Current issues of ACP Journal Club are published in Annals of Internal Medicine


Acarbose improved glycemic control in type 2 diabetes

ACP J Club. 1995 May-June;122:70. doi:10.7326/ACPJC-1995-122-3-070

Source Citation

Chiasson JL, Josse RG, Hunt JA, et al. The efficacy of acarbose in the treatment of patients with non-insulin-dependent diabetes mellitus. A multicenter controlled clinical trial. Ann Intern Med. 1994 Dec 15;121: 928-35.



To evaluate the effectiveness and safety of acarbose in improving glycemic control in patients with type 2 diabetes mellitus.


Randomized, double-blind, placebo-controlled trial with 1-year follow-up.


7 university-affiliated, tertiary care diabetes clinics in Canada.


354 patients (mean age 57 y, 60% men) with type 2 diabetes of ≥ 6 months duration and hemoglobin A1c (HbA1c) levels > 6.5%. 77 patients were being treated with diet alone (mean HbA1c level 6.7%), 83 with diet and metformin (mean HbA1c level 7.8%), 103 with diet and sulfonylurea (mean HbA1c level 8.0%), and 91 with diet and insulin (mean HbA1c level 7.7%). Exclusion criteria were gastrointestinal disease, receipt of medication to lower serum lipid levels, receipt of medications likely to alter gut motility or absorption, or receipt of β-blockers or thiazide diuretics. 89% of patients were included in the analysis.


All patients were placed on a weight-maintaining diet and continued to receive their hypoglycemic medication. After 6 weeks, 172 patients were assigned to acarbose and 182 to placebo. The initial dose of acarbose was 50 mg 3 times daily; if necessary, the dose was titrated to 100 mg and then to 200 mg 3 times daily. The dose was increased if the postprandial plasma glucose level was > 10 mmol/L. Concurrent hypoglycemic medication was decreased by 25% if the postprandial plasma glucose level was ≤ 10 mmol/L.

Main outcome measures

Levels of HbA1c, fasting and postprandial plasma glucose and serum C-peptide, and fasting serum lipids. Postprandial values were measured after a liquid test meal.

Main results

After 1 year, the HbA1c levels were lower in patients receiving acarbose than in those receiving placebo; the difference was 0.9% for the group treated by diet alone (P = 0.005), 0.8% for the metformin group (P = 0.011), 0.9% for the sulfonylurea group (P = 0.002), and 0.4% for the insulin group (P = 0.077). 52% of patients receiving acarbose were classified as responders (a decrease in HbA1c levels of 15% or to < 7%, or both) compared with 26% of those who received placebo (P < 0.001). The mean postprandial plasma glucose peak level (90 min) was lower in patients receiving acarbose than in those receiving placebo (19.0 to 15.5 mmol/L, P < 0.001). More symptoms of abdominal cramps and discomfort (25% vs 9%), diarrhea (44% vs 20%), and flatulence (73% vs 39%) occurred in patients receiving acarbose than in those receiving placebo.


Acarbose improved glycemic control in patients with non-insulin-dependent diabetes mellitus regardless of concurrent hypoglycemic medication.

Source of funding: In part, Miles Canada, Inc.

For article reprint: Dr. J.L. Chiasson, Centre de Recherche/Hôtel-Dieu de Montréal, 3850 Rue Saint-Urbain, Montréal, Québec H2W 1T8, Canada. FAX 514-843-2709.


In patients with type 2 diabetes, impaired secretion of insulin causes both fasting hyperglycemia and an excessive increase in plasma glucose levels after meals. Acarbose, an inhibitor of intestinal α-glucosidase, impairs carbohydrate digestion and delays, and partially prevents, absorption of carbohydrate. This tends to reduce postprandial hyperglycemia and contributes to overall glycemic control.

Chiasson and colleagues compared acarbose with placebo in a large group of patients whose diabetes was moderately well controlled by diet, oral hypoglycemic agents, or insulin (mean hemoglobin A1c levels were within 2% of the normal range). The addition of acarbose while other treatments were unchanged, modestly improved glycemic control during 1 year of therapy. As expected from its mechanism of action, acarbose produced gastrointestinal side effects, including abdominal cramps, diarrhea, and flatulence.

Chiasson and colleagues showed that although acarbose was a moderately effective agent for long-term treatment of NIDDM, its side effects may affect quality of life. Before clinicians begin using acarbose as part of the standard therapy for diabetes, they should await the results of more clinically pertinent trials, that is, comparisons of acarbose treatment with efforts to optimize glycemic control with current therapy. These trials should include assessments of quality of life. Only with this information can we advise our patients about whether the benefits of acarbose outweigh its side effects.

William E. Clutter, MD
Washington University School of MedicineSt. Louis, Missouri, USA