Environment/Climate Change
Air pollution and migraine: Same-day and lagged associations of smartphone app attack records following environmental exposures Andrea Portt* Andrea Portt Erjia Ge Christine Lay Hong Chen Peter Smith
Background and aims: Although migraine affects over 1 billion people worldwide, little is known about its environmental triggers. While some research has observed an association between ambient air pollution and migraine events, most studies have been limited to single pollutants and/or relied on emergency-department visit data. Our objective was to estimate the associations between environmental exposures and migraine events captured using a smartphone app in the province of Ontario, Canada.
Materials and methods: Migraine Buddy is a well-established smartphone app with approximately 3 million users worldwide. Environment and Climate Change Canada provided exposure estimates for 2017-2019.
Previous work used statistical methods with pooled data. The case time series is a newly developed modeling technique that harnesses longitudinal individual-level data in relation to multiple environmental exposures Associations between environmental exposures and migraine events were estimated using the case time series method with lagged multi-pollutant models, accounting for demographic covariates.
Results: There were 14,526 migraine attacks reported by 2,989 research-consenting Migraine Buddy participants. On average, participants were more likely to report new migraine attacks on days with relatively higher temperatures or ozone, and on days with relatively large increases in barometric pressure. In contrast, participants were more likely to report new migraine attacks the day after relatively large negative changes in barometric pressure, and the day after relatively high nitrogen dioxide or particulate matter 2.5 exposure.
Conclusion: This is the first study to estimate associations of environmental exposures and migraine using smartphone data and the case time series. Our results demonstrate the application of smartphone data in epidemiologic research, and suggest that this neurologic health outcome may occur with different time lags depending on the exposure.