Diagnosis of Diabetes Mellitus

To the Editor.—  In the article by Turkington and Weindling entitled "Insulin Secretion in the Diagnosis of Adult-Onset Diabetes Mellitus" (240:833, 1978), the authors suggest that the demonstration of "insulin deficiency should be essential to establish the diagnosis of diabetes mellitus." Their data are interesting and provocative. However, before accepting their interpretation, it appears that several variables in this study were uncontrolled, suggesting that the interpretation of the study be de-emphasized. From a general point of view, the authors tend to stress insulin deficiency in absolute terms as essential in the diagnosis of diabetes mellitus without correlating this with the degree of glycemia, carbohydrate intolerance, or obesity. Many normal persons with relatively low insulin responses clearly do not have hyperglycemic or diabetic complications. Perhaps the authors' conclusions would be more firm if they employed one of the time-honored insulin-glucose relationships.Information concerning glucose levels or glucose tolerance tests during the