Dynamic risk factors associated with non-severe hypoglycemia in patients treated with insulin glargine or exenatide once weekly.

BackgroundGlycemic control in patients with type 2 diabetes is a dynamic process, and changes in risk factors affecting the incidence of hypoglycemia are not well understood. This study explored the association of longitudinal interactive effects of clinical risk factors and concomitant medications on hypoglycemia risk in patients treated with insulin glargine (IG) or exenatide once weekly (EQW). MethodsPooled patient-level 52-week longitudinal data of treatment with EQW (n=541) or IG (n=223) from three controlled trials were analyzed. ResultsProportions of patients with at least one episode of hypoglycemia in the EQW and IG groups were 23% and 54%, respectively. Compared with patients with HbA1c 7% (53mmol/mol) over time, patients with HbA1c <7% had significantly higher hypoglycemia risk in both groups (95% confidence intervals [CI] of odds ratios [OR]: EQW-1.21, 2.81; IG- 6.26, 9.84). The patterns of interaction effect of changing body mass index (BMI) and HbA1c on hypoglycemia risk differed in the two treatment groups: patients with with BMI >35kg/m(2) had a 119% increased hypoglycemia risk in the EQW group, but a 57% reduced risk in the IG group. Sulfonylurea-treated patients in the EQW and IG groups had 4.7- and 3-fold additional hypoglycemia risk, respectively, versus non-sulfonylurea-treated patients. ConclusionThis study revealed differential effects of tight glycemic control and itsinteraction with treatment-induced changes in BMI on hypoglycemia risk its interaction in patients treated with EQW and IG. The residual adverse effect of sulfonylurea was higher in EQW-treated patients.