Chen · Environmental epidemiology (Philadelphia, Pa.) 2026 · time-stratified case-crossover study · n=6,000,000 ED visits

Short-term associations between ambient air pollution and emergency department visits for Parkinson disease, multiple sclerosis, migraine, and seizure in California, 2005-2018.

Cited 1 times in the scientific literature.

Level 4 - case-series / case-control

Time-stratified case-crossover observational study

PubMed 41450755 · doi:10.1097/EE9.0000000000000444 · record verified 2026-08-27

What was done

Researchers evaluated short-term associations between emergency department (ED) visits for four neurological conditions (Parkinson disease, multiple sclerosis, migraine, and seizure) and three ambient air pollutants: fine particulate matter (PM2.5), carbon monoxide (CO), and nitrogen dioxide (NO2). Using a time-stratified case-crossover design with conditional quasi-Poisson regression, daily estimated pollutant exposures modeled at 12 km resolution were aggregated to the ZIP-code level and linked to 6 million ED visits in California from 2005 to 2018. Risk ratios (RRs) were evaluated across lag days per interquartile range increase in exposure.

What was found

For Parkinson disease, CO showed the strongest short-term association at lag 0 (RR = 1.007, 95% CI: 1.003, 1.011). No associations were observed for multiple sclerosis. For migraines, same-day PM2.5 (RR = 1.003, 95% CI: 1.001, 1.005), CO (RR = 1.006, 95% CI: 1.003, 1.008), and NO2 (RR = 1.009, 95% CI: 1.005, 1.015) were positively associated with ED visits. For seizures, same-day NO2 showed the strongest effect (RR = 1.006, 95% CI: 1.003, 1.010). Multi-day cumulative effects decreased for CO and NO2 but remained similar for PM2.5.

Why it matters

This study provides large-scale US evidence that acute increases in ambient air pollutants are associated with modest same-day increases in emergency department visits for migraines, Parkinson disease, and seizures.

Limits

The relative risk increases are very small (all point estimates between 1.003 and 1.009), leaving findings susceptible to residual confounding. Exposure was estimated at a 12 km grid resolution and aggregated to ZIP codes rather than measured individually, introducing potential exposure misclassification. The abstract does not report visit counts broken down by specific diagnosis.

Cited by