Prenatal acrylamide exposure linked to higher depression and anxiety risk in offspring

Children exposed to higher levels of acrylamide in the womb were significantly more likely to be diagnosed with depression and anxiety decades later, according to a prospective birth cohort study published in eClinicalMedicine. The study, which followed offspring for up to 34 years, found that those whose mothers had high blood levels of the dietary and tobacco-smoke contaminant during pregnancy showed a 1.6-fold higher incidence rate of depression and a 2.3-fold higher incidence rate of anxiety compared with children whose mothers had lower levels, even after accounting for maternal smoking during pregnancy.

The research is the first to quantify prenatal exposure to acrylamide using a blood biomarker and to track clinically diagnosed neuropsychiatric disorders over such a long follow-up period. Acrylamide forms in a wide range of commonly consumed carbohydrate-containing foods and drinks during high-temperature cooking, with high concentrations found in potato chips, fried potatoes, biscuits, bread, coffee and baked cereals. Tobacco smoke is another major source. The compound is classified as a Group 2A “probable human carcinogen” by the International Agency for Research on Cancer, is neurotoxic and genotoxic, crosses the human placenta, and has been detected in human second trimester amniotic fluid.

Researchers used the Danish Foetal Origins 1988–1989 (DaFO88) cohort, in which 991 of 1212 eligible women with singleton pregnancies were recruited at a routine antenatal visit at Aarhus University Hospital in gestational week 30. Blood samples collected around gestational week 30 from 896 women were analysed for haemoglobin adducts of acrylamide (HbAA), a stable chemical product formed when acrylamide covalently binds to amino acid residues of haemoglobin inside red blood cells. These adducts reflect internal dose accumulated over the previous four months from all exposure sources, including diet and tobacco smoke, and provide a more accurate measure than food-frequency questionnaires, since acrylamide concentrations in food vary widely depending on cooking methods, food composition and storage.

Offspring health outcomes were obtained by linking cohort data to nationwide Danish registries covering 1988 to 2023. Cases of depression, anxiety and attention-deficit/hyperactivity disorder (ADHD) were identified from redemption of at least two prescriptions for medications indicated for these disorders or one diagnosis according to International Classification of Diseases codes, whichever occurred first. Cases of autism spectrum disorder (ASD) and any neuropsychiatric diagnoses were identified from the first occurrence of a primary diagnosis. By the end of follow-up, 22 percent of offspring had received one or more registrations of any neuropsychiatric diagnosis, 21 percent had depression, 11 percent had anxiety, 8 percent had co-occurring ASD and ADHD, and 7 percent had ADHD.

The analysis included 861 participants in models with imputed covariates, with descriptive characteristics reported for 687 complete cases. The median HbAA level was 93.0 pmol/g haemoglobin, with a range from the 5th to 95th percentile of 45.0 to 249.7. Women who smoked during pregnancy had substantially higher HbAA levels than non-smoking women, 148.0 versus 74.0 pmol/g haemoglobin, and 87 percent of the high-exposure group were smokers compared with 27 percent of the low-exposure group. Women with high HbAA levels were also more commonly of lower education and income and had more often received a neuropsychiatric diagnosis themselves.

When HbAA was modelled as a continuous variable, each 10-pmol/g haemoglobin increase was associated with an adjusted incidence rate ratio of 1.02 (95 percent confidence interval 0.99 to 1.05) for depression and 1.03 (0.99 to 1.07) for anxiety, neither of which reached statistical significance. However, when comparing offspring of mothers with high exposure, defined as at or above the 75th percentile of 146 pmol/g haemoglobin, with those below that threshold, the incidence rate ratio was 1.64 (1.05 to 2.56) for depression and 2.29 (1.23 to 4.25) for anxiety. The incidence rate of depression was 6.7 per 1000 person-years among the low-exposure group and 9.3 per 1000 person-years among the high-exposure group, while the corresponding rates for anxiety were 2.7 and 5.3 per 1000 person-years.

