Abstract
INTRODUCTION: This study compared basal analog (BA: glargine U100/mL and detemir) and premix (PM: human, lispro and aspart biphasic) insulin regimens in terms of their efficacy and safety in type 2 diabetes mellitus patients. METHODS: Searches of MEDLINE, Embase, and CENTRAL identified primary randomized controlled trials (RCTs) ≥ 12 weeks in duration that compared BA or PM insulin regimens in adults with T2DM, with ≥ 30 patients per arm. A systematic literature review and a pairwise meta-analysis were performed using a random effects model adjusted for between-study variability. Analyses were conducted based on frequency of bolus insulin and PM injections, PM ratio and type, BA type, race, follow-up period, and baseline glycosylated hemoglobin (HbA1c). RESULTS: Twenty-two primary RCTs with 9691 patients were included. The BA and PM regimens yielded similar changes in HbA1c and postprandial glucose levels, with a statistically significant reduction in fasting glucose [mean difference (MD) - 0.61 mmol/L (95% confidence interval (CI) - 0.90, - 0.32), I(2) = 89.6%]. The BA regimens showed significantly reduced rates of total hypoglycemia [odds ratio (OR) 0.77 (95% CI 0.64, 0.92), I(2) = 65.3%] and changes in body weight [MD - 0.48 kg (95% CI - 0.86, - 0.11), I(2) = 75.7%] compared to PM regimens. Stratification by PM type and dosing ratio demonstrated statistically significant reductions in HbA1c favoring BA compared to human [MD - 0.39% (95% CI - 0.60, - 0.18), I(2) = 61.8%] or 50/50-ratio [MD - 0.22% (95% CI - 0.40, - 0.04), I(2) = 0.0%] PM regimens. Other subgroup analyses found no difference in HbA1c change between the BA and PM regimens. CONCLUSION: When compared to PM regimens, BA regimens yielded similar efficacies and better safety profiles in patients with type 2 diabetes mellitus. FUNDING: Sanofi (Shanghai, China).