The hispanic paradox and predictors of mortality in an aging biethnic cohort of Mexican Americans and European Americans: The san antonio longitudinal study of aging

Objectives To examine predictors of mortality in aging Mexican Americans (MAs) and European Americans (EAs). Design Longitudinal, observational cohort study. Setting Socioeconomically diverse neighborhoods in San Antonio, Texas. Participants Community-dwelling adults aged 65 and older (394 MA; 355 EA) who completed the baseline examination (1992-96) of the San Antonio Longitudinal Study of Aging (SALSA) and for whom vital status was ascertained over an average 8.2 years of follow-up. Measurements Ethnic group was classified using a validated algorithm. Hazard ratios (HRs) for mortality were estimated using Cox proportional hazards models with age, sex, ethnic group, education, income, frailty, diabetes mellitus with and without complications, comorbidity, cognition, depressive symptoms, and body mass index included as predictors in sequential models. Results At baseline, MAs had a higher prevalence of diabetes mellitus and frailty and lower socioeconomic status (SES) than EAs. The age- and sex-adjusted ethnic HR (MA vs EA) for mortality was 1.54 (95% confidence interval (CI) = 1.17-2.03). After adjusting for SES, the ethnic HR was no longer significant (HR = 1.16, 95% CI = 0.83-1.61). In the final model, comorbidity, diabetes mellitus with complications, depressive symptoms, and cognitive impairment were significant independent risk factors for mortality. Conclusion Contrary to the Hispanic paradox, MAs were at greater risk of mortality than EAs. SES differences largely explained this ethnic disparity. Significant independent predictors of mortality, regardless of ethnic group, were diabetes mellitus with complications, comorbidity, depressive symptoms, and cognitive impairment. Mortality reduction in older MAs requires attention to socioeconomic disparities and disease factors.