arznei-telegramm 2002; 33: 72


The STOP-NIDDM Trial: Does Acarbose (Glucobay) Prevent Type 2 Diabetes?

Patients with impaired glucose tolerance possess a very high risk potential of developing type 2 diabetes mellitus. The STOP-NIDDM* trial consisted of 1,429 participants with impaired glucose tolerance who received either 300 mg acarbose (GLUCOBAY) per day or a placebo (1). Impaired glucose tolerance was defined as an increase in plasma sugar levels two hours after oral intake of 75 grams of glucose (oral glucose tolerance test = OGTT) to concentrations of between 7.8 and 11.1 mmol/l (140 to 200 mg/dl), and a fasting plasma glucose level of between 5.6 and 7.7 mmol/l (100 to 140 mg/dl). Four percent of participants were excluded from the final interpretation of the study. In the acarbose group, 31% of patients discontinued the treatment prematurely, mostly due to adverse effects, as did 19% in the placebo group. After a mean study period of 3.3 years, 32% of patients in the acarbose group and 42% in the placebo group showed 2-h OGTT plasma glucose concentrations of over 11.1 mmol/l (200 mg/dl) and thus were diagnosed as diabetics (1). Fasting plasma glucose concentrations and the HbA1c-values were not reported. According to a statement by one of the lead authors of the publication, there were no significant differences in these values in both the acarbose group and the placebo group (2).

The mean 2-h OGTT plasma glucose concentration differed by less than 20 mg/dl between the acarbose and the placebo group (2). This may be due to an isolated effect of acarbose on the results of the OGTT. The OGTT is a test with poor reproducibility, and acarbose can exert an influence on its results (3). In the STOP-NIDDM trial, the incidence of diabetes decreased when a second OGTT was performed. After discontinuing acarbose at the end of the trial, a three months lasting single-blind all-placebo run-out-phase revealed an increased incidence of new-onset diabetes in patients originating from the acarbose group as compared to the originally assigned placebo group (15% vs. 11%) (1).

The relative risk reduction of developing diabetes under acarbose is 24%. According to previous studies, metformin (e.g. GLUCOPHAGE) lowers the relative risk for developing diabetes in patients with impaired glucose tolerance by 31% (4). A change in lifestyle including weight reduction and increased activities lowers the relative risk by more than 50% (4,5).

It is rather disquieting that the financial connections between the authors of the trial and Bayer Corporation, the manufacturer of acarbose, have not been revealed. The Lancet publication noticed that the authors did not indicate a conflict of interest. However, at least the lead author and the German author of the study had received repeatedly financial support by Bayer by way of grants (6,7) and/or other kinds of reimbursements (2,8,9). The statutes of Lancet clearly state that such remunerations are considered a conflict of interest and have to be announced whether or not the author is aware of a conflict (10). Thus, the authors have violated the publication regulations of the journal Lancet. This behaviour may probably explain the incomplete account of study data and their skewed interpretation.

Acarbose (GLUCOBAY) lowers the plasma glucose concentrations only marginally in patients with impaired glucose tolerance, following an oral glucose tolerance test. This effect is clearly less significant than following non-drug measures.
Acarbose has no influence on the fasting glucose levels or the HbA1c value.
Conflicts of interest due to financial connections between the authors and the manufacturer Bayer Corporation have not been publicised and are probably reflected in the suppression and misinterpretation of trial data.


(R = randomised study)



CHIASSON, J.L. et al.: Lancet 2002; 359: 2072



HANEFELD, M.: Pers. Communication, Duesseldorf, 26 June 2002



CHIASSON, J.L.: Diabet. Med. 1996; 13, 3, Suppl. 2: S23-4



Diabetes Prevention Program Research Group.: N. Engl. J. Med. 2002; 346: 393-403



TUOMILEHTO, J. et al.: N. Engl. J. Med. 2001; 344: 1343-50












STOP-NIDDM - Study TO Prevent Non-Insulin Dependent Diabetes Mellitus

© arznei-telegramm 07/02