Weather, air pollution, and migraine: A case-time series analysis examining environmental exposures and transient health outcomes recorded via smartphone application

Background: Most studies of environmental exposures and migraine have been limited to weather or single-pollutant models and aggregate outcome measures. We estimated the associations between multiple environmental exposures and migraine events captured using a smartphone application (app) in the province of Ontario, Canada. Methods: We obtained Ontario-wide local daily weather and pollution estimates for 2017–2019 from Environment and Climate Change Canada. The Migraine Buddy app team provided records from research-consenting users. Associations between environmental exposures and migraine attack onsets were examined with case-time series lagged multipollutant models, accounting for demographic and temporal covariates. Results: We analyzed 69,808 migraine attacks reported by 7418 participants in Ontario, Canada. Overall cumulative estimates for 0–3 days suggested increased odds for migraine attacks with higher nitrogen dioxide (NO 2 ) and ozone (O 3 ), drops in barometric pressure, and cold winter temperatures. No association was observed for PM 2.5 . Positive associations with NO 2 and O 3 were strongest 1 and 2 days after exposures, while associations with barometric pressure changes and cold winter temperatures were strongest on the first day. Estimates for warmer summer temperatures suggested a complex temporal pattern, with positive same-day association followed by 3 days of negative associations. Conclusion: This study presents a novel examination of the associations between air pollution and migraine attacks using small-area exposures, smartphone data, and the case-time series method. Onset of migraine was associated with higher NO 2 and O 3 , colder winter weather, and positive and negative changes in barometric pressure.