A randomized double-blind trial of acarbose in type 2 diabetes shows improved glycemic control over 3 years (U.K. Prospective Diabetes Study 44)

OBJECTIVE To determine the degree to which alpha-glucosidase inhibitors, with their unique mode of action primarily reducing postprandial hyperglycemia, offer an additional therapeutic approach in the long-term treatment of type 2 diabetes. RESEARCH DESIGN AND METHODS We studied 1,946 patients (63% men) who were previously enrolled in the U.K. Prospective Diabetes Study (UKPDS). The patients were randomized to acarbose (n = 973), titrating to a maximum dose of 100 mg three times per day, or to matching placebo (n = 973). Mean +/- SD age was 59 +/- 9 years, body weight 84 +/- 17 kg, diabetes duration 7.6 +/- 2.9 years, median (interquartile range) HbA1c 7.9% (6.7-9.5), and fasting plasma glucose (FPG) 8.7 mmol/l (6.8-11.1). Fourteen percent of patients were treated with diet alone, 52% with monotherapy, and 34% with combined therapy. Patients were monitored in UKPDS clinics every 4 months for 3 years. The main outcome measures were HbA1c, FPG, body weight, compliance with study medication, incidence of side effects, and frequency of major clinical events. RESULTS At 3 years, a lower proportion of patients were taking acarbose compared with placebo (39 vs. 58%, P < 0.0001), the main reasons for noncompliance being flatulence (30 vs. 12%, P < 0.0001) and diarrhea (16 vs. 8%, P < 0.05). Analysis by intention to treat showed that patients allocated to acarbose, compared with placebo, had 0.2% significantly lower median HbA1c at 3 years (P < 0.001). In patients remaining on their allocated therapy, the HbA1c difference at 3 years (309 acarbose, 470 placebo) was 0.5% lower median HbA1c (8.1 vs. 8.6%, P < 0.0001). Acarbose appeared to be equally efficacious when given in addition to diet alone; in addition to monotherapy with a sulfonylurea, metformin, or insulin; or in combination with more complex treatment regimens. No significant differences were seen in FPG, body weight, incidence of hypoglycemia, or frequency of major clinical events. CONCLUSIONS Acarbose significantly improved glycemic control over 3 years in patients with established type 2 diabetes, irrespective of concomitant therapy for diabetes. Careful titration of acarbose is needed in view of the increased noncompliance rate seen secondary to the known side effects.