The chemistry, pharmacology, pharmacokinetics, and clinical efficacy o
f acarbose, a new antidiabetic agent, are reviewed. Acarbose reversibl
y inhibits intestinal alpha-glucosidases, enzymes responsible for the
metabolism of complex carbohydrates into absorbable monosaccharide uni
ts. This action results in a diminished and delayed rise in blood gluc
ose following a meal, resulting in a reduction in postprandial hypergl
ycemia, area under the glucose concentration-time curve, and glycosyla
ted hemoglobin. Other effects include a reduction in postprandial insu
lin and variable changes in plasma lipid concentrations. In placebo-co
ntrolled trials, acarbose caused significant improvements in glycemic
control indicators, including glycosylated hemoglobin. Acarbose has de
monstrated additional glycemic control when added to other antidiabeti
c therapies, including sulfonylureas and insulin. Efficacy of acarbose
appears to be comparable to or slightly less than that of sulfonylure
as or metformin, although it has not been compared with maximal doses
of these agents. The most commonly reported adverse drug reactions wit
h acarbose are abdominal pain, diarrhea, and flatulence, which tend to
lessen with time. Acarbose may affect the bioavailability of metformi
n and may be less effective when used in conjunction with intestinal a
dsorbents and digestive enzyme preparations. Concurrent use with hypog
lycemic agents (sulfonylureas and insulin) may cause an increased freq
uency of hypoglycemia. Acarbose should not be used in individuals with
certain intestinal disorders, including inflammatory bowel disease. T
he dosage should start at 25 mg one to three times daily given with th
e first bite of each main meal and should be adjusted to a maximum of
50 mg three times daily for patients weighing up to 60 kg or 100 mg th
ree times daily for heavier patients. Acarbose may be considered for f
irst-line antidiabetic therapy in certain patients and may be useful a
s combination therapy in selected instances. Acarbose is efficacious i
n improving metabolic control in non-insulin-dependent diabetes mellit
us. Further evaluation of its effects on the long-term complications o
f diabetes is needed.