The Malignancy of Dementia

OBJECTIVE: To evaluate the effect of dementing illnesses on the risk of dying, taking into account other conditions that would shorten survival. DESIGN: Five-year follow-up of community survey of dementia. SETTING: Five-year data were obtained for the 3531 subjects, aged 65 years and older, who participated in the 1987 population survey of dementia in Shanghai, China. MAIN OUTCOME MEASURE: Time to death. Relative risks of dying were calculated for demographic variables, dementia diagnoses based on findings of clinical evaluations, and 15 reported prevalent medical conditions using the proportional hazards model. RESULTS: In those subjects aged 65 to 74 years, the mortality risk ratio was 5.4 (95% confidence interval, 2.0 to 14.6) for Alzheimer's disease and 7.2 (95% confidence interval, 3.6 to 14.4) for vascular dementia. The risk ratio for Alzheimer's disease was similar to the mortality risk ratio for cancer (5.6 range, 2.9 to 10.9). In this age group, dementing illnesses were uncommon, and few deaths were therefore attributable to the dementing illnesses. In those subjects aged 75 years and older, the mortality risk ratios were 2.8 (95% confidence interval, 2.1 to 3.6) for Alzheimer's disease, 3.5 (95% confidence interval, 2.4 to 5.1) for vascular dementia, and 3.6 (95% confidence interval, 2.0 to 6.7) for "other dementias." Because these dementing disorders were common in those subjects aged 75 years and older, 23.7% of the risk of death could be attributed to these disorders. CONCLUSIONS: Both Alzheimer's disease and vascular dementias are truly malignant and constitute major risk factors for death in persons older than 75 years.