Impact of insulin lispro on HbA1c values in insulin pump users

To compare the therapeutic efficacy of the short-acting insulin analogue insulin lispro (Humalog) with that of buffered regular human insulin (Velosulin) in patients on insulin pump therapy.

Sixty-two (45 women and 17 men) young patients with type 1 diabetes using insulin pump therapy were compared while using buffered regular human insulin for a mean +/- s.e.m. of 20.1+/-1.2 months or insulin lispro for a mean +/- s.e.m. of 19.7+/-0.5 months. The initial mean +/- s.e.m. age and duration of diabetes were 29.1+/-0.9 and 17.7+/-0.9 years, respectively. The mean HbA1c values, basal insulin dosages, premeal insulin dosages and number of low blood sugars were recorded during treatment with both insulins.

Mean +/- s.e.m. HbA1c values were significantly lower (p < 0.001; paired Wilcoxon t-test) during insulin lispro treatment (7.4+/-0.1%) as compared to treatment with buffered regular human insulin (7.9+/-0.1%). Total units of insulin (mean +/- s.e.m.)/kg/day was significantly (p = 0.03) lower (0.61+/-0.02) during the insulin lispro treatment period as compared to the buffered regular human insulin treated period (0.65+/-0.03). Total mean +/- s.e.m. (U/kg/day) of basal insulin administered per day was higher when patients received insulin lispro treatment (0.44+/-0.02 vs. 0.42+/-0.01 for buffered regular human insulin treated period; p = 0.002). The premeal insulin boluses (mean +/- s.e.m.) for the two treatment groups were significantly different with less insulin required for the insulin lispro treatment period for all three meals (p < 0.001, t-test). The number of mild/moderate and severe hypoglycaemic episodes were similar in the two groups.

We conclude that use of insulin lispro in pump therapy significantly lowers HbA1c values in comparison to therapy with buffered regular human insulin insulin without increasing hypoglycaemic episodes.