Comparative Effectiveness of Basal-Bolus Versus Premix Analog Insulin on Glycemic Variability and Patient-Centered Outcomes during Insulin Intensification in Type 1 and Type 2 Diabetes: A Randomized, Controlled, Crossover Trial

Context: In patients with diabetes, intraday glucose variability might predict health outcomes independently from glycosylated hemoglobin (HbA(1c)). Objective: Our objective was to evaluate patient satisfaction (PS), quality of life (QoL), glycemic control, and variability during insulin intensification to HbA(1c) below 7.0%. Patients, Design, and Setting: Eighty-two type 1 and 306 insulin-treated type 2 diabetes patients (47% male; age 54 +/- 11 yr; HbA(1c) = 7.8 +/- 0.7%) participated in this multicenter, randomized, crossover trial at 52 U.S. centers. Interventions: Interventions included insulin glargine plus premeal glulisine (n = 192) vs. twice-daily premix 75/25 or 70/30 analog insulin (n = 196) for 12 wk and crossed to the alternate arm for 12 wk. Main Outcome Measures: Main outcome measures included PS and QoL questionnaires, 3-d continuous glucose monitoring (CGM), and HbA(1c) every 4-8 wk. Results: Mean +/- SE HbA(1c) change was -0.39 +/- 0.09% for glargine-glulisine and -0.05 +/- 0.09% for premix (P < 0.0001). The PS net benefit scale (0-100) improved from 51.1 to 60.5 +/- 1.2 for glargine-glulisine and worsened to 45.4 +/- 1.2 for premix (P < 0.0001). The PS regimen acceptance scale was comparable (P = 0.33). Overall QoL favored glargine-glulisine (P < 0.001), as did perceived health (P < 0.0001), symptom distress (P < 0.0001), general health perceptions (P < 0.01), and psychosocial (P < 0.02). CGM daily glucose mean, daily glucose SD (glycemic variability), and percent time over 140 mg/dl were lower for glargine-glulisine by 13.1 +/- 2.7 mg/dl, 5.9 +/- 1.4 mg/dl, and 7.3 +/- 1.6%, respectively (all P < 0.0001), with no difference in CGM percent time below 70 mg/dl (P = 0.09). Symptomatic hypoglycemia rates were comparable. HbA(1c), mean CGM daily glucose, and glycemic variability were independent predictors of PS net benefit. Conclusions: Patient satisfaction was impacted more positively by improved QoL, reduced glucose variability, and better glycemic control with a basal-bolus regimen than negatively by the burden of additional injections, thereby facilitating insulin intensification and the ability to achieve HbA(1c) below 7.0%. (J Clin Endocrinol Metab 97: 3504-3514, 2012)