Incidence and outcomes of diabetes mellitus in elderly people: report fromthe Canadian Study of Health and Aging

Background: The epidemiology of diabetes in elderly people is not well unde rstood. The purpose of this study was to estimate the incidence of diabetes mellitus among elderly people in Canada and the relative risks of death an d admission to an institution among elderly diabetic patients. Methods: The study was a secondary analysis of data for a community-dwellin g sample from the Canadian Study of Health and Aging, a nationwide, represe ntative cohort study of 9008 elderly people (65 years of age or older at ba seline) in Canada. Diabetes was identified primarily by self-reporting, aci d a clinician's diagnosis and the presence of treatments for diabetes were used to identify diabetic patients who did not report that they had the con dition. Results: The reliability of self reported diabetes (the kappa statistic) wa s 0.85. The estimated annual incidence of diabetes was 8.6 cases per thousa nd for elderly Canadians. Incidence decreased with age, from 9.5 for subjec ts 65-74 years of age, to 7.9 for those 75-84 years of age and then to 3.1 for those 85 years of age and older. Diabetes was associated with death (re lative risk 1.87, 95% confidence interval 1.59-2.19) and admission to an in stitution (relative:risk 1.58, 95% confidence interval 1.28-1.94). Interpretation: Diabetes mellitus is common among elderly people, but the i ncidence declines among the very old.