We appreciate the opportunity to respond to Imai et al. (1) regarding our recently published article on the impact of food order on postprandial glucose and insulin levels (2). As we mentioned in our article, the data on this subject are quite limited and we were unable to elaborate on their referenced study due to limitations of word count (3). There are, however, substantial differences in the design and results of the two studies that merit further discussion.
In their study (3), Imai et al. investigated the effect of eating vegetables (sliced cabbage and tomato) before carbohydrates (rice) or vice versa on postprandial glucose and insulin excursions. The study showed a significant reduction in glucose levels at 30 min, with a trend toward significance at 60 min and no difference at 120 min when vegetables were consumed before carbohydrates.
Our study was designed to simulate real-world Western eating behavior and included a typical complete meal incorporating vegetables, protein (grilled chicken), and carbohydrates (bread). Subjects consumed protein and vegetables together before carbohydrates or vice versa. A very significant reduction in glucose levels was observed at all measured time points after the meal when the protein and vegetables were consumed together before carbohydrates. Ingestion of either protein (4) or vegetables (3) prior to carbohydrates has been demonstrated in previous studies to lower postmeal glucose spikes; a combination of these two meal components may have an additive effect on lowering postmeal glycemia. This may partly explain the greater magnitude of effect seen in our study. Indeed, eating patterns vary across the globe and across ethnicities. Further studies are needed to understand their impact on postmeal glycemia and to determine what works best for individual groups of patients.
We appreciate the contribution of Imai et al. (5) in performing continuous glucose monitoring in patients with a dietary intervention of sequential ingestion of vegetables-protein-carbohydrates or carbohydrates-protein-vegetables. We believe, however, this does not completely address the limitation of the 2-h follow-up in our own study as the meal pattern examined was different and did not include insulin data. The insulin data are particularly relevant for the broader application of this intervention to other subgroups of patients with diabetes and may have implications for insulin requirements. We have already initiated a study with multiple meal patterns and longer follow-up that will examine glucose, insulin, and gut hormone responses to changes in food order.
Finally, recommending a change in food order as a public health measure will require more research in larger numbers of subjects to test both the feasibility and the effectiveness of this intervention.
Article Information
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
