[LYS(B28), PRO(B29)]-HUMAN INSULIN - EFFECT OF INJECTION TIME ON POSTPRANDIAL GLYCEMIA

Background: [Lys(B28), Pro(B29)]-human insulin (lispro) is an insulin analogue with a reduced capacity for self-association and faster absor ption from subcutaneous injection sites. We hypothesized that administ ration of lispro closer to a meal would result in better glucose contr ol than that achieved with regular insulin. Methods: This trial used a randomized crossover design that consisted of a period of metabolic s tabilization lasting 9 days followed by an evaluation period lasting 5 days. The patients received weight-maintenance diets, and insulin dos es were adjusted as needed, Calorie intake, insulin dose, and activiti es were kept constant once the evaluation period began. During the eva luation period, we varied the time between insulin injection and mealt ime and assessed glucose control. Results: During the evaluation perio d, the lowest mean glucose concentrations were 117.9 mg/dl for lispro and 119.8 mg/dl (p = 0.817) for regular insulin. To obtain these, we g ave lispro, on average, 22.5 minutes before meals and regular insulin 63.8 minutes before meals (p = 0.006). A similar pattern was evident t hroughout the glucose control parameters. The exception was mean ampli tude of glucose excursion, which was lower after lispro (59 versus 75 mg/dl; p = 0.007) compared with regular insulin. Conclusions: We achie ved equal or slightly better glucose control, as reflected by mean amp litude of glucose excursion, with insulin lispro given much closer to meal time than that achieved with regular insulin. As a result of thes e findings, we propose that a rapidly absorbed analogue of insulin is capable of achieving better control of postprandial glucose at a more convenient injection time.