Descriptive epidemiology of dementia in the US
Using routine data to uncover underlying trends presents challenges Population estimates of dementia have, in recent decades, become of national interest to policy makers, politicians and the wider community. In a linked study (doi:10.1136/bmj-2024-083034), Blass and colleagues report estimates of the incidence and prevalence of dementia in the US Medicare fee-for-service health insurance system for people aged 66 years and over.1 Consistent with findings from cross-generational cohort studies,23 they found that age standardised incidence of dementia declined from 3.5% in 2015 to 2.8% in 2021 and prevalence increased from 10.5% to 11.8% owing to population ageing. Clear differences existed between subgroups by race/ethnicity and deprivation, but interpreting these differences is not simple using this dataset alone. Studies that rely on routine data have both strengths and limitations. These include the nature, availability, and configuration of services, who accesses them and how, and
