Optimal provision of daytime NPH insulin in patients using the insulin analog lispro

OBJECTIVE - Insulin lispro improves early postprandial blood glucose control but can result in late interprandial hyperglycemia. As an approach to resolving this problem, we per formed a randomized, crossover study with four treatment arms, comparing the daytime metabolic profile after either premeal lispro alone or premeal lispro with optimal daytime NPH insulin and with standard human regular insulin. RESEARCH DESIGN END METHODS - Twelve C-peptide negative type 1 diabetic patients were studied on four separate study days, at least 7 days apart. On each study day, patients received one of the four study insulin treatments, in random order, with identical meals and snacks. The four treatments were 1) premeal human regular insulin before lunch and supper at unchanged dose; 2) premeal lispro (unchanged dose) at lunchtime and dinner; 3) pre-lunch reduced-dose lispro (70%) before lunch and supper with supplemental lunchtime NPH and with a 6-h interval until dinner; and 4) pre-lunch reduced-dose lispro (70%) before lunch and supper with supplemental lunchtime NPH and with a 8-h interval until dinner. All patients were using their usual premeal plus basal insulin regimen during the period of the study with human regular insulin before meals and NPH insulin at bedtime. RESULTS - Postprandial blood glucose concentrations (1230-1500) were lower after reduced or usual lispro dose compared with human regular insulin (5.5 +/- 0.2 and 5.6 +/- 0.2 vs. 8.2 +/- 0.5 mmol/l, P < 0.001), with no difference between the lispro doses. However, preprandial (1800) blood glucose levels deteriorated to higher levels after usual-dose lispro alone compared with either human regular insulin (P < 0.05) or reduced-dose lispro plus NPH (P < 0.05) (8.9 +/- 0.3 vs. 7.1 +/- 0.8 and 6.4 +/- 0.4 mmol/l), with no difference between human regular insulin and reduced-dose lispro plus NPH. During the 2 h between the usual and delayed mealtime, blood glucose concentrations remained controlled on lispro plus NPH (2000: 6.5 +/- 0.4 mmol/l). CONCLUSIONS - Reduced-dose lunchtime lispro plus NPH maintained the improvement in postprandial blood glucose control with no deterioration in interprandial blood glucose control, even up to a late meal.