Prevalence and progression of impaired glucose homeostasis assessed by the different criteria for IFG in Japanese adults

Abstract

To the Editor: Several studies have indicated that isolated IGT (that is, IGT with normal fasting plasma glucose) is more common than isolated IFG (that is, IFG with normal 2-h plasma glucose after OGTT). Based on current diagnostic thresholds, 30–60% of subjects with IGT have normal fasting plasma glucose (FPG) levels [1–3]. Furthermore, a recent longitudinal study on the natural history of type 2 diabetes has demonstrated that the threshold of 6.1 mmol/l (110 mg/dl) for normal FPG level does not create an equivalent category of glucose homeostasis to IGT for either the subsequent development of diabetes or in terms of coronary heart disease risk factors [4]. To reconcile these inconsistencies and to provide an equivalent predictive power to both categories, the American Diabetes Association most recently redefined IFG as FPG levels of at least 5.55 mmol/l (100 mg/dl) but below 7.0 mmol/l (126 mg/dl) [5]. To date, however, there is little information about differences in prevalence and progression of impaired glucose homeostasis (IFG and IGT) when different criteria for IFG are used.We assessed the feasibility of these criteria by comparing the proportion of categories for IFG and IGT with their progression to type 2 diabetes in Japanese adults. The study was approved by the institutional ethics committee. The subjects of the present study comprised the group of general health check-up examinees who enrolled between 1996 and 2003 and underwent a series of examinations at the Department of Health Care in our hospital. Of 853 examinees, 46 already had diabetes and were eliminated, and the remaining 807 subjects (54.4±9.8 years [mean±SD], 81.2%men) underwent a standard 75-g OGTT after an overnight fast. A substantial number of individuals participated in this programme on several occasions over the study period. Their serial OGTTs were observed on occasions to convert to different diagnostic categories, for example NGT at one time point and IGT at a subsequent time point or vice versa. Therefore, we analysed all of them (1,406 OGTTs) separately to avoid selection bias. According to the modified criteria [5], NGT is defined as FPG below 5.55 mmol/l (100 mg/dl) and 2-h PG below 7.8 mmol/l (140 mg/dl), isolated IFG as FPG between 5.55 and 6.9 mmol/l (100–125 mg/dl) and 2-h PG below 7.8 mmol/l (140 mg/dl), isolated IGT as FPG below 5.55 mmol/l (100 mg/dl) and 2-h PG between 7.8 and 11.0 mmol/l (140–199 mg/dl), and combined IFG and IGT (IFG/IGT) as FPG between 5.55 and 6.9 mmol/l (100–125 mg/dl) and 2-h PG between 7.8 and 11.0 mmol/l (140–199 mg/dl). At baseline, the prevalence of NGT was 58.4% (821 of 1,406 OGTTs) by the current criteria and was reassessed at 41.1% (578/1,406) by the modified criteria: IFG 4.9% at 22.2%, IGT 20.9% at 12.2%, and IFG/IGT 6.0% at 14.7%, respectively.Theprevalenceofdiabeteswas thesame(9.7%) by both criteria as expected. Thus, based on the current criteria, isolated IFGwas found only in 25% of isolated IGT (4.9 vs 20.9%), which was consistent with previous studies [1–3]. However, using the modified criteria, its prevalence increased approximately two-fold that of IGT (22.2 vs 12.2%). Among 137 diabetic OGTTs, only 20 OGTTs (14.6%) were diagnosed as diabetic by FPG criteria alone. In contrast, 64 OGTTs (46.7%) were diagnosed as diabetic by 2-h PG criteria alone and their FPG concentrations were normal in more than half of the cases (38 OGTTs). Thus, there is a serious flaw when as many as 46.7% of diabetic individuals could be left undiagnosed using the FPG criteria alone. We analysed the progression from NGT or impaired glucose homeostasis to each category for glucose tolerance on the subset of 95 subjects (51.2±9.0 years, 84 K. Hanai . Y. Kiuchi . T. Wasada (*) Saitama-ken Saiseikai Kurihashi Hospital, 714-6 Kouemon Kurihashi-machi, Kitakatsushika-gun, Saitama-ken, 349-1105, Japan e-mail: tounyo@saikuri.org