What Environmental Factors Cause Autism?

Multiple prenatal and early-life environmental exposures have been linked to autism spectrum disorder, though genetics remains the dominant contributor. Large twin studies estimate the heritability of autism at roughly 90 percent or higher, which means environmental factors play a supporting rather than a starring role for most individuals.¹ Still, that remaining slice of risk is substantial across a population, and the exposures that fill it span air pollution, maternal health conditions, nutrient deficiencies, certain medications, and more. Importantly, most of these factors involve the prenatal period, and none of them operate like a simple on-off switch.

Air Pollution During Pregnancy

Of all the environmental exposures studied, fine particulate matter (PM2.5) in ambient air has some of the most consistent evidence linking it to autism risk. A nested analysis within the Nurses’ Health Study II found that higher PM2.5 exposure during pregnancy was associated with about a 60 percent increase in the odds of autism in offspring, and that the association held even after accounting for exposure before and after pregnancy.² That study pointed to the third trimester as a particularly sensitive window. A separate large cohort study, however, found the strongest PM2.5 associations during the first two trimesters, with the link decreasing linearly toward birth.³

The exact timing of vulnerability appears to vary across studies, which is not surprising given that brain development is a rolling process. What remains consistent is that pregnancy itself is the critical period, not the months before or after. A population-based case-control study also reported strong positive associations between PM2.5 and autism, with the highest risk during gestational months four through seven.⁴

One pattern that keeps appearing is a sex difference. Boys seem more susceptible to the effects of prenatal air pollution. In the cohort study that tracked weekly PM2.5 exposures, the association with autism was clear in boys but smaller and not statistically significant in girls.⁵ Given that boys are diagnosed with autism at roughly four times the rate of girls, this sex-dependent vulnerability is an active area of research.

Maternal Infections and Immune Activation

The idea that a mother’s immune response during pregnancy can affect fetal brain development has decades of evidence behind it. Epidemiological data links maternal infections during pregnancy to a higher risk of autism in children, and animal studies have shown that activating the maternal immune system alone, without an actual pathogen reaching the fetus, is enough to produce lasting brain changes and autism-like behaviors in offspring.⁶ The mechanism appears to involve inflammatory molecules crossing the placenta and interfering with how the fetal brain wires itself.

This is not about one specific infection. Researchers have found similar patterns with influenza, rubella, and even bacterial infections. The key variable seems to be the intensity and timing of the mother’s immune response rather than the identity of the bug. Animal models in mice, rats, and nonhuman primates all converge on the same conclusion: the maternal immune response is the critical link between prenatal infection and altered brain development.⁷ Genetic susceptibility, the timing and severity of the infection, and postnatal stressors all influence whether the outcome looks like autism, schizophrenia, or another neurodevelopmental condition.

Prenatal psychological stress also fits within this broader framework. Stress activates the maternal immune and hormonal systems, and several lines of research suggest that severe prenatal stress is associated with increased autism risk and with biological markers, like neuroinflammation and immune irregularities, that are elevated in autistic individuals.

Maternal Obesity and Diabetes

A mother’s metabolic health before and during pregnancy has emerged as a meaningful risk factor, though the picture has some nuance. In a large study, prepregnancy obesity and pregestational diabetes were each individually associated with higher odds of autism in offspring. But the real jump in risk came when the two conditions appeared together: mothers who had both obesity and pregestational diabetes had roughly four times the risk compared to mothers without either condition.⁸ The combination of obesity and gestational diabetes also tripled the risk.

Interestingly, paternal weight may matter too. A Scandinavian study found that maternal obesity alone was only weakly associated with autism risk, while paternal obesity was linked to higher rates of autistic disorder and Asperger’s disorder.⁹ The mechanism behind paternal obesity’s influence is not fully understood, but one hypothesis involves epigenetic changes to sperm caused by metabolic dysfunction. Fathers’ environmental exposures and lifestyle factors before conception can alter the chemical tags on their DNA, changes that are potentially passed to offspring.¹⁰

Prenatal Vitamins as a Protective Factor

Not every environmental influence raises risk. Some lower it, and prenatal folic acid and multivitamin supplementation are the clearest examples. An umbrella review that pooled results across multiple meta-analyses found that mothers who took folic acid or multivitamins during pregnancy had about a 30 percent lower risk of having a child diagnosed with autism compared to mothers who did not.¹¹ A separate meta-analysis found that folic acid supplementation specifically during early pregnancy was associated with roughly a 40 percent reduction in autism risk, and that a daily intake of at least 400 micrograms appeared to be the threshold for benefit.¹²

