Risk of hypoglycemia following intensification of metformin treatment with insulin versus sulfonylurea

Background: Hypoglycemia remains a common life-threatening event associated with diabetes treatment. We compared the risk of first or recurrent hypoglycemia event among metformin initiators who intensified treatment with insulin versus sulfonylurea. Methods: We assembled a retrospective cohort using databases of the Veterans Health Administration, Medicare and the National Death Index. Metformin initiators who intensified treatment with insulin or sulfonylurea were followed to either their first or recurrent hypoglycemia event using Cox proportional hazard models. Hypoglycemia was defined as hospital admission or an emergency department visit for hypoglycemia, or an outpatient blood glucose value of less than 3.3 mmol/L. We conducted additional analyses for risk of first hypoglycemia event, with death as the competing risk. Results: Among 178 341 metformin initiators, 2948 added insulin and 39 990 added sulfonylurea. Propensity score matching yielded 2436 patients taking metformin plus insulin and 12 180 taking metformin plus sulfonylurea. Patients took metformin for a median of 14 (interquartile range [IQR] 5–30) months, and the median glycated hemoglobin level was 8.1% (IQR 7.2%–9.9%) at intensification. In the group who added insulin, 121 first hypoglycemia events occurred, and 466 first events occurred in the group who added sulfonylurea (30.9 v. 24.6 events per 1000 person-years; adjusted hazard ratio [HR] 1.30, 95% confidence interval [CI] 1.06–1.59). For recurrent hypoglycemia, there were 159 events in the insulin group and 585 events in the sulfonylurea group (39.1 v. 30.0 per 1000 person-years; adjusted HR 1.39, 95% CI 1.12–1.72). In separate competing risk analyses, the adjusted HR for hypoglycemia was 1.28 (95% CI 1.04–1.56). Interpretation: Among patients using metformin who could use either insulin or sulfonylurea, the addition of insulin was associated with a higher risk of hypoglycemia than the addition of sulfonylurea. This finding should be considered by patients and clinicians when discussing the risks and benefits of adding insulin versus a sulfonylurea.