New research suggests that exposure to intense wildfire smoke during pregnancy may be associated with increased likelihood of autism in children. The study, by researchers at UC Davis Health and UCLA, was published in the journal Environment International.

The study of more than 8.6 million births in California is the largest to date examining how wildfire-specific air pollution may impact early neurodevelopment. Scientists combined detailed wildfire smoke data with state birth records from 2001 to 2019. They matched these with autism diagnoses from California Department of Developmental Services through 2022.

The researchers considered the timing and intensity of wildfire smoke exposure, as well as the type of background air pollution that families usually experienced. They focused on wildfire-related PM 2.5, or tiny particles of smoke about 30 times smaller than the diameter of human hair.

High‑intensity wildfire smoke events show strongest associations

Overall, average wildfire smoke levels during pregnancy showed only weak associations with autism diagnosis. But during high‑intensity smoke events — days with particulate levels in the top 10 percent — the associations grew stronger. In regions with generally cleaner air, such as rural areas, it was most pronounced.

“The most intense wildfire smoke exposure is where we see the clearest links,” said UC Davis Health’s Rebecca J. Schmidt, senior author and professor in the Department of Public Health Sciences and the UC Davis MIND Institute. “And these intense wildfire events are becoming more common in the West.”

The researchers estimated wildfire smoke exposure for three months before, during and for one year after pregnancy. They used advanced models of particulate matter that included sources like wildfire maps, air monitors and satellites and adjusted for wind currents and geography.

They separated particulate matter from vegetation burning and from wildland urban interface sources such as homes, vehicles and other materials.

“These models help us understand not just how much smoke was in the air, but where it came from,” said Schmidt, who is also affiliated with the UC Davis Environmental Health Sciences Center.

Among the key findings:

• Pregnancies in the top 10 percent of wildfire smoke exposure had a 6% higher likelihood of autism after adjusting for other kinds of air pollution.
• In areas with cleaner air, known as low background pollution, the link was much stronger. In these areas, pregnancies in the top 10 percent of wildfire smoke exposure had a 50% higher likelihood of autism.
• Very smoky days mattered. When wildfire smoke particles reached 35 micrograms per cubic meter or higher, a level considered very smoky, autism likelihood increased. Lower‑level smoky days did not show the same pattern.
• The strongest links were seen in non‑metropolitan areas, where there is generally less pollution from traffic and industry and smoke spikes stand out more.
• Smoke from wildland‑urban interface fires — where homes and buildings burn — was linked to higher autism likelihood even in large cities, possibly because these fires release more toxic particles.

The most intense wildfire smoke exposure is where we see the clearest links. And these intense wildfire events are becoming more common in the West.”—Rebecca Schmidt, professor, Department of Public Health Sciences

Environmental factors and neurodevelopment

Schmidt cautions that the findings do not show that wildfire smoke exposure causes autism. “This study complements existing research that shows environmental exposures during pregnancy can influence development,” she said.

The findings are in line with results from another recent study focused on Southern California. Exposure to intense wildfire smoke during pregnancy has also been associated with preterm delivery and increased birthweight.

Karl O’Sharkey, lead author and epidemiologist at UCLA, noted that the wildfire smoke findings align with earlier work by this team. That research linked prenatal exposure to traffic‑related pollutants like nitrogen dioxide, benzene and nickel to increased likelihood of autism identification.

“Together, these results suggest that air pollution during pregnancy is not a single, uniform exposure, and that the source and composition of pollution, as well as timing and intensity of exposure matter when it comes to neurodevelopment,” O’Sharkey explained.

The authors note that this research underscores the need for targeted policies such as vegetation management, land use planning, indoor filtration upgrades and clean-air shelters.

Study limitations include that fact that estimated smoke exposure relies on the home address at birth, and can’t account for moves, workplace exposures or time spent outdoors. In addition, the state Department of Developmental Services system doesn’t serve all autistic children, so the data are likely an undercount.

The study was funded by a $1.35 million grant from the U.S. Environmental Protection Agency.

Other authors include Beate Ritz, Sanjali Mitra and Ting Chow of UCLA; and Amanda J. Goodrich, Kathryn C. Conlon, Miriam Nuño, Deborah H. Bennett, Sean Raffuse, Mohammed Astaneh, Yusheng Zhao and Michael Kleeman of UC Davis.

Related Resources

Check real-time air quality index levels using AirNow or PurpleAir.

10 ways to protect your health when the air is smoky