Diabetes in older people: Prevalence, incidence and its association with medium- and long-term mortality from all causes.

OBJECTIVE To estimate the prevalence and incidence of self-reported diabetes and to study its association with medium- and long-term mortality from all causes in persons ≥ 65 years. DESIGN A population-based cohort study begun in 1993. SETTING "Envejecer en Leganés" cohort (Madrid). PARTICIPANTS A random sample of persons ≥ 65 years (n=1277 in the 1993 baseline sample). METHODS Participants were classified as having diabetes if they so reported and had consulted a physician for this reason within the last year. Diabetes history was categorized in <10 and ≥ 10 years in 1993. Incidence density was calculated in 2-year periods in non-diabetic individuals (1965 persons/2 years). Vital status was recorded on 31 December 2011. The association between diabetes history ≥ 10 years and mortality at 6 and 18 years follow-up was studied by the Kaplan-Meier and Cox regression analyses after adjusting for age, sex, heart disease and comorbidity. RESULTS The prevalence of self-reported diabetes rose from 10.3% in 1993 to 16.1% in 1999 (p ≤ 0.001) and was higher in women than men (p ≤ 0.05). Total incidence density was 2.6 cases/100 persons/2 years (95% CI: 2.0-3.3). Medium- and long-term mortality was higher in persons with diabetes history ≥ 10 years than in non-diabetic individuals (HR: 2.0; 95% CI: 1.2-3.3 and HR: 1.7; 95% CI: 1.1-2.5, respectively). In diabetics with history <10 years the HR was 1.3 (95% CI: 0.9-1.9) and HR: 1.5 (95% CI: 1.2-1.9, respectively). CONCLUSIONS Although diabetes is clearly associated with increased risk of mortality, it is significant only for patients with ≥ 10 years' history of diabetes.