Sulfonylureas are associated with a higher likelihood of serious hypoglycaemia than metformin, meglitinides or thiazolidinediones, according to the first study to “examine and compare” the risk in patients treated with monotherapy in a real-world setting.

The authors defined serious hypoglycaemia as needing hospitalisation or treatment in an emergency department. Based on this definition, the incidence of serious hypoglycaemia varied from 68.0 per 1,000 patient years’ follow-up for glyburide to 11.9 per 1,000 patient years for metformin.

After allowing for confounders, the risk of serious hypoglycaemia was highest with glyburide (hazard ratio [HR] 3.95 versus metformin), glimepiride (HR 3.28) and glipizide (HR 2.57). The increased risk compared to metformin was less marked with repaglinide (HR 2.03), nateglinide (HR 1.21), rosiglitazone (HR 0.90) and pioglitazone (HR 0.80).

The risk of serious hypoglycaemia increased with higher average daily doses. For example, the incidence rose from 8.8 to 13.4 per 1,000 patient years with low and high doses of metformin respectively. The incidence was 37.2 and 66.8 with repaglinide and 64.5 and 74.3 with glyburide.

Sulfonylureas and meglitinides produce a dose-dependent stimulation of insulin secretion. In addition, patients receiving higher doses usually have serious diabetes and are more likely to show fluctuating glycaemic control – factors which may account for the dose-response relationship.

^{Pharmacoepidemiol Drug Saf DOI:10.1002/pds.4337}