New research suggests that exposure to certain synthetic chemicals during early pregnancy may be linked to a higher likelihood of autistic traits in children. The study found that children whose mothers had higher levels of a chemical called perfluorooctanoic acid (PFOA) in their blood during early pregnancy were more likely to show signs of social and communication difficulties by age four. Additionally, another chemical, perfluorohexane sulfonate (PFHxS), was more strongly associated with autistic traits in children who had a higher inherited risk for autism.
Per- and polyfluoroalkyl substances, or PFAS, are a large group of human-made chemicals used in a wide variety of consumer and industrial products, including non-stick cookware, stain-resistant fabrics, firefighting foams, and food packaging. Because they do not easily break down in the environment or in the human body, PFAS are often referred to as “forever chemicals.” These compounds can persist in the body for years and have been linked to various health risks. Due to their persistence and potential harm, some PFAS have been banned or restricted in many countries. Still, production and exposure remain widespread, especially in developing countries such as China.
In the United States, PFAS have become a growing focus of environmental and public health regulation. In April 2024, the Environmental Protection Agency under the Biden administration finalized the first legally enforceable national limits for six PFAS compounds in drinking water, including strict limits of 4 parts per trillion for PFOA. However, in May 2025, the Trump administration rolled back these standards by delaying enforcement for PFOA and repealing limits for four other PFAS.
Previous research has raised concerns that exposure to PFAS during pregnancy could interfere with a child’s brain development. Animal studies and limited human research have shown that PFAS may alter hormone levels, disrupt neurotransmitters, and interfere with normal signaling in the brain. Yet the human evidence linking PFAS exposure to autism-related traits has been inconsistent. Some studies suggest a connection, while others do not. Many of these earlier studies were conducted in populations where PFAS exposure was relatively low due to stricter regulations.
The new study, published in the Journal of Hazardous Materials, sought to fill some of the gaps in current understanding by analyzing a large sample of mother-child pairs in China, where PFAS levels are typically higher. The researchers focused on whether prenatal exposure to these chemicals might increase the likelihood of autism-related behaviors in children, and whether a child’s genetic makeup influenced the effect of this exposure.
The study used data from 1,610 mother-child pairs in the Shanghai Birth Cohort, a long-term study that follows children from early pregnancy through childhood. Blood samples were collected from mothers during the first trimester of pregnancy, and concentrations of ten different PFAS were measured. These included PFOA, PFHxS, and several other commonly detected compounds. When the children reached age four, their social and communication skills were assessed using a parent-completed questionnaire called the Social Responsiveness Scale (short form), which measures behaviors often associated with autism.
The researchers also calculated a polygenic risk score for autism, based on the presence of specific genetic variants known to be linked to autism. This score allowed the researchers to assess how a child’s genetic risk for autism might interact with environmental exposures like PFAS.
After adjusting for other factors that could influence child development—such as maternal age, education level, and body mass index—the researchers found that higher levels of PFOA during early pregnancy were associated with a three-fold increase in the odds of a child scoring in the range associated with autistic traits. However, among mothers who reported taking folic acid supplements before pregnancy, the association between PFOA and autistic traits appeared to be weaker.
In addition, the study found a potential interaction between PFHxS and genetic risk. Children with both higher prenatal exposure to PFHxS and a higher polygenic risk score were more likely to show signs of autistic traits. This interaction suggests that children who are already genetically more susceptible to autism may be more vulnerable to environmental exposures like PFAS.
When the researchers analyzed the effects of the entire PFAS mixture as a group, they did not find a clear association with autistic traits. This suggests that the risks may be driven by specific compounds like PFOA and PFHxS, rather than PFAS exposure in general.
The study also included a number of additional analyses to test the reliability of the results. These included separating the sample by child sex and by whether mothers took folic acid supplements. Interestingly, the researchers found that some PFAS appeared to have different associations with autistic traits in boys versus girls. For example, one chemical, PFDoA, was associated with fewer autistic traits in boys but not in girls. Another compound, PFBS, was linked to increased risk in boys but reduced risk in girls. These findings point to possible sex-specific effects that need more investigation.
The authors note that the study has several strengths, including a relatively large sample size, prospective design, and the ability to account for both genetic and environmental factors. However, they also caution that there are important limitations.
One limitation is that the measure of autistic traits relied on parent reports, which can be influenced by subjective perceptions. Also, only 24 children in the sample had scores in the range that suggested possible autism-related traits, which may limit the ability to detect small effects or to generalize the findings. In addition, the genetic risk score was based on data from mostly European populations, which may not perfectly reflect the genetic architecture of autism in a Chinese population. Differences in how genetic variants are inherited across populations could affect how well the polygenic risk score captures true genetic susceptibility.
Another limitation is that the PFAS levels in this study were relatively high compared to studies in Western countries. For example, median PFOA levels in this Shanghai sample were roughly three times higher than those reported in the United States. This means that the findings may not apply to populations with lower exposure levels. Also, the researchers could not rule out the influence of other environmental chemicals, such as heavy metals or plasticizers, although these are not strongly correlated with PFAS.
The study also raises new questions about the potential protective role of folic acid supplementation. While the researchers observed that pre-pregnancy supplementation appeared to lessen the association between PFOA and autistic traits, the biological reasons for this are not yet understood. Folic acid plays an important role in brain development, and previous studies have linked adequate folate intake to reduced risk of developmental disorders. It is possible that folic acid helps counteract some of the negative effects of PFAS on fetal brain development, but more research is needed to clarify the mechanisms involved.
Overall, the findings contribute to a growing body of evidence suggesting that early-life exposure to environmental chemicals like PFAS may have lasting effects on child development. The study highlights the importance of considering both environmental and genetic factors in understanding neurodevelopmental outcomes. While the results do not prove that PFAS cause autism or autistic traits, they suggest that certain exposures during pregnancy may increase the risk, particularly in children who are genetically more susceptible.
The study, “Prenatal exposure to per- and polyfluoroalkyl substances, genetic factors, and autistic traits: Evidence from the Shanghai birth cohort,” was authored by Yun Huang, Weiran Chen, Yuexin Gan, Xin Liu, Ying Tian, Jun Zhang, and Fei Li.