The reason for this discrepancy lies in the shape of the exposure-response relationship. Using natural cubic splines with three degrees of freedom, the researchers found that risk estimates increased especially for depression and anxiety when HbAA levels exceeded approximately 150 pmol/g haemoglobin, and likelihood ratio tests rejected linearity for anxiety (p < 0.001) and depression (p < 0.05). The authors note that the exposure-response relationships for these two outcomes are therefore not adequately summarised by a single linear slope, which explains why the estimates per 10-pmol/g increment did not reach statistical significance while the comparison of high with low exposure did. When HbAA was categorised into quartiles, the fourth quartile showed an incidence rate ratio of 1.78 (1.02 to 3.11) for depression compared with the first quartile, with a p-value for trend of 0.07, while the quartile estimates for anxiety did not show a significant trend (p for trend = 0.80).

Because tobacco smoking is a source of acrylamide, the researchers modelled both exposures separately and concurrently. In mutually adjusted models including maternal active smoking, HbAA levels, ethylene oxide adducts as an internal biomarker of tobacco smoke exposure, alcohol intake, a healthy eating index, education, household income, parity, offspring sex and gestational age, the association between high acrylamide exposure and depression and anxiety remained significant, while the estimate for maternal smoking itself was 1.00 (0.66 to 1.52) for depression and 1.10 (0.60 to 2.02) for anxiety. Maternal smoking was, however, associated with increased risk for ADHD (2.29, 1.14 to 4.62) and co-occurrence of ASD and ADHD (2.89, 1.52 to 5.48). The increased risks from high acrylamide exposure were observed in both offspring of nonsmokers and smokers, with a depression estimate of 1.79 (0.58 to 3.48) in offspring of nonsmokers and 1.47 (0.82 to 2.63) in offspring of smokers, although with wide confidence intervals.

The study has important limitations. The population was relatively small, restricting statistical power for rarer outcomes: with 45 ADHD cases and roughly a quarter of the population classified as highly exposed, the study had 80 percent power to detect an incidence rate ratio of 2.6 for ADHD, and the estimates for ADHD, co-occurring ASD and ADHD, and any neuropsychiatric disorder pointed in the same direction but were underpowered and did not reach statistical significance. The authors stress these should not be interpreted as evidence of an absence of an effect. HbAA was measured at a single time point, so later changes in maternal and offspring dietary or lifestyle habits may not be captured, and subsequent exposure may contribute to the observed associations. As an observational study, residual confounding by behavioural factors or maternal mental health cannot be ruled out, and confidence intervals and p-values were not adjusted for multiple comparisons, so chance findings cannot be excluded. The findings do not establish a causal effect or demonstrate that an intervention reducing prenatal exposure would reduce the incidence of these disorders. The cohort, recruited at a single Danish hospital and almost exclusively of Danish origin, also limits generalisation to more diverse populations.

Biological plausibility is supported by extensive animal and in vitro studies demonstrating that acrylamide causes neurotoxicity and developmental neurotoxicity, potentially involving axonal degeneration, neuronal apoptosis and autophagy, oxidative stress, neuroinflammation, mitochondrial dysfunction, disrupted neuronal migration and differentiation, and impaired axonal transport. Rodent studies show that gestational exposure to acrylamide causes structural brain changes in offspring, including altered hippocampal development, reduced neurotrophic factor expression, and persistent motor and cognitive deficits into adulthood. The authors conclude that if the findings are confirmed, given that exposure to acrylamide is ubiquitous and the large burden of depression and anxiety, they support continued efforts to decrease exposure to acrylamide from both diet and tobacco smoke among pregnant women.

Subject of Research: Prenatal acrylamide exposure and long-term risk of neuropsychiatric disorders in offspring

Article Title: Prenatal exposure to acrylamide and neuropsychiatric disorders: a biomarker-based birth cohort study with 34 years of follow-up

Article References: Chandia-Poblete, D., Tuffier, S., Vryonidis, E., Halldorsson, T. I., Bjerregaard, A. A., Strøm, M., Rytter, D., Bech, B. H., Henriksen, T. B., Olsen, S. F., Törnqvist, M., & Pedersen, M. (2026). Prenatal exposure to acrylamide and neuropsychiatric disorders: a biomarker-based birth cohort study with 34 years of follow-up. eClinicalMedicine, 101, Article 104286. https://doi.org/10.1016/j.eclinm.2026.104286

Image Credits: AI Generated

DOI: 10.1016/j.eclinm.2026.104286

Keywords: acrylamide, prenatal exposure, depression, anxiety, birth cohort, haemoglobin adducts, neuropsychiatric disorders, ADHD, autism spectrum disorder, maternal smoking, neurotoxicity, Danish registries