A Canadian cohort study of cognitive impairment and related dementias (ACCORD): study methods and baseline results

The overall objective of the Canadian Collaborative Cohort of Related Dementias (ACCORD) study is to describe the diagnostic distribution, natural history and treatment outcomes of individuals referred from the community to dementia clinics in Canada. Between 1997 and 1999, an inception cohort of 1,136 subjects entered into this longitudinal study. At the baseline assessment, 10.9% of the subjects were classified as "not cognitively impaired" (NCI), 30.1% as "cognitively impaired not demented" (CIND), and 59% as demented. A subclassification of CIND included amnestic 25.1%, vascular cognitive impairment 18.1%, psychiatric 17.2%, neurologic 7.3%, medical/toxic metabolic 3.5%, mixed 7.6% and not specified 19.0%. The percentage of the cohort referred with dementia increased progressively each decade, while the proportions of CIND and NCI decreased. Within the dementia group, Alzheimer's disease accounted for 47.2% of the subjects, mixed dementias 33.7%, vascular dementia 8.7%, frontotemporal degenerations 5.4%, dementia with Lewy bodies 2.5%, and unclassifiable 1.8%. The ACCORD cohort will allow a detailed study of the longitudinal course of CIND, and the longer-term outcomes of both treated and untreated dementia subjects.