The article by Pasanisi et al. (1) in the February issue of Diabetes Care evaluated the development of diabetes in people with metabolic syndrome who were receiving one of four different treatments: 1) placebo, 2) Mediterranean diet, 3) metformin, and 4) metformin plus Mediterranean diet. After 5 years, the development of diabetes was markedly less in the two metformin groups compared with the placebo and diet-alone groups. The authors concluded that metformin prevents diabetes. The purpose of this comment letter is to put these results in the proper context. Metformin (and other antihyperglycemic drugs) do not change the progressive decrease in insulin sensitivity and β-cell function noted in prediabetes (2). They simply treat the dysglycemia of prediabetes, retarding its increase to the diagnostic levels of diabetes. Metformin (as this article confirmed), thiazolidinediones, and basal insulin lower the rate of development of diabetes in people with prediabetes, but when these drugs are discontinued, diabetes incidence quickly becomes similar to the rate in the placebo groups (3). Furthermore, over time, the progressive β-cell dysfunction weakens the effects of the drugs that were keeping the dysglycemia of prediabetes from reaching the glycemic levels diagnostic of diabetes. For example, diabetes incidence rates were similar in the metformin and placebo groups in the follow-up Diabetes Prevention Program Outcomes Study (4). Finally, Pasanisi et al. stated that the 31% reduction in diabetes incidence in the Diabetes Prevention Program in the metformin-treated group persisted 21 years later, citing Goldberg et al. (5), which was reference 28 in their article. This is incorrect. The article by Goldberg et al. only shows mean A1C levels, with the final values approximated from Fig. 1 of 6.33% in the metformin group and 6.48% in the placebo group (5). The slightly lower A1C level in the metformin group might be expected if some individuals were still responding to the drug.
Article Information
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Handling Editors. The journal editors responsible for overseeing the review of the manuscript were John B. Buse and Justin B. Echouffo Tcheugui.
