Abstract
Objective
Life expectancy varies between 3-12 years following the diagnosis of Alzheimer’s disease (AD) and is an important clinical question for patients and families. Current literature is limited by relatively small sample sizes and a reliance on clinical diagnoses. This study sought to evaluate predictors of AD life expectancy in a large autopsy-confirmed sample.
Methods
Baseline data from individuals 50 years and older clinically and neuropathologically diagnosed with AD (N=764) were obtained from the National Alzheimer’s Coordinating Center. Life expectancy was calculated in months from AD diagnosis to death. Nineteen variables (demographic, medical/health, disease severity, and psychiatric) obtained at dementia diagnosis were examined. Variables that showed significant differences in life expectancy using t-tests and Pearson correlations (14 of 19) were then entered into a forward multiple regression.
Results
Seven predictors in the model explained 27% of the variance in life expectancy (F= 40.7, R-squared= 0.267). Lower MMSE scores (β= 0.339, p < .001), male sex (β= -0.144, p < .001), older age (β= -0.130, p < .001), non-Hispanic Caucasian race/ethnicity (β= 0.115, p < .001), greater impairment on the Functional Activities Questionnaire (β= -0.091, p=.042), abnormal neurological/physical exam (β= -0.083, p=.011), and higher Neuropsychiatric Inventory Questionnaire total scores (β= -0.079, p=.016) predicted shorter life expectancy.
Conclusions
Global cognitive impairment, sex, age, race/ethnicity, functional impairment, abnormal neurological exam findings, and psychiatric symptoms explain a significant proportion of life expectancy following an AD diagnosis. Future studies should explore the relationship between life expectancy, specific neurological abnormalities, and psychiatric symptoms. These 7 predictors could potentially be used to predict life expectancy in individuals diagnosed with AD.