Timing matters. Data from a large Israeli cohort showed that women who took folic acid or multivitamins before pregnancy had significantly lower autism risk in their children compared to women who started only during pregnancy, and both groups fared better than those who took no supplements at all.¹³

Vitamin D tells a similar story. In a large nested case-control study, mothers with deficient vitamin D levels during pregnancy had about 44 percent higher odds of having a child with autism compared to mothers with sufficient levels.¹⁴ A Dutch cohort study (the Generation R Study) found that children born to mothers who were vitamin D deficient at mid-gestation scored higher on a measure of autism-related social difficulties.¹⁵ Neither folic acid nor vitamin D is a guaranteed shield, but the consistency of these findings across different populations and study designs is striking.

Medications During Pregnancy

Certain medications taken during pregnancy carry documented autism risk, and the most thoroughly studied is valproate, an anti-seizure drug also used for bipolar disorder and migraine prevention. A Danish population study found that children exposed to valproate in the womb had roughly three times the odds of being diagnosed with an autism spectrum disorder compared to unexposed children, and over five times the odds of childhood autism specifically, even after adjusting for the mother’s epilepsy.¹⁶

A more recent study in the New England Journal of Medicine confirmed this, reporting a hazard ratio of about 2.7 for autism with prenatal valproate exposure compared to no antiseizure medication, while two other commonly prescribed antiseizure drugs, topiramate and lamotrigine, showed no elevated risk.¹⁷ The implication is that valproate’s effect on brain development is specific to its chemistry, not a general consequence of treating epilepsy during pregnancy. Most clinical guidelines now recommend avoiding valproate in women of childbearing age when alternatives are available.

Pesticides, Metals, and Industrial Chemicals

Organophosphate pesticides have drawn attention because of their known effects on the developing nervous system. One study of children living near agricultural areas found that higher prenatal organophosphate exposure (measured through metabolites in the mother’s urine) was associated with poorer social behavior in offspring at age 14, though the researchers did not find a clear link between simply living near pesticide-sprayed fields and autism-related traits.¹⁸ Another study in a high-risk cohort found that organophosphate metabolite concentrations during pregnancy were not associated with increased autism risk when boys and girls were analyzed together, though there was a suggestion of elevated risk for girls exposed to one specific metabolite.¹⁹ The evidence here is mixed enough that researchers describe pesticides as a plausible risk factor that has not yet been proven.

Heavy metals sit in a similar spot. A review of the literature found that both the presence of certain toxic metals (like lead and mercury) and the absence of certain essential metals (like zinc and iron) have been flagged as autism risk factors. The two may be connected: toxic metals can compete with essential metals for the same biological pathways, so exposure to one can functionally create a deficiency of the other.²⁰

Per- and polyfluoroalkyl substances (PFAS), sometimes called “forever chemicals,” are widespread in the environment, and their role in autism is genuinely unclear. One study of high-risk families (the MARBLES study) found that certain PFAS compounds, especially PFNA and PFOA, were associated with increased autism risk.²¹ But a larger case-control study using prenatal blood samples found the opposite: PFOA and PFOS concentrations were actually lower in mothers of children later diagnosed with autism, and the overall results did not support a positive link between PFAS and autism.²² A third study focused on autism-related traits found that one PFAS compound (PFNA) showed a modest association, while the overall PFAS mixture did not.²³ The contradictory results may reflect differences in study populations, timing of measurement, or confounders that are hard to control for.

Phthalates and bisphenol A (BPA), the common endocrine disruptors found in plastics and consumer products, have been the subject of persistent concern. The evidence so far is mostly reassuring. A recent study measuring both maternal and paternal preconception exposures and maternal pregnancy exposures found no association between phthalate or BPA concentrations and autistic behaviors in children.²⁴ An earlier study found mostly null results as well, with one exception: higher prenatal exposure to a specific phthalate metabolite was linked to increased autism symptoms in boys only.²⁵ Overall, the case against these particular chemicals as autism risk factors is weaker than many people assume.

Parental Age

Advanced parental age, particularly paternal age, is one of the more consistent demographic risk factors. The most intuitive explanation involves the accumulation of new genetic mutations in sperm over time: unlike eggs, which are formed before a woman is born, sperm-producing cells divide continuously throughout a man’s life, introducing new copying errors with each round.²⁶ Epigenetic changes in aging sperm, like altered patterns of chemical tags on DNA, could be an additional pathway.

