Objectives: To examine the extent to which cognitive status and decline in cognitive status predict mortality in older Mexican Americans.
Design: Longitudinal cohort.
Setting: Older Mexican Americans residing in five southwestern states in the United States.
Participants: Two thousand six hundred twenty-five persons aged 65 and older living in Texas, New Mexico, Colorado, Arizona, and California.
Measurements: The cognitive function of participants was assessed using the Mini-Mental State Examination (MMSE) in 1993-94 and again in 1995-96. Cognitive decline was defined using two sets of criteria: a drop to 17 or less (moderate-severe cognitive impairment) on the MMSE at 2-year follow-up and a decrease of at least four points, a threshold change in scores that was used to predict mortality in this sample. Cox proportional hazards models were conducted to examine the association between the MMSE and increased risk for mortality, after controlling for sociodemographic characteristics, medical conditions, and depressive symptoms.
Results: Five-year mortality risk was significantly associated with persons in the moderately-severely cognitively impaired category (hazard ratio (HR) = 2.35, P <.001). Moreover, mild cognitive impairment was also predictive of mortality in older Mexicans (HR = 1.45, P <.001). Two-year declines in cognitive functioning, particularly in persons who had declined to the moderate-severe impairment category (HR = 2.23, P <.001) and those who dropped at least four points on the MMSE (HR = 1.30, P <.001), were predictive of mortality 3 years later, after controlling for important variables.
Conclusions: Baseline moderate-severe and mild cognitive status and 2-year decline in cognitive functioning independently predict mortality in older Mexican Americans. Although age and selected medical conditions have been reported as the more salient predictors of mortality, cognitive functioning should be considered part of identifying older persons at high risk for underlying medical conditions and mortality.