Meta-analysis of insulin aspart versus regular human insulin used in a basal–bolus regimen for the treatment of diabetes mellitus

Background: The objective of the current study was to compare the efficacy of two different insulin formulations, insulin aspart (IAsp) and regular human insulin (RHI), for prandial insulin coverage with neutral protamine Hagedorn (NPH) insulin as basal insulin using a meta-analysis approach. The primary endpoint was change in A1c over time. Secondary endpoints included incidence of hypoglycemia and postprandial glycemic control. Methods: Clinical trials (Type 1 and Type 2 diabetes) complying with Good Clinical Practice, and with individual patient data, were included in the meta-analysis. Trials were randomized, consisting of (at least) two treatment arms and had a minimum duration of 12 weeks. Estimates were calculated using fixed-effects and random-effects models. Heterogeneity was assessed for each analysis. The effect of baseline parameters on A1c was analyzed in extended simultaneous models. Results: The mean difference in A1c was 0.1% (95% confidence interval [CI] [−0.15; −0.04], P < 0.001) in favor of IAsp. Higher accumulated dose of IAsp, higher age and increased rates of hypoglycemia were associated with improved A1c outcome. Fasting plasma glucose was not significantly different between regimens. Postprandial glucose was significantly lower after treatment with IAsp compared with RHI, but the analysis did present a significant level of heterogeneity (P < 0.001). The overall rate of hypoglycemia was the same with both regimens, but nocturnal hypoglycemia was significantly lower with IAsp. Conclusions: A basal–bolus regimen with IAsp as bolus insulin provided minimal, but statistically significant, improvement in overall glycemic control with a lower rate of nocturnal hypoglycemic episodes, compared with a corresponding regimen with bolus RHI.