But the picture is more complex. A large international study found that the risk of autism increased not just with advancing paternal or maternal age, but also when the age gap between parents was ten years or more, regardless of which parent was older.²⁷ That pattern suggests that the mutation-accumulation theory cannot explain everything. Some researchers have proposed that assortative mating patterns, fertility treatments more common in older parents, or social factors associated with large age gaps also play a role.

Birth Complications and Oxygen Deprivation

Events around the time of birth, particularly those involving reduced oxygen supply, have been linked to autism in several studies. One found that autistic children were significantly more likely to have been exposed to at least one hypoxic condition during the prenatal period compared to neurotypical children.²⁸ A population-level analysis found a modest excess risk of autism among boys who experienced oxygen deprivation at birth, though the effect was not enormous.²⁹

One complication with this evidence is the direction of causation. Babies who are already developing atypically in the womb may be more likely to experience a difficult delivery, meaning that the birth complication is a consequence of atypical development rather than a cause. The research so far has not fully disentangled these possibilities, and both directions likely contribute in different cases.

How Genes and Environment Interact

Framing autism risk as “genes versus environment” misses the point. The emerging picture is that environmental exposures act on individuals who are already genetically vulnerable, and the same exposure may have very different effects depending on someone’s genetic background. One leading hypothesis is that environmental chemicals alter DNA methylation, the system of chemical markers that controls which genes are active in developing brain cells. These changes to gene regulation could push neurodevelopment off course in someone whose genetic starting point already puts them near a tipping point.³⁰

A related framework, the “redox/methylation hypothesis,” proposes that environmental toxins trigger oxidative stress in genetically susceptible individuals, which in turn disrupts the methylation processes needed for normal neural communication.³¹ This gene-environment interplay is also why identifying specific environmental culprits has been so difficult. The effect of a given chemical exposure likely varies depending on the genetic profile of the exposed individual, making it hard to detect in studies that lump everyone together.³²

The gut microbiome adds yet another layer. Research has shown that the composition of gut bacteria differs between autistic and non-autistic individuals, and that early-life factors like mode of delivery (vaginal versus cesarean), breastfeeding, and antibiotic use influence the gut microbiome in ways that may be relevant to brain development via the gut-brain axis.³³ Whether gut microbiome differences are a cause, a consequence, or a mediator of autism remains an open question.

The Vaccine Question

Vaccines are not an environmental cause of autism. This deserves a direct statement because it remains one of the most persistent misconceptions in public health. The original 1998 paper that claimed a link was retracted for fraud, and the decades of research that followed have been overwhelming. Across numerous large, well-designed studies, vaccines have not been shown to increase the risk of autism.³⁴ The misconception persists in part because autism is typically diagnosed around the same age children receive several routine vaccinations, creating a coincidence of timing that can feel causal to parents. But timing is not causation, and the evidence against a link is among the strongest in all of epidemiology.

Green Space and Neighborhood Environment

A growing body of work has started looking beyond individual chemical exposures to broader features of the environment where pregnant women live. One of the more intriguing findings involves green space. A study of a national Medicaid cohort found that higher greenness around a mother’s residence during pregnancy was associated with about a 17 percent lower risk of autism in offspring.³⁵ The protective effect was strongest for Black and Hispanic children and for those in urban areas. A separate study found that higher greenness during the first trimester specifically was linked to roughly a 23 percent reduction in autism odds, while exposure during other trimesters was weaker.³⁶

The effect sizes are modest, and a third population-based birth cohort found smaller, borderline associations between greenness and autism.³⁷ How green space might protect against autism is unclear. Reduced air pollution exposure is one pathway. Lower maternal stress is another. Physical activity, social interaction, and noise reduction could all contribute. The real takeaway may be less about trees per se and more about the cumulative environmental quality of the places where pregnancies unfold, which circles back to the broader theme running through all of this research: autism risk is not driven by any single environmental villain, but by the total burden of harmful exposures and the absence of protective ones acting on a genetically shaped foundation.

