Autism prevalence following prenatal exposure to hurricanes and tropical storms in Louisiana.

The team tracked every hurricane and tropical storm that hit Louisiana. They mapped how close each pregnant woman lived to the worst weather. Later they counted how many of those babies were later diagnosed with autism.

They split storm exposure into three dose levels: mild, moderate, and severe. They also noted the week of pregnancy when the storm hit.

The harder the storm hit, the more autism cases showed up. The link was strongest when storms arrived between weeks 21 and 32 of pregnancy.

This dose-response pattern points to mid-to-late gestation as a sensitive window.

Beversdorf et al. (2005) asked moms to recall stressful events during pregnancy. Their surveys also peaked at 21–32 weeks, matching the hurricane window.

Taylor et al. (2017) went further. They showed that kids with autism who had two or more prenatal stressors had worse social and communication scores. This extends the Louisiana finding from "more cases" to "more severe symptoms."

Cryan et al. (1996) looked at classic obstetric complications and saw no extra autism risk. That null result seems to clash with the storm study, but the 1996 paper used broader medical charts, not a single intense stressor. The difference is likely what was measured, not a true conflict.

You can’t stop hurricanes, but you can spot the families who lived through them. When a child’s file shows prenatal storm exposure, flag the 21–32-week window and track social-communication milestones a little closer. Share this history with assessment teams; it adds context that standard checklists might miss.