Prevalence and severity of cognitive impairment with and without dementia in an elderly population

Background: Not all cognitively impaired people have dementia, but those who do not meet current criteria for dementia have received little study. We report a comprehensive estimate of the prevalence of 'cognitive impairment, no dementia' (CIND) in an elderly population. Methods: The Canadian Study of Health and Aging gathered population representation information about elderly Canadians aged 65 and over from 36 cities and surrounding areas in five regions. In each region, the sample size was 1800 people in the community and 250 people in institutions. Patients in the community were screened for cognitive impairment by means of the modified mini-mental state examination. Those who scored below the cut-off point (n = 1106) and a randomly selected sample of those who scored above the cut-off point (n = 494) were referred for clinical examination. 59 individuals unable to take the screening test were also assessed clinically. We selected 17 long-term care institutions in each region, and then randomly selected consenting residents of these institutions for clinical assessment (n = 1255). Results: The prevalence of CIND was 16.8%, which was more than all types of dementia combined (8.0%). The prevalence of all types of cognitive impairment, including dementias, increased with age. Patients with CIND were three times more likely to be living in institutions than were cognitively unimpaired patients (odds ratio 3.1 [95% CI 2.4-3.9]). Circumscribed memory loss has a prevalence of 5.3% in the elderly Canadian population, and was the most common category. CIND was related to some degree of functional impairment in elderly patients. Interpretation: CIND is commonly associated with functional disability and a need for institutional care. This diagnostic category includes a costly group of disorders that merit further study.