1
JAMA Psychiatry. Etiology of Autism Spectrum Disorders and Autistic Traits Over Time ↩︎
2
Environmental Health Perspectives. Autism Spectrum Disorder and Particulate Matter Air Pollution before, during, and after Pregnancy: A Nested Case–Control Analysis within the Nurses’ Health Study II Cohort ↩︎
3
PubMed Central. Prenatal Exposure to Air Pollution and Autism Spectrum Disorder: Sensitive Windows of Exposure and Sex Differences ↩︎
4
PubMed Central. Maternal exposure to fine particulate matter and autism spectrum disorder in children: population based case-control study ↩︎
5
PubMed Central. Prenatal Exposure to Air Pollution and Autism Spectrum Disorder: Sensitive Windows of Exposure and Sex Differences ↩︎
6
PubMed Central. Maternal immune activation: Implications for neuropsychiatric disorders ↩︎
7
PubMed Central. Maternal Immune Activation and Autism Spectrum Disorder: From Rodents to Nonhuman and Human Primates ↩︎
8
PubMed Central. The Association of Maternal Obesity and Diabetes With Autism and Other Developmental Disabilities ↩︎
9
PubMed Central. Parental obesity and risk of autism spectrum disorder ↩︎
10
SpringerLink. Multifaceted paternal exposures before conception and their epigenetic impact on offspring ↩︎
11
PLoS One. The association between maternal prenatal folic acid and multivitamin supplementation and autism spectrum disorders in offspring: An umbrella review ↩︎
12
PubMed Central. Prenatal Folic Acid Supplements and Offspring’s Autism Spectrum Disorder: A Meta-analysis and Meta-regression ↩︎
13
JAMA Psychiatry. Association of Maternal Use of Folic Acid and Multivitamin Supplements in the Periods Before and During Pregnancy With the Risk of Autism Spectrum Disorder in Offspring ↩︎
14
PubMed Central. Maternal Vitamin D Levels During Pregnancy and Offspring Autism Spectrum Disorder ↩︎
15
Molecular Psychiatry. Gestational vitamin D deficiency and autism-related traits: the Generation R Study ↩︎
16
PubMed Central. Prenatal valproate exposure and risk of autism spectrum disorders and childhood autism ↩︎
17
PubMed Central. Risk of Autism after Prenatal Topiramate, Valproate, or Lamotrigine Exposure ↩︎
18
PubMed Central. Prenatal Organophosphate Pesticide Exposure and Traits Related to Autism Spectrum Disorders in a Population Living in Proximity to Agriculture ↩︎
19
PubMed Central. Prenatal exposure to organophosphate pesticides and risk of autism spectrum disorders and other non-typical development at 3 years in a high-risk cohort ↩︎
20
PubMed Central. Metal Profiles in Autism Spectrum Disorders: A Crosstalk between Toxic and Essential Metals ↩︎
21
PubMed Central. Prenatal exposure to per- and polyfluoroalkyl substances in association with autism spectrum disorder in the MARBLES study ↩︎
22
PubMed Central. Prenatal Maternal Serum Concentrations of Per- and Polyfluoroalkyl Substances in Association with Autism Spectrum Disorder and Intellectual Disability ↩︎
23
PubMed Central. Prenatal exposure to per- and polyfluoroalkyl substances and childhood autism-related outcomes ↩︎
24
PubMed. Associations of maternal and paternal preconception and maternal pregnancy urinary phthalate biomarker and bisphenol A concentrations with offspring autistic behaviors: The PEACE study ↩︎
25
PubMed Central. Prenatal phthalate exposures and autism spectrum disorder symptoms in low-risk children ↩︎
26
JAMA Pediatrics. Maternal and Paternal Age and Risk of Autism Spectrum Disorders ↩︎
27
Molecular Psychiatry. Autism risk associated with parental age and with increasing difference in age between the parents ↩︎
28
PubMed Central. Prenatal exposure to hypoxic risk conditions in autistic and neurotypical youth: Associated ventricular differences, sleep disturbance, and sensory processing ↩︎
29
PubMed Central. Autism spectrum disorders and fetal hypoxia in a population-based cohort: Accounting for missing exposures via Estimation-Maximization algorithm ↩︎
30
Environmental Epigenetics. DNA methylation: a mechanism linking environmental chemical exposures to risk of autism spectrum disorders? ↩︎
31
NeuroToxicology. How environmental and genetic factors combine to cause autism: A redox/methylation hypothesis ↩︎
32
PubMed Central. Gene×environment interactions in autism spectrum disorders ↩︎
33
PubMed Central. Role of Gut Microbiome in Autism Spectrum Disorder and Its Therapeutic Regulation ↩︎
34
PubMed Central. The myth of vaccination and autism spectrum ↩︎
35
PubMed Central. Preconception, prenatal and early childhood exposure to green space and risk of neurodevelopmental delays: a national cohort study among Medicaid enrollees ↩︎
36
Journal of Exposure Science & Environmental Epidemiology. Association between greenspace exposure before, during, and after pregnancy and autism spectrum disorder in offspring ↩︎
37
PubMed. The association between prenatal greenspace exposure and Autism spectrum disorder, and the potentially mediating role of air pollution reduction: A population-based birth cohort study ↩︎