Recently, a joint consensus statement by the American Diabetes Association/European Association for the Study of Diabetes (1) recommended starting insulin therapy for type 2 diabetes with basal insulin and increasing doses until a fasting glucose <130 mg/dl was obtained. The use of rapid-acting, meal-time insulin is considered only when the AlC target is not reached, despite optimal control of fasting glucose.
We observed a consecutive series of 490 outpatients (267 women and 223 men) whose type 2 diabetes was unsatisfactorily controlled (A1C ≥7%) by combined treatment with metformin and insulin secretagogues and for whom glucose self-monitoring data were available (at least four determinations after an overnight fast 2 h after breakfast in the previous month). The average of available values was considered for analysis. Patients included in the analysis had mean ± SD age of 64.8 ± 10.2 years, duration of diabetes 16.0 ± 11.1 years, and BMI 28.3 ± 4.8 kg/m2.
Fasting plasma glucose (FPG), postprandial glucose (PPG), and A1C were 187.4 ± 51.3 mg/dl, 229.0 ± 57.9 mg/dl, and 9.0 ± 3.5%, respectively; 59 (12.0%), 181 (36.9%), and 250 (51.0%) patients showed FPG <130, 130–180, and >180 mg/dl, respectively. PPG was >30% of FPG in 190 (38.8%) patients. The proportion of patients with postprandial hyperglycemia (PPH) was 69.5, 55.2, and 19.6% among those with FPG <130, 130–180, and >180 mg/dl, respectively.
Insulin treatment was initiated in 156 (31.8%) patients. A 6-month follow-up was available for 151 subjects. Of those, 46 (30.5%) patients showed FPG <180 mg/dl and PPG >30% of PPG (PPH), while 39 (25.8%) patients had FPG >130 mg/dl and PPG <30% of FPG (fasting hyperglycemia [FPH]). Among patients with PPH, 30.4, 45.7, and 23.9% received treatment with basal (NPH/glargine) insulin only, prandial (regular/rapid-acting analogs) insulin only, or both, respectively. Corresponding estimates for patients with FPH were 76.6, 29.0, and 42.3%.
Of the patients treated with insulin, 62 (41.1%) showed a reduction of A1C >15% of baseline and/or A1C <7% at 6 months. The proportion of success, defined as above, in patients with PPH was 21.0, 71.4, and 18.2% in those receiving basal insulin only, prandial insulin only, or both, respectively (P < 0.05 for prandial only vs. basal only).
Unsatisfactory glucose control in patients on oral therapy can be due to FPH, PPH, or both. Not surprisingly, patients with PPH alone seem to have a better response to treatment with prandial insulin than with basal insulin. These limited data, obtained through an observational approach, do not have the strength of randomized clinical trials. However, they are coherent with known pharmacokinetics of available insulin formulations.
The greater body of available evidence leads many authors to prefer basal insulin as a first choice for insulin treatment of type 2 diabetes. On the other hand, there is no demonstration of the superiority of basal over prandial insulin in the treatment of oral therapy failure. We feel there is a need for clinical trials specifically designed to compare the two approaches, in which an accurate assessment of phenotype of glucose profiles is obtained through self-monitoring. Until results of such trials are available, any recommendation in favor of either basal or prandial insulin (or both) is somewhat arbitrary.
E.M. has received research grants from Novo Nordisk and Sanofi Aventis Pharmaceuticals and speaking fees from Eli Lilly and Sanofi Aventis Pharmaceuticals; M.M. has received consulting fees from Sanofi Aventis Pharmaceuticals and speaking fees from Eli Lilly; G.M. has received research grants from Sanofi Aventis Pharmaceuticals; N.M. has received research grants from Novo Nordisk; and C.M.R. has received speaking fees from Eli Lilly and Sanofi Aventis Pharmaceuticals.