Comparison of morning or bedtime insulin with and without glyburide in secondary sulfonylurea failure

New treatment options are needed for glycemic control in NIDDM. We evaluated the effects of bedtime or morning insulin treatment, combined with daytime glyburide or given alone.

Twenty-nine male patients with NIDDM, mean age 63 +/- 1.7 yr, body weight 124 +/- 2.98% of DBW, received a maximum glyburide dose (20 mg/day) for a minimum of 6 wk, to confirm sulfonylurea failure. Human lente insulin was added for 12 wk either AM (n = 14) or HS (n = 15) and adjusted to obtain fasting euglycemia (FPG; combination treatment phase). Glyburide was then stopped, and insulin was continued for 6 wk, aiming for normal FPG (insulin phase).

After combination treatment phase, FPG decreased (P < 0.02) from 12.43 +/- 0.68 to 5.73 +/- 0.65 mM (AM) and from 12.68 +/- 0.76 to 5.51 +/- 0.48 mM (HS) (AM vs. HS, NS). Postbreakfast, presupper, and 0200 AM plasma glucose levels fell equally (P < 0.02) except for 1-h postprandial (AM 12.46 +/- 0.51 mM, HS 10.88 +/- 0.62 mM, AM vs. HS, P < 0.1). Mean HBA1c fell similarly in both AM and HS groups. At 2 wk of the insulin phase, FPG was higher in AM than HS, 9.8 +/- 0.76 vs. 7.56 +/- 0.7 mM (P < 0.1). At the end of insulin phase, plasma glucose levels were similar to the end of combination treatment phase, but the insulin dose had to be raised in AM by 39% (P < 0.02) and HS by 30% (P < 0.05). After the combination treatment phase, fasting C-peptide was significantly suppressed in HS group only, from 1.22 +/- 0.12 to 0.82 +/- 0.09 nM (P < 0.02). At the end of insulin phase, fasting C-peptide was further suppressed in both groups, but 2-h postprandial C-peptide levels decreased significantly in AM group only, from 1.85 +/- 0.23 to 1.42 +/- 0.13 nM (P < 0.02). Triglycerides and total and HDL cholesterol did not change significantly after either combination treatment phase or insulin phase. Mean weight gain was 6.5 lb during combination treatment phase (NS from baseline), without further change during insulin phase. Hypoglycemic reactions, all mild, were recorded at a rate of 1.35/patient in the AM group and 0.4/patient in the HS group (P < 0.025).

Normal fasting glycemia and near-normal postprandial glucose profile could be obtained with combination therapy in NIDDM. Results were similar if insulin, alone or in combination with glyburide, was given before breakfast or at bedtime, but hypoglycemic reactions were more common with conventional morning insulin injections.