regression for odds of macrosomia and ever-overweight (tertile 3 vs tertile 1 exposure), adjusted for child sex, maternal smoking during pregnancy, parental education, gestational age at delivery, siblings, breastfeeding duration (not for birth outcomes), and maternal age; GEE models with cubic splines (knot points at 12, 25, 38, 61, 98 months) for longitudinal BMI (1–8 years), stratified by sex.
• In vitro adipogenesis: Human adipose-derived mesenchymal stem cells (ATCC) differentiated into adipocytes with exposure to nBuP at 0.5, 1, 10 µM (0.05% ethanol vehicle) throughout differentiation. Outcomes: Oil Red O triglyceride staining; impedance-based real-time monitoring (xCELLigence) of differentiation; qPCR for LEP, ADIPOQ, PPARG; ELISAs for leptin and adiponectin secretion; cell viability (MTT). Validation with murine adipose-derived MSCs exposed similarly; rosiglitazone used as positive control.
• Reporter gene assays: nBuP tested for activation of nuclear receptors (PPARγ, AR, ERα, PR, GR) in human reporter cell lines; concentration–response up to cytotoxicity IC10; effects quantified as EC10/EC15 or induction ratios.
• Mouse in vivo model: BALB/cByJ dams received subcutaneous nBuP (1.75 µg in 100 µL corn oil) twice weekly during gestation and lactation until weaning (perinatal exposure); controls received vehicle. Urinary nBuP levels in exposed mice were measured and compared to human cohort levels. Offspring outcomes: body weight tracked twice weekly to 12 weeks; body composition by EchoMRI at 12 weeks; weekly food intake (post-weaning to week 11); glucose tolerance test (GTT) and insulin tolerance test (ITT) at 9 weeks; serum adipokines/hormones (leptin, adiponectin, resistin, insulin, acylated ghrelin; estradiol) by ELISA; adipocyte size in visceral fat (H&E histology). Hypothalamic gene expression (lepr, pomc, mc4r, agrp, insr) by qPCR. DNA methylation of POMC regulatory regions (promoter, neuronal enhancers nPE1 and nPE2) by bisulfite pyrosequencing. Epigenetic intervention: offspring from exposed dams treated with 5-Aza-2'-deoxycytidine (160 µg/kg i.p., 3×/week) from postnatal week 1 to weaning to test reversibility of nPE1 methylation and phenotypes. Longitudinal mouse data analyzed via GEE; other comparisons by t-test or ANOVA.

• Source of exposure: Among 414 mothers with cosmetic data, 26% used at least one paraben-containing leave-on product during pregnancy. Users of paraben-containing leave-on products had higher urinary parabens than users of paraben-free leave-on products (median µg/L; Mann–Whitney U): MeP 68.80 vs 28.05 (p=0.0018), EtP 2.90 vs 1.89 (p=0.0466), nPrP 7.40 vs 3.20 (p=0.0025), nBuP 1.24 vs 0.41 (p=0.0004), iBuP 0.20 vs 0.10 (p=0.1875).
• Human epidemiology: High prenatal exposure (tertile 3 vs tertile 1) to iBuP and nBuP increased risk of ever being overweight between ages 2–8 years (adjusted OR [95% CI]): iBuP 2.40 [1.16–4.98], p=0.018; nBuP 2.17 [1.06–4.47], p=0.035. No significant effects on macrosomia at birth for any paraben. Longitudinal BMI (1–8 years) showed higher intercepts with higher exposure to BuPs: iBuP tertile 2 beta 0.35 (0.08,0.62), p=0.011; iBuP tertile 3 beta 0.26 (0.02,0.50), p=0.035. Effects were stronger in girls (e.g., iBuP tertile 3 beta 0.53 [0.16,0.89], p=0.005). nBuP showed a positive trend overall (tertile 3 beta 0.21 [−0.04,0.47], p=0.095) and in girls (0.36 [−0.03,0.74], p=0.069).
• In vitro adipogenesis: nBuP (0.5–10 µM) did not promote human or murine MSC adipocyte differentiation by Oil Red O staining or impedance monitoring, and did not increase PPARG expression or activate PPARγ, AR, PR, or GR reporters. nBuP strongly affected ERα activity. Leptin (LEP) mRNA decreased significantly even at 0.5 µM; secreted leptin decreased significantly at 10 µM. Adiponectin secretion increased in a concentration-dependent manner. No cytotoxicity at tested concentrations.
• Mouse in vivo: Female offspring of nBuP-exposed dams had persistently higher body weight (GEE estimate +3 g; SE 0.5; p=5.3×10⁻⁹), increased fat mass, decreased lean mass, and higher weekly food intake (+0.5 g/week; SE 0.2; p=0.006). Fasting glucose was elevated but GTT/ITT were not impaired. Male offspring showed no significant differences. Adult female mice exposed to nBuP showed no changes in weight, food intake, or leptin.
• Adipose and hormonal phenotypes: Female offspring from exposed dams had larger adipocyte area and higher serum leptin; adiponectin, resistin, ghrelin, insulin were unchanged.
• Hypothalamic mechanism: In female offspring, hypothalamic lepr and pomc mRNAs were downregulated; mc4r, agrp, insr were unchanged. DNA methylation was increased at the neuronal POMC enhancer nPE1 (no change at promoter or nPE2), detectable as early as weaning. Treating offspring with the DNA methyltransferase inhibitor Aza reversed nBuP-induced nPE1 hypermethylation, restored pomc expression, and reduced elevated body weight, food intake, adipocyte area, serum leptin, and restored lepr expression.

