Acarbose, an α-glucosidase inhibitor, controls postprandial hyperglycemia by slowing carbohydrate digestion and absorption. The slowing is accomplished by competitive enzyme inhibition at the brush border of the small intestine. Rosenstock and colleagues studied the effect of acarbose added to metformin in the treatment of persons with type 2 diabetes (formerly known as non–insulin-dependent diabetes) not adequately controlled by either agent alone.

A total of 168 patients with type 2 diabetes who exhibited inadequate glucose control with diet plus 2,000 or 2,500 mg metformin daily were included in the study. In this multicenter, double-blind, placebo-controlled study, patients were randomized to receive metformin and either placebo or 25 mg of acarbose three times a day, taken orally at the beginning of each meal. Four weeks after randomization, the acarbose was increased to 50 mg three times a day. After eight weeks, the dosage was raised to 100 mg three times a day if the one-hour postprandial blood glucose level was greater than 160 mg per dL (8.88 mmol per L). This became the dosage that study patients received throughout the study. Test meals were given at randomization and at weeks 4, 12 and 24. Blood samples were obtained at zero, 60, 90 and 120 minutes after the meal to determine plasma glucose, serum insulin and triglyceride levels. Blood samples for determination of glycosylated hemoglobin (HbA_1c) levels were also taken during screening and randomization, and at all study visits.

Of the 147 patients whose results were evaluable, the 73 patients in the acarbose and metformin arm of the study had an initial mean HbA_1c level of 8.46 percent and a mean decrease of 0.57 percent, while the mean HbA_1c level in the patients taking placebo and metformin was initially 8.17 percent, and the mean increase was 0.08 percent. The mean fasting plasma glucose level at end point increased 1.8 mg per dL (0.10 mmol per L), from 195.2 mg per dL (10.83 mmol per L), in the placebo-treated patients and decreased 12.7 mg per dL (0.71 mmol per L), from 203.7 mg per dL (11.31 mmol per L), in the acarbose-treated patients. Plasma glucose levels were significantly lower at 60, 90 and 120 minutes after the standard meals in patients in the acarbose group compared with patients in the placebo group. A positive clinical response was defined as an HbA_1c value of 7 percent or less, associated with a reduction from baseline of at least 1 percent. Such a response was noted in 36 percent of the acarbose group compared with 16 percent of the placebo group. No significant differences in liver transaminase elevations were observed between patients treated with acarbose and those treated with placebo.

The authors conclude that the addition of acarbose therapy in patients already receiving metformin reduced postprandial plasma glucose levels. Control of these postprandial glucose levels ultimately lowered HbA_1c levels. The authors caution that 12 percent of the patients receiving acarbose withdrew from the study because of diarrhea, flatulence and abdominal pain.

editor's note: Although a number of patients taking acarbose and metformin experienced modest improvement in HbA_1c and plasma glucose levels, many of them had significant gastrointestinal complaints, particularly flatulence. Acarbose may have a therapeutic role in certain specific circumstances, but its use may be limited by side effects.—t.o.