Income, Race-Ethnicity, and Dementia Risk Factors in the United States.

OBJECTIVES: To determine population attributable fractions (PAFs) of modifiable dementia risk factors by income, and independent and interactive associations of race-ethnicity and income with each risk factor and cumulative number of midlife risk factors. METHODS: This nationally representative cross-sectional study of the 1999-2008, 2011-2014, and primarily the 2015-2018 National Health and Nutrition Examination Surveys comprised of individuals aged 18-44, 45-64, and 65+ years for early-life, midlife, and late-life analyses, respectively. Income was operationalized using the poverty-income ratio. The primary outcomes were PAFs and prevalence ratios (PR) of the 13 individual-level dementia risk factors and cumulative number of midlife risk factors. RESULTS: There were 13,145 individuals with risk factor data between 2015 and 2018 (51.1% aged 18-44, 31.0% aged 45-64, and 17.9% aged ≥65 years; 51.5% female after survey-weighting). Higher income was associated with lower prevalence of each dementia risk factor except obesity, high LDL, and TBI. The highest PAF for those with incomes <100% of the federal poverty level was late-life vision loss (20.9%, 95% CI 16.8%-25.2%). Higher income was associated with lower number of midlife risk factors (PR: 0.91, 95% CI 0.89-0.94). Race-ethnicity categories historically-underrepresented in dementia studies were associated with midlife diabetes, obesity, physical inactivity, and late-life vision loss. DISCUSSION: Lower income and historically underrepresented race-ethnicity categories are associated with many dementia risk factors. Dementia prevention efforts may be more successful by targeting modifiable risk factors in these higher-risk populations.