This study evaluated the proportions of prandial (PHG) vs fasting hyperglycaemia (FHG) over 24h in a group of patients with type 2 diabetes (overall and for Caucasian vs Asian patients), and tested the hypothesis that an insulin regimen with a prandial component allows a greater response than basal insulin at low glycated haemoglobin (HbA_1c) levels with a higher proportion of PHG than FHG.

Relative contributions of PHG and FHG to overall hyperglycaemia were analyzed by baseline HbA_1c quartiles and by ethnicity at baseline and after 24-week treatment with either insulin glargine or insulin lispro mix 25 in the DURABLE study.

With increasing baseline HbA_1c, the mean relative contribution of PHG to the total area under the curve decreased (from 41% to 27%) while FHG was increased (from 59% to 73%). Both insulins decreased FHG, but only insulin lispro mix 25 decreased PHG. More patients with baseline HbA_1c<9%, where PHG was more relevant, achieved the target HbA_1c of<7% at endpoint with insulin lispro mix 25 compared with glargine. On average, Asians had a 10% larger contribution of PHG at all HbA_1c quartiles, and a lower proportion of Asians reached the HbA_1c target of<7% with either insulin treatment compared with Caucasians.

At baseline, the contribution of FHG to overall hyperglycaemia predominated at all HbA_1c quartiles, whereas PHG was more clinically relevant at lower HbA_1c levels and with a greater response to insulin lispro mix 25. Asians had a greater proportion of PHG and a lesser response to either insulins compared with Caucasians. Thus, responses to diabetes drugs by baseline HbA_1c and ethnicity are worth investigating to better target and individualize treatment.