Objective: To evaluate mean survival and to identify prognostic factors in a cohort of patients with Alzheimer's disease (AD). Design: Multicentric 9-year cohort analytic study. Setting: Seven neurology departments throughout Italy between April 1982 and January 1984. Patients: We recruited a consecutive sample of 145 patients affected by probable AD (Multicenter Italian Study on Dementia protocol, National Institute of Neurological Disorders and Stroke-Alzheimer's Disease and Related Disorders Association criteria). Five were misdiagnosed, and 21 could not participate in the longitudinal study. The clinicodemographic characteristics of the 119 enrolled patients (49 men, 70 women; mean age, 64.7 years; SD, 4.1 years; mean duration of disease, 3.1 years; SD, 1.8 years) did not differ from those of the 26 excluded patients. All underwent extensive cliniconeuropsychological testing every 6 months for at least 2 years until the patient died or our survey ended (April 30, 1991). Mean follow-up was 5.1 years (SD, 2.5 years). Main Outcome Measures: Death, severe functional impairment (a score greater than or equal to 7 on the Blessed Dementia Scale),and severe cognitive impairment (a score of less than or equal to 7 on the Information-Memory-Concentration Test). Results: Survival curves obtained by the Kaplan-Meier method indicated that (1) patients with early- and late-onset disease tie, before or after age 65 years) showed no difference either in relative survival or in time to reach predetermined functional and cognitive end points; (2) severely aphasic patients became profoundly demented significantly sooner than those with mild to moderate aphasia (P<.0001). Among clinicodemographic variables analyzed by a Cox model, severe language disability and functional loss proved to be the best predictors of death independent of age at onset or degree of dementia. Conclusions: Age at onset did not influence course and survival in AD. Severe aphasia appears to be the best predictor of death and unfavorable course.
