Abstract
Geographical differences in Alzheimer's disease and related dementias (ADRD) prevalence may be driven by under-diagnosis due to insufficient healthcare access. Telehealth may improve ADRD detection in remote regions but relies on broadband service availability. We evaluated the influence of county-level broadband availability on Appalachian and rural variation of ADRD prevalence in six US Central Appalachian states and hypothesized that accounting for broadband access may accentuate Appalachian/non-Appalachian and rural/urban differences. An ecologic analysis evaluated county-level ADRD prevalence among the Medicare fee-for-service population across 591 Central Appalachian counties from 2015 to 2018. ADRD prevalence by Appalachian/non-Appalachian and rural/urban county-designation was estimated using negative binomial regression sequentially adjusting for age/education, diagnostic access, broadband presence/usage, poverty, and internet access/device ownership. Appalachian counties had lower ADRD prevalence than non-Appalachian counties in rural compared to urban counties (β(AppXRural) = -0.028; 95% Confidence Interval (CI): -0.052, -0.005). This variation attenuated with adjustment for broadband access (β(AppXRural) = -0.014; 95% CI: -0.038, 0.009). Broadband presence (β= -0.020; 95% CI: -0.032, -0.008) and a higher proportion of households with broadband in a county (β= -0.405; 95% CI: -0.534, -0.277) were negatively associated with ADRD prevalence after adjustment for poverty. Future research should implement alternative study designs to investigate mechanisms linking broadband access to ADRD prevalence.