Mortality caused by dementia has mainly been examined in population-based studies relying on prevalent cases. This study aims to investigate the impact of incident dementia on mortality as well as to identify factors influencing the course of dementia and those predicting early death in demented individuals. A representative community sample of 1,692 individuals aged 75 years and over was examined by neuropsychological testing in a four-wave study. Data were analyzed with the Cox proportional hazards model after making necessary adjustments for potential covariates. At the third follow-up 51% of the incident demented and 19% of the participants without dementia had died. The mean survival time was 3.1 years (95% CI = 2.8-3.4) for the demented subjects and 4.0 years (95% CI = 3.9-4.0) for those without dementia (p < 0.001). In the total sample, the relative risk of dying after developing dementia was estimated to be 2.4 (95% CI = 1.6-3.6) with age, sex, education, co-morbidity, and institutionalization being taken into consideration. Those persons with incident dementia who died had a more severe dementia. Population-based studies relying on incident cases are especially valuable in describing course and outcome of dementia. Studies relying on prevalent cases and clinical samples tend to overestimate mortality and propose course-modifying factors that are challenged by studies relying on incident cases.
Copyright (c) 2006 S. Karger AG, Basel