Insulin treatment may not be associated with worse prognosis in acute heart failure diabetic patients

BACKGROUND: Diabetic patients with heart failure (HF) treated with insulin appear to have a worse prognosis compared to oral anti-diabetic (OAD) agents. Whether insulin is a risk factor or a marker of disease severity is unknown. We studied the prognostic impact of insulin treatment in an acute HF diabetic population. METHODS: From a prospectively recruited population of hospital-admitted acute HF patients we retrospectively selected a convenience sample. Pair-matched analysis: each insulin-treated patient was matched with a diabetic patient with similar glycated hemoglobin (HbA1c +/- 0.2%) under OAD. End-point: all-cause death. Follow-up: 5 years. Insulin-treated and OADtreated patients were compared. Multivariate Cox-regression analysis was used to analyze the prognostic impact of insulin. RESULTS: We studied 92 patients: 46 insulin-treated and 46 HbA1c-matched under OAD. Mean age: 74 +/- 9 years, 46.7% male and 63.5% had HF with reduced ejection fraction. HbA1c was 7.8 +/- 1.5% in both groups. In the subgroup under OAD: 87.0% metformin, 41.3% sulphonylurea, 28.3% dipeptidyl peptidase-4 inhibitors and 4.3% other agents. Insulintreated patients had lower hemoglobin, higher creatinine and discharged B-type natriuretic peptide (BNP) levels (816.6 [289.2-1805.8] vs. 613.3 [205.6-1110.8]; P= 0.02). Seventy three patients died. There were no differences in mortality up to 5 years. After multivariate adjustment accounting for hemoglobin, creatinine and discharge BNP, HR for 1-, 2-, 3-, and 5-year all-cause death in insulin-treated patients were 1.48 (0.62-3.54); P= 0.38, 1.11 (0.55-2.25); P= 0.77, 1.08 (0.562.08); P= 0.28 and 1.24 (0.70-2.19); P= 0.46, respectively. CONCLUSIONS: Insulin-treated diabetic patients with HF and HbA1c-matched patients treated with OAD have similarly ominous prognosis. Our results favor insulin as a marker of poor health condition.