Aims: Administration of bolus insulin after eating may be a useful therapeutic option for some patients. This 6-month, crossover study compared metabolic effects of routine use of preprandial vs. postprandial injection of bolus insulin lispro.
Methods: Thirty-one patients with Type 1 diabetes injected insulin lispro either preprandially or postprandially for a 3-month period followed by the alternate regimen for a further 3 months. HbA1c, fructosamine and eight-point self-determined blood glucose profiles were measured and analysed using an anova model appropriate for a crossover design.
Results: Mean HbA1c decreased slightly from baseline with preprandial (-0.15 +/- 0.41%) and increased slightly with postprandial (0.11 +/- 0.48%) insulin lispro so that there was a significant (P = 0.008) difference between treatments in final HbA1c level. Mean fructosamine also decreased slightly with preprandial (-15 +/- 31 micro mol/l) but was almost unchanged (1 +/- 39 micro mol/l) with postprandial insulin lispro. Overall daily blood glucose was not different (P = 0.312) for preprandial compared with postprandial administration. However, mean preprandial glucose was lower (7.5 +/- 2.01 vs. 6.6 +/- 1.22 mmol/l; P = 0.026), whereas mean postprandial glucose was higher (7.7 +/- 1.8 vs. 8.7 +/- 2.1 mmol/l; P = 0.031) with postprandial insulin lispro administration. Mean blood glucose excursions were higher with postprandial compared with preprandial insulin lispro, indicating greater daily fluctuations. No difference in incidence of hypoglycaemia was observed with the two treatment regimens.
Conclusions: Postprandial insulin lispro administration appeared to be an acceptable treatment regimen and may be of benefit in certain situations. However, the benefits of postprandial administration may have to be balanced against poorer glycaemic control with continuous long-term use.