Acarbose delays release of glucose from complex carbohydrates and disaccharides by inhibiting intestinal a-glucosidases, thereby attenuating postprandial increments in blood glucose and insulin. This multicenter, double-blind, placebo-controlled study compared the efficacy and safety of diet alone, acarbose, tolbutamide, and acarbose-plus-tolbutamide in non-insulin-dependent diabetes mellitus (NIDDM) patients.

A total of 290 patients with NIDDM and fasting plasma glucose levels of at least 140 mg/dL were randomized to receive treatment tid with acarbose 200 mg, tolbutamide 250 to 1,000 mg, a combination of both drugs, or placebo. A 6-week run-in period was followed by double-blind treatment for 24 weeks, then a 6-week follow-up period.

All active treatments were superior (P <0.05) to placebo in reducing postprandial hyperglycemia and HbA_1c levels. The ranking in order of efficacy was: acarbose-plus-tolbutamide, tolbutamide, acarbose, and placebo. The postprandial reductions in glucose were approximately 85 mg/dL for acarbose-plus-tolbutamide, 71 mg/dL for tolbutamide, 56 mg/dL for acarbose, and 13 mg/dL for placebo. Tolbutamide was associated with increases in body weight and postprandial insulin levels when taken alone, but these were ameliorated when tolbutamide was taken in combination with acarbose. Acarbose alone or in combination with tolbutamide caused significantly more gastrointestinal adverse events (mainly flatulence and soft stools or diarrhea) than tolbutamide or placebo, but these were generally well tolerated. Clinically significant elevations in hepatic transaminase levels occurred in 3 patients in the acarbose group and 2 in the acarbose-plus-tolbutamide group. Transaminase levels returned to normal when therapy was discontinued.

Acarbose was effective and well tolerated in the treatment of NIDDM. Control of glycemia was significantly better with acarbose compared with diet alone. Acarbose-plus-tolbutamide was superior to tolbutamide alone.