The study integrates cohort data with mechanistic experiments to address whether prenatal paraben exposure contributes to childhood overweight. Epidemiological analyses linked higher maternal urinary BuP (iBuP and nBuP) to increased odds of overweight between ages 2–8 years, particularly in girls, and to higher BMI trajectories. The mouse model causally corroborated these findings: perinatal nBuP exposure led to increased food intake, adiposity, and weight gain in female offspring. In vitro experiments did not support a direct adipogenic effect of nBuP on human adipocyte differentiation, suggesting the weight gain is less likely due to enhanced adipogenesis per se. Instead, central appetite regulation emerged as a key pathway: nBuP exposure reduced hypothalamic pomc and lepr expression, with epigenetic hypermethylation of the neuronal enhancer nPE1 implicated as a mechanistic driver. Reversal of phenotypes by Aza supports a causal role of DNA methylation changes in mediating increased appetite and weight. The sex-specific effects may relate to estrogenic activity (ERα modulation) and sex hormone milieu during development, although precise mechanisms require further investigation. The findings underscore cosmetic products as a significant exposure source and suggest that perinatal exposure to long-chain parabens may predispose offspring to overweight via epigenetic programming of hypothalamic satiety pathways.

This translational study shows that prenatal exposure to butyl parabens, particularly nBuP, is associated with and can causally induce increased childhood/female offspring adiposity through altered central appetite regulation. The mechanism involves epigenetic hypermethylation of the neuronal POMC enhancer nPE1 and reduced hypothalamic pomc expression, leading to increased food intake. Cosmetic leave-on products are a relevant exposure source for pregnant women. These results highlight the importance of minimizing environmental paraben exposure during pregnancy to reduce later obesity risk. Future research should: (1) delineate sex-specific mechanisms, including roles of ERα signaling; (2) map genome-wide epigenetic changes and cell-type specificity in hypothalamic circuits; (3) evaluate dose–response relationships at current regulatory limits and across mixtures of parabens; (4) replicate findings in other cohorts with repeated exposure measures; and (5) explore intervention strategies to mitigate epigenetic programming effects.

• Human exposure assessment relied on a single maternal urine sample at 34 weeks, which may not capture temporal variability.
• Cosmetic product composition was assessed retrospectively via the TOXFOX database (2016) and may underestimate paraben content for products used in 2006–2008; other exposure sources (food, medications) were not fully accounted for.
• Regulatory changes in 2014 limiting paraben concentrations may make current exposures lower than during recruitment, affecting generalizability.
• Several parabens (iPrP, sBuP, BzP) had >70% of values below LOQ and were excluded.
• Overweight classification lacked reference data at age 1; analyses of ever-overweight were limited to ages 2–8 with a subcohort (n=223).
• Observational human analyses are subject to residual confounding despite adjustments.
• Mouse dosing via subcutaneous injection may differ from typical human exposure routes (dermal/oral), and species differences limit direct extrapolation.
• Aza is a global DNA methylation inhibitor; reversal experiments support epigenetic involvement but lack locus-specific causality.
• Some experimental groups had modest sample sizes, potentially limiting power for certain endpoints and sex-stratified analyses.