| Abstract: | This study evaluated the risk of heart failure hospitalization in a 1:1 matched pair sample of sitagliptin ever and never users derived from the Taiwan's National Health Insurance. A total of 85,859 ever users and 85,859 never users matched on 8 digits of propensity score were followed for the first event of heart failure hospitalization until December 31, 2011. The treatment effect (for ever versus never users, and for tertiles of cumulative duration of therapy) was estimated by Cox regression incorporated with the inverse probability of treatment weighting using propensity score. Additionally, adjusted hazard ratios for heart failure were estimated for the baseline characteristics in sitagliptin ever users. Results showed that the incidence of heart failure hospitalization was 1,020.16 and 832.54 per 100,000 person-years, respectively, for ever and never users, with an overall hazard ratio (95% confidence intervals) of 1.262 (1.167-1.364). While compared to never users, the respective hazard ratio for the first, second, and third tertile of cumulative duration < 3.7, 3.7-10.3 and >10.3 months was 2.721 (2.449-3.023), 1.472 (1.318-1.645) and 0.515 (0.447-0.594). Older age, longer diabetes duration, male sex, and use of insulin, sulfonylurea, calcium channel blockers, aspirin, ticlopidine, clopidogrel and dipyridamole were significantly associated with a higher risk in sitagliptin users, but dyslipidemia and use of metformin and statin were protective. In conclusion, sitagliptin increases the risk of heart failure hospitalization within one year of its use, but reduces the risk thereafter. Some factors predisposing to sitagliptin-related heart failure are worthy of attention in clinical practice. |
