Dissecting the racial/ethnic disparity in frailty in a nationally representative cohort study with respect to health, income, and measurement

Background: Racial/ethnic frailty prevalence disparities have been documented. Better elucidating how these operate may inform interventions to eliminate them. We aimed to determine whether physical frailty phenotype (PFP) prevalence disparities (i) are explained by health aspects, (ii) vary by income, or (iii) differ in degree across individual PFP criteria. Methods: Data came from the 2011 National Health and Aging Trends Study baseline evaluation. The study sample (n = 7,439) included persons in all residential settings except nursing homes. Logistic regression was used to achieve aims (i)–(iii) listed above. In (i), health aspects considered were body mass index (BMI) status and number of chronic diseases. Analyses incorporated sampling weights and adjusted for sociodemographic factors. Results: Comparisons are versus non-Hispanic whites: Non-Hispanic blacks (odds ratio [OR] = 1.46, 95% confidence interval [CI]: 1.21–1.76) and Hispanics (1.56, 1.20–2.03) continued to have higher odds of frailty after accounting for BMI status and number of chronic diseases. Non-Hispanic blacks had elevated odds of frailty in all income quartiles, including the highest (OR = 2.19, 1.24–3.397). Racial/ethnic disparities differed considerably across frailty criteria, ranging from a twofold increase in odds of slowness to a 25%–30% decrease in odds of self-reported exhaustion. Conclusions: BMI and disease burden do not explain racial/ethnic frailty disparities. Black–white disparities are not restricted to low-income groups. Racial/ethnic differences vary considerably by NHATS PFP criteria. Our findings support the need to better understand mechanisms underlying elevated frailty burden in older non-Hispanic black and Hispanic Americans, how phenotypic measures capture frailty in racial/ ethnic subgroups and, potentially, how to create assessments more comparable by race/ethnicity.