Five-year follow-up of cognitive impairment with no dementia

Context: The importance of early identification of dementia has prompted numerous investigations of mild cognitive impairment and the preclinical stages of progressive degenerative disorders. To date, there is limited information from large-scale studies regarding outcomes of persons specifically identified with cognitive impairment but no dementia (CIND). Objectives: To investigate outcomes for persons with no cognitive impairment (NCI) or CIND, focusing on its etiologic subcategories, from the Canadian Study of Health and Aging (CSHA) and to examine the predictive validity of a set of core features thought to be early manifestations of subsequent dementia. Design: Five-year, longitudinal follow-up of all persons without dementia examined during the first phase of the CSHA in 1991. Setting: Community and institutional settings. Participants: Population-based sample of 883 persons with NCI and 801 persons with CIND at the first phase of the CSHA. At follow-up, 517 persons with NCI (59%) and 327 persons with CIND (41%) were alive and received clinical diagnoses. Main Outcome Measures: Mortality, institutionalization, and clinical diagnoses using the Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition criteria for dementia. Results: Persons with CIND were more likely than those with NCI to die (49% vs 30%), to be admitted to facility care (29% vs 14%), or to receive diagnoses of dementia (47% vs 15%) at follow-up. Those subsequently diagnosed as having dementia were more likely to have had impaired memory, informant-reported change in memory, and functional impairment at baseline. Conclusions: Persons with CIND were more likely to have a negative outcome than persons with NCI during a 5-year interval. This was true across etiologic subcategories and suggests that the use of specific diagnostic criteria sets does not improve our identification of those who develop dementia compared with a broader, more. inclusive approach. More factors contributed to the prediction of all forms of dementia than to AD, but the most accurate prediction was for those who remained dementia free.