A growing body of research is pushing scientists to reconsider whether some cases of autism spectrum disorder can be prevented through changes in prenatal and early-life exposures. Helen Tager-Flusberg, director of Boston University’s Center for Autism Research Excellence and a founder of a coalition of autism researchers, is among those raising questions about whether avoiding certain toxins before birth or in the first months of life could lower a child’s risk, as recently described in reporting on her work. The shift comes as the FDA initiated a label change process for acetaminophen on September 22, 2025, reflecting evidence of possible associations with autism and ADHD, and as new peer-reviewed findings link parental avoidance of environmental toxins to reduced rates of neurodevelopmental disorders. One widely discussed study, highlighted in coverage of emerging autism prevention questions, suggested that parents who took steps to reduce chemicals in their homes before conception and during pregnancy had children with lower rates of autism diagnoses, intensifying debate over how much risk can realistically be modified.
Prenatal Drug Exposures and Measurable Risk
The clearest evidence that prenatal chemical exposure can raise autism risk comes from valproate, an anti-seizure medication used to treat epilepsy and bipolar disorder. A large Danish registry analysis, available through a national birth cohort, found that children exposed to valproate in the womb had an absolute autism spectrum disorder risk of approximately 4.42 percent, compared with a background rate of roughly 1.53 percent among unexposed children. After adjusting for maternal epilepsy and other factors, the hazard ratio for any ASD was about 2.9, and for childhood autism specifically it climbed to around 5.2, underscoring that valproate is an outlier among medications in terms of documented neurodevelopmental risk. These findings helped prompt earlier regulatory warnings that prenatal exposure to the drug was associated with lower cognitive scores and structural birth defects, and they now serve as a reference point when researchers consider whether other, more common medications might exert subtler effects on brain development.
Acetaminophen has drawn more recent regulatory attention because it is so widely used and long considered relatively safe in pregnancy. On September 22, 2025, the Food and Drug Administration announced it had initiated a label change process for the drug, with a notice to clinicians describing possible links to autism and ADHD based on observational data. Yet the science here is far from settled. A nationwide Swedish cohort study covering approximately 2.48 million births with follow-up through 2021 found that small associations between prenatal acetaminophen and neurodevelopmental outcomes attenuated or disappeared in sibling comparisons, suggesting that shared genetics and family environment explained much of the signal. The National Institutes of Health summarized those findings by emphasizing that sibling-controlled designs reduce confounding and that, in this case, they revealed no convincing causal relationship. The gap between the FDA’s precautionary move and the Swedish data illustrates how regulators and researchers can reasonably reach different interim conclusions from overlapping evidence, leaving expectant parents and clinicians to navigate uncertainty rather than clear-cut guidance.
Folic Acid and Other Potentially Protective Factors
If certain exposures appear to raise risk, others may lower it. The Norwegian Mother and Child Cohort Study, tracking 85,176 births from 2002 to 2008, reported that mothers who used folic acid from four weeks before conception to eight weeks after their last menstrual period had an adjusted odds ratio of 0.61 for autistic disorder in their children, with a 95 percent confidence interval of 0.41 to 0.90, according to findings in a large prospective analysis. An Israeli birth cohort and case-control study following 45,300 births through 2015 pointed in the same direction, reporting a relative risk of roughly 0.39 for ASD among children whose mothers took folic acid before pregnancy, as documented in related follow-up research. Although the two studies differed in design and population, both suggested that periconceptional folic acid supplementation was associated with a meaningful, though not absolute, reduction in autism risk.
Broader syntheses have reinforced those patterns. A science update from the Eunice Kennedy Shriver National Institute of Child Health and Human Development noted that folic acid and multivitamin use before and during pregnancy may reduce autism risk, and a systematic review of maternal supplementation reached similar conclusions about modest protective effects. The National Institute of Environmental Health Sciences has highlighted folate as a nutrient of interest in autism research, noting that early intake of folic acid appears particularly important for families already at elevated risk. For expectant parents, the practical takeaway is that a supplement already recommended to prevent neural tube defects may confer additional neurodevelopmental benefits, even though no study has established a guaranteed preventive effect against autism or fully disentangled nutrition from other health behaviors.
Environmental Exposures Beyond the Medicine Cabinet
The prevention question extends well past pharmaceuticals into the air, water, and household environments that surround pregnant people and infants. A population-based case-control study in California linked state pesticide application records with maternal residential proximity and found associations between ambient pesticide exposures during pregnancy and infancy and autism spectrum disorder, with stronger associations in cases where ASD co-occurred with intellectual disability, according to analyses of nearby agricultural spraying. Separately, a nested case-control analysis within the Nurses’ Health Study II evaluated fine particulate matter exposure and identified pregnancy (especially the third trimester) as a key window when higher PM2.5 levels were associated with increased autism risk. Researchers have also examined maternal infections and fevers, with a prospective Norwegian cohort reporting that prenatal fever, particularly when recurrent in the second trimester, correlated with higher odds of ASD in offspring, hinting that immune activation may be one of several pathways linking environment and neurodevelopment.
Maternal metabolic health appears to play a role as well. A PRISMA meta-analysis of studies through April 2023 calculated a pooled odds ratio of about 1.48 for autism among children of mothers with type 2 diabetes, with similar or slightly higher estimates for gestational diabetes and significant heterogeneity across study designs. Obesity and metabolic syndrome have been implicated in several of the same cohorts, raising the possibility that systemic inflammation, altered placental function, or shared genetic risk contributes to the observed associations. At the same time, environmental health researchers caution that many of these factors cluster together: families living near heavy traffic or large-scale agriculture may face higher exposures to air pollution and pesticides while also contending with limited access to healthy foods and preventive care. Recent reporting on one peer-reviewed paper, which found that parents who deliberately reduced environmental exposures in their homes before conception had children with lower rates of autism diagnoses, underscored how preconception behaviors and exposures can be tightly intertwined with socioeconomic and cultural factors that are difficult to separate statistically.
Balancing Hope, Uncertainty, and Equity
The emerging science of modifiable autism risk sits at a fraught intersection of hope and uncertainty. On one hand, the valproate data show that certain prenatal exposures can substantially increase risk, and the folic acid literature suggests that targeted interventions may modestly lower it. On the other hand, many of the most discussed signals, such as those around acetaminophen or low-level pesticide exposure, come from observational studies vulnerable to confounding, measurement error, and publication bias. Scientists interviewed about this work have stressed, as reflected in coverage of environmental changes in families’ homes, that associations do not prove causation and that even strong statistical links typically translate into modest absolute risk differences at the population level. For most individual families, the likelihood of having a child on the spectrum remains driven primarily by genetics and chance, not by any single behavior or exposure.
Advocates and researchers also warn that a narrow focus on parental choices can inadvertently feed stigma or blame toward mothers, particularly those with fewer resources to avoid certain exposures. Tager-Flusberg and colleagues have emphasized that the goal of this research is not to scrutinize families’ behavior in retrospect but to identify practical, evidence-based steps, such as smoking cessation, diabetes management, folic acid supplementation, and reasonable limits on high-dose medications, that can be supported through public health systems. As regulators weigh precautionary actions like label changes and as new cohort findings accumulate, the challenge will be translating complex, sometimes conflicting data into guidance that is transparent about uncertainty, attentive to equity, and respectful of neurodiversity. For now, experts generally converge on a cautious middle ground: encourage proven prenatal health measures, reduce clearly harmful exposures where feasible, and continue rigorous research before declaring that autism, in any broad sense, can be prevented.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
