A single-center, randomized, double-blind, three-way crossover study examining postchallenge glucose responses to human insulin 70/30 and insulin lispro fixed mixtures 75/25 and 50/50 in patients with type 2 diabetes mellitus

Clin Ther. 2006 Oct;28(10):1649-57. doi: 10.1016/j.clinthera.2006.10.017.


Objective: The aim of this study was to test the ability of human insulin 70/30, insulin lispro mixture 75/25 (75% neutral protamine lispro [NPL], 25% insulin lispro), and insulin lispro mixture 50/50 (50% NPL, 50% insulin lispro) to control postprandial glucose (PPG) concentrations in patients with type 2 diabetes mellitus (DM).

Methods: This single-center, randomized, double-blind, 3-way crossover study was conducted at the Diabetes and Glandular Disease Research Center, San Antonio, Texas. We measured serum glucose responses after a standardized breakfast test meal (500 kcal; 50% carbohydrate, 34% fat, 16% protein) in patients with type 2 DM receiving a single preprandial dose of human insulin 70/30, insulin lispro 75/25, or insulin lispro 50/50 by SC injection. All patients previously used insulin for at least 1 month prior to the study. The glucose responses were compared with those of healthy, untreated subjects administered the identical meal.

Results: A total of 33 patients were enrolled (23 with type 2 DM and 10 healthy controls). The baseline characteristics of the patients were as follows: sex ratio (M:F), 17:6; mean (SD) age, 61.3 (10.0) years; mean (SD) body weight, 98.5 (13.2) kg; mean (SD) body mass index, 33.0 (3.8) kg/m(2); mean (SD) glycosylated hemoglobin, 8.1% (1.6%); mean (SD) fasting serum glucose (FSG), 158.7 (27.6) mg/dL; and 56.5% white, 8.7% black, and 34.8% Hispanic. The mean (SD) doses (U/d) of the fixed-mixture preparations were similar: human insulin 70/30, 44.1 (18.2); insulin lispro 75/25, 44.1 (18.2); and insulin lispro 50/50, 43.8 (17.8). FSG levels obtained before the test meal were not significantly different between treatments. Compared with those in healthy subjects, incremental glucose AUC values for the 4 hours after the meal (AUCglucose 0-4) for patients with type 2 DM were 6.4-fold higher with human insulin 70/30, 4.6-fold higher with insulin lispro 75/25, and 3.0-fold higher with insulin lispro 50/50. Each insulin regimen produced AUC(glucose) 0-4 and 2-hour PPG values significantly different from all other regimens (all, P < 0.05). Mean (SD) 2-hour PPG values (mg/dL) were lower with mixtures containing insulin lispro than with human insulin 70/30 and decreased as the proportion of insulin lispro within the fixed mixtures increased: human insulin 70/30, 212.6 (47.0); insulin lispro 75/25, 198.0 (67.5); and insulin lispro 50/50, 158.8 (52.3).

Conclusions: In this small study in patients with type 2 DM and healthy controls, preprandial administration of a fixed mixture containing rapid-acting or regular insulin and intermediate-acting components was associated with attenuation of the rise in PPG in patients with type 2 DM administered a test meal. Mixtures containing insulin lispro were associated with greater decreases in PPG concentrations compared with human insulin 70/30. Furthermore, greater amounts of rapid-acting insulin contained within the mixture were associated with better PPG control.

Publication types

  • Randomized Controlled Trial
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Aged
  • Blood Glucose / analysis*
  • Cross-Over Studies
  • Diabetes Mellitus, Type 2 / drug therapy*
  • Double-Blind Method
  • Female
  • Humans
  • Hypoglycemic Agents / therapeutic use*
  • Insulin / analogs & derivatives
  • Insulin / therapeutic use*
  • Male
  • Middle Aged
  • Postprandial Period


  • Blood Glucose
  • Hypoglycemic Agents
  • Insulin