We read with great interest the comments by Scaramuzza et al. (1) regarding our article on optimal timing of mealtime bolus insulin (2). We found that administration of insulin aspart 15 min before mealtime resulted in better postprandial glycemic profiles, without increased risk of hypoglycemia. However, administration 30 min before mealtime provided similar postprandial glycemic profiles to that of administration at the start of the meal. Scaramuzza et al. raise the possible explanation that administration of insulin 15 min before the meal better mimics the cephalic response. They refer to an article by Ahrén and Holst (3) in which the cephalic response in healthy subjects was suppressed using trimethaphan, resulting in significantly higher glucose levels between 25 and 60 min postprandially. Additionally, the insulin response in the first 10 min after ingestion of the meal, when blood glucose levels had not yet risen, was reduced by 73 ± 11%. However, the cephalic response does not result in insulin secretion before the start of a meal (3,4). Therefore, administration of insulin analogs 15 min before mealtime cannot truly mimic the cephalic response; it only shifts the plasma insulin concentration curve forward in time, thereby apparently delivering insulin at a more appropriate moment. Injecting 15 min before the meal results in higher insulin concentrations directly after the meal, but also in already increased insulin levels at the start of the meal, unlike in the cephalic response. In our study, insulin levels measured by luminescent oxygen channeling immunoassay (5) at 10 min after the meal were indeed increased by 28 ± 27% in the −15 min group and 36 ± 25% in the −30 min group compared with the 0 min group; however, insulin levels at the start of the meal were also increased by 32 ± 24% in the −15 min group and 71 ± 31% in the −30 min group compared with the 0 min group. To truly mimic the cephalic response, insulin analogs are needed that are more rapidly absorbed in order to establish more physiological plasma insulin levels postprandially without leading to increased insulin levels preprandially.
Acknowledgments
No potential conflicts of interest relevant to this article were reported.
The authors thank O. Skyggebjerg (Novo Nordisk A/S, Diabetes Research Unit, Denmark) for determining the insulin levels.