The joint association of physical activity and fine particulate matter exposure with incident dementia in elderly Hong Kong residents

•Habitual physical activity is negatively related to incident dementia.•PM2.5 exposure is positively associated with incident dementia.•Benefits of physical activity remain in people exposed to various levels of PM2.5.•Physical activity is still recommended to people living in relatively polluted areas. The evidence for the beneficial effects of physical activity (PA) and potentially detrimental effects of long-term exposure to fine particulate matter (PM2.5) on neurodegeneration diseases is accumulating. However, their joint effects remain unclear. We evaluated joint associations of habitual PA and PM2.5 exposure with incident dementia in a longitudinal elderly cohort in Hong Kong. A total of 57,775 elderly participants (≥65 years) without dementia were enrolled during 1998–2001 and followed up till 2011. Their information on PA and other relevant covariates were collected at baseline (1998–2001) by a standard self-administered questionnaire, including PA volumes (high, moderate, low, and inactive) and types (aerobic exercise, traditional Chinese exercise, stretching exercise, walking slowly, and no exercise). Their annual mean PM2.5 exposures at the residential address were estimated using a satellite-based spatiotemporal model. We then adopted the Cox proportional hazards model to examine the joint associations with the incidence of all-cause dementia, Alzheimer’s diseases, and vascular dementia on additive and multiplicative scales. During the follow-up period, we identified 1,157 incident cases of dementia, including 642 cases of Alzheimer’s disease and 324 cases of vascular dementia. A higher PA level was associated with a lower risk of incident all-cause dementia (hazard ratio (HR) for the high-PA volume was 0.59 (95% CI, 0.47, 0.75), as compared with the inactive-PA), whereas a high level of PM2.5 was related to the higher risk with an HR of 1.15 (95%CI: 1.00, 1.33) compared with the low-level of PM2.5. No clear evidence was observed of interaction between habitual PA (volume and type) and PM2.5 inhalation to incident dementia on either additive or multiplicative scale. Habitual PA and long-term PM2.5 exposure were oppositely related to incident dementia in the Hong Kong aged population. The benefits of PA remain in people irrespective of exposure to air pollution.