Background/aim Only recent studies have explored the role of air pollution exposure on neurodegenerative processes. The aim of this study is to evaluate the association of long-term exposure to particulate matter, nitrogen oxides and summer ozone with the first hospitalisation for dementia (overall, vascular dementia, Alzheimer’s disease and senile dementia) in a large administrative cohort.

Methods We selected 3 50 872 subjects (free of dementia) aged 65–100 years at inclusion (21/10/2001) and we followed them until 31/12/2013. From the hospital discharge registry, we selected all subjects hospitalised for the first time with primary or secondary diagnoses of dementia (ICD-9CM: 46.1, 290.0–290.4, 294, 331.0, 331.1, 331.82). We estimated exposure at residence using land use regression models for nitrogen oxides and particulate matter, and a chemical dispersion model for summer ozone. We used Cox models (HR, hazard ratio) to estimate the association between air pollution exposure and first hospitalisation for dementia and its subtypes, adjusted for age, education, marital status, place of birth, socioeconomic position index, and stratified by gender.

Results During the follow-up we selected 21 105 first hospitalizations for dementia (7500 for vascular dementia, 7671 for Alzheimer and 7571 for senile dementia). Overall we observed a negative association between exposure to NO2 (10 μg/m³) and dementia hospitalisation (HR=0.97; 95% CI: 0.96 to 0.99) and a positive association with exposure to O3 (10 μg/m³), HR=1.06 (95% CI: 1.04 to 1.09).

Considering different type of dementia we found a positive and statistically significant association between exposure to NOx, NO2, PM2.5, PM10, and vascular dementia; whereas, a negative association between exposure to NOx, NO2, PM2.5, PM10, and Alzheimer’s disease was observed. Hospitalisation for senile dementia was positively associated with exposure to O3, HR=1.20 (95% CI: 1.15 to 1.24).

Conclusion Our results showed a positive association between exposure to O3 and hospitalizations for dementia and a negative association with NO2. In the separate analysis by type of dementia, exposure to all pollutants (except O3) was associated positively to vascular dementia and Ozone exposure was associated with senile dementia.