Air pollution and the risk of multiple sclerosis relapses in the Paris region, the CONFISEP study

Background Air pollution has been proposed as a short-term trigger of multiple sclerosis (MS) relapses, but delayed effects remain poorly studied in the era of modern disease-modifying therapies (DMTs). We investigated whether weekly air pollution exposure was associated with delayed MS relapse risk up to 15 weeks. Methods We conducted a bicentric retrospective cohort study in the Île-de-France region (2018–2021) including adults with relapsing MS from the Observatoire Français de la Sclérose en Plaques registry. Weekly exposure to particulate matter (PM) 10 , PM 2.5 , nitrogen dioxide (NO 2 ) and ozone (O 3 ) was estimated using Airparif street-level modelling data. Associations (lags 0–15 weeks) were assessed using self-controlled case series and cohort analyses adjusted for time-varying DMTs, disease duration, number of relapses during the previous year, respiratory infections, temperature, sunlight exposure, lockdown and holiday periods and additionally for age, sex, deprivation index and urbanicity in cohort analyses. Results Among 1152 patients (73.6% women; median age 37 years), 898 relapses occurred (0.20/patient-year). In self-controlled analyses, PM 10 exposure was associated with increased relapse risk at 2-week (incidence rate ratio (IRR) 1.28, 95% CI 1.05 to 1.57) and 11-week lags (IRR 1.26, 95% CI 1.02 to 1.54), PM 2.5 exposure was associated with increased risk at 2-week (IRR 1.20, 95% CI 1.03 to 1.39), while NO 2 exposure was associated with increased risk at 2-week (IRR 1.21, 95% CI 1.02 to 1.44), 3-week (IRR 1.26, 95% CI 1.04–1.52) and 12-week lags (IRR 1.32, 95% CI 1.12 to 1.56). Cohort analyses showed similar but weaker associations for PM 10 and PM 2.5 . DMTs were strongly associated with lower relapse risk, with effect sizes greater than those observed for air pollution exposure. Conclusions Weekly exposure to PM 10 , PM 2.5 and NO 2 was associated with increased MS relapse risk over delayed time windows ranging from 2– 3 weeks to 11–12 weeks. These temporal windows should be considered exploratory given the multiple lag-specific analyses. DMTs remained the strongest determinants of relapse risk. Trial registration number NCT05033782 .