Abstract
OBJECTIVES: There is highly suggestive evidence for an effect of air pollution exposure on dementia-related outcomes, but evidence is not yet present to clearly pinpoint which pollutants are the probable causal agents. The aims of this study was to assess the longitudinal association between exposures of fine ambient particulate matter (PM2.5) from residential wood burning, and vehicle exhaust, with dementia. METHOD: We used data from the Betula study, a longitudinal study of dementia in Umeå, Northern Sweden. The study size was 1 806 and the participants were followed from study entry (1993-1995) to 2010. Modelled levels of source-specific fine particulate matter at the residential address were combined with information on wood stoves or wood boilers, and with validated data on dementia diagnosis and individual-level characteristics from the Betula study. Cox proportional hazards models were used to estimate Hazard Ratios (HRs) and their 95% CIs for dementia incidence (vascular dementia and Alzheimer's disease), adjusted for individual-level characteristics. RESULTS: The emission of PM2.5 from local residential wood burning was associated with dementia incidence with a hazard ratio of 1.55 for a 1 μg/m3 increase in PM2.5 (95% Confidence Interval (CI): 1.00-2.41, p-value 0.05). Study participants with an address in an area with the highest quartile of PM2.5 from residential wood burning and who also had a wood-burning stove were more likely to develop dementia than those in the lower three quartiles without a wood-burning stove with hazard ratios of 1.74 (CI: 1.10-2.75, p-value 0.018). Particulate matter from traffic exhaust seemed to be associated with dementia incidence with hazard ratios of 1.66, (CI: 1.16-2.39), p-value 0.006, and 1.41 (CI: 0.97-2.23), p-value 0.07, in the third and fourth quartiles, respectively. CONCLUSIONS: If the associations we observed are causal, then air pollution from residential wood burning, and air pollution from traffic, might be independent important risk factors for dementia.