Association of Personal Wealth Indicators and Health Care Costs in Persons With Alzheimer's Disease

Abstract

ObjectivesThere is paucity of studies on the relationship between personal wealth and health care costs among persons with dementia, and earlier studies on other indicators of socioeconomic position have assessed costs after dementia diagnosis only. We investigated how different indicators of personal wealth (disposable income, supplementary income, assets subject to taxation, taxes and tax-like payments, and liabilities) are associated with health care costs in persons with Alzheimer's disease (AD) before and after AD diagnosis. DesignRegister-based nationwide cohort study of persons with AD. Setting and ParticipantsCohort of 70,531 people who received a clinically verified AD diagnosis in Finland between 2005 and 2011 and were community-dwelling at time of diagnosis. MethodsData on income indicators were obtained from Statistics Finland. Data on medication costs and hospital care costs for the 12-month period from 5 years before to 2 years after AD diagnosis were obtained from national registers. Associations of wealth indicators with costs were investigated with multivariate mixed-effect negative binomial regression. ResultsAfter adjustment for age, region, sex, marital status, comorbidities, expensive medications, use of psychotropic and antidementia medication, and highest occupational class before AD, people with higher levels of personal wealth indicators were more likely to have higher total health care costs along the whole follow-up period. The incidence rate ratios (IRRs), 95% CI for highest quintile vs lowest quintile were 1.17, 1.15−1.19 for disposable income, 1.10, 1.08−1.12 for taxable income, 1.18, 1.16−1.19 for supplementary income, 1.07, 1.06−1.09 for taxes, and 1.05, 1.04−1.07 for taxable wealth. Conclusions and ImplicationsOur observation on the association between income/wealth indicators and health care costs in a country with a strong public health care system call for more effective measures in targeting health inequalities in the aging population. Although the different indicators were not completely interchangeable, associations of different indicators were toward the same direction.