The impacts of air pollution on mortality and hospital readmission among Medicare beneficiaries with Alzheimer's disease and Alzheimer's disease-related dementias: a national retrospective cohort study in the USA

Background: Alzheimer's disease and Alzheimer's disease-related dementias (AD/ADRD) are prevalent neurodegenerative disorders, posing a critical worldwide public health challenge. Ambient air pollution has been identified as a potential risk factor for AD progression based on toxicological and epidemiological studies. We aimed to evaluate the impacts of air pollution—including fine particulate matter (PM_2·5), nitrogen dioxide (NO₂), summer ozone (O₃), and oxidant—on readmission or death among Medicare enrollees previously hospitalised with an AD/ADRD diagnosis code. Methods: We constructed a population-based nationwide retrospective cohort including all Medicare fee-for-service beneficiaries (aged ≥65 years) in the contiguous USA (2000–16) hospitalised with AD/ADRD, and followed them up from the year after their first hospitalisation until (1) year of death (mortality cohort) and (2) year of second hospitalisation for any cause (readmission cohort). We calculated annual average PM_2·5, NO₂, summer O₃, and oxidant concentrations for each individual at their residential ZIP code in each year after their first hospitalisation with AD/ADRD. We applied Cox proportional hazard models for the mortality and readmission cohorts stratifying on individual risk factors and adjusting for socioeconomic status, seasonal temperatures, and relative humidity. Findings: Our cohort consisted of 5 544 118 individuals, of whom 4 543 759 (82·0%) died and 3 880 894 (70·0%) were readmitted to the hospital during the study period. The average follow-up times were 3·34 years (SD 2·60) for the mortality cohort and 1·98 years (SD 1·65) for the readmission cohort. In both the mortality and readmission cohorts we found significant associations with each pollutant. For an IQR increase in NO₂, we found a hazard ratio (HR) for mortality of 1·012 (95% CI 1·009–1·015) and an HR for readmission of 1·110 (1·104–1·117). In the readmission cohort, we found an HR of 1·084 (1·079–1·089) for an IQR increase (3·87 μg/m³) in PM_2·5. The results slightly decreased in multi-pollutant models. The results of effect modification for mortality and readmission varied by pollutant, but higher risks were found among Black males and among those eligible for Medicaid in general. Interpretation: We provide new evidence that among a susceptible population with previous AD/ADRD-related hospitalisations, annual air pollution exposure since first hospitalisation is associated with risk of readmission and death. Funding: National Institute on Aging.