Risk of heart failure in patients with recent-onset type 2 diabetes: population-based cohort study.

BACKGROUND: Although thiazolidinediones precipitate fluid retention in clinical trials, current guidelines advocate their use for patients with diabetes who are felt to be at low risk for heart failure (HF). METHODS AND RESULTS: An inception cohort study was conducted using Saskatchewan Health databases spanning the years 1991 to 1999 (before use of thiazolidinediones) to compare incidence rates of new HF in patients with recent-onset diabetes vs. the general population. Of 12,272 patients with new-onset type 2 diabetes (mean age 63 years), 718 (6%) developed HF over 5.2 years; median time until development of HF was 2.8 years. The adjusted rate of incident HF for the diabetes cohort was 794 cases per 100,000 person years compared with 275 per 100,000 person-years in the general population. Patients with recent-onset diabetes were more likely to develop HF than the general population (adjusted rate ratio 2.9; 95% CI 2.6 to 3.2) and the relative risk was most pronounced in those younger than 60 years (adjusted rate ratio 12.8; 95% CI 8.2 to 20.0). CONCLUSIONS: The incidence of HF is relatively high within 5 years of diabetes onset, calling into question the ease with which individuals with diabetes "at low risk of HF" can be identified.