To determine the relationship between carbohydrate intake and the effect of acarbose on HbA1c in subjects with type 2 diabetes treated with acarbose alone, acarbose plus sulfonylurea, acarbose plus metformin, or acarbose plus insulin.
We conducted a double-blind randomized placebo-controlled study in which subjects with diabetes in four treatment strata (77 on diet alone, 83 treated with metformin, 103 treated with sulfonylurea, and 91 treated with insulin) were randomized to treatment with placebo or acarbose for 12 months. Before randomization, and 3, 6, 9, and 12 months after randomization, fasting blood was obtained for HbA1c, and 3-day diet records were collected. Subjects who completed at least 6 months of acarbose therapy and provided at least three 3-day diet records were included.
In the 114 subjects included in this analysis, carbohydrate intake varied from ∼30–60% of energy. There was no significant relationship between carbohydrate intake and change in HbAlc in any of the four treatment strata (diet: n = 26, r = 0.35, P = 0.076; metformin: n = 27, r = 0.26, P = 0.19; sulfonylurea: n = 35, r = 0.24, P = 0.16; insulin: n = 25, r = −0.27, P = 0.19). In the 80 subjects consuming <50% of energy from carbohydrate, the fall in HbAlc (7.83 ± 0.17% at baseline to 6.72 ± 0.13% on acarbose, P < 0.001) was no different from that of the 34 subjects consuming >50% of energy from carbohydrate (7.55 ± 0.25% at baseline to 6.66 ± 0.23% on acarbose, P < 0.001). There was no difference in carbohydrate intake between those who dropped out of the study because of gastrointestinal side effects and those who did not, and there was no relationship between severity of symptoms and the composition of the diet.
In subjects with type 2 diabetes consuming 30–60% of energy from carbohydrate, the effect of acarbose on HbAlc and gastrointestinal symptoms was not related to carbohydrate intake. Because most people consume at least 30% of energy from carbohydrate, we conclude that no special diet is needed for acarbose to be effective in improving blood glucose control in the treatment of type 2 diabetes.
