[ effect of impaired fasting glucose and impaired glucose tolerance in prediction of incident type 2 diabetes: Tehran lipid and glucose study]

ABSTRACT

The aim of the current study was to ascertain which impaired glucose homeostasis at baseline is predictive of conversion to type 2 diabetes in an Iranian non-diabetic population. This is a population-based cohort study conducted in 4134 non-diabetic residents of Tehran [the capital city of Iran], aged over 20 years, from 1999 to 2004. Fasting plasma glucose and glucose levels after ingestion of 75 g glucose were measured at baseline [1999-2001] and at follow up [2002-2004] with mean follow up duration of 3.4 years. ADA 1997 and 2004 criteria were used to determine the glucose tolerance status of the participants at baseline and follow up. Using ADA 1997 criteria, the cumulative incidence of diabetes for participants with IFG and IGT, isolated IGT and isolated IFG was 32.3, 10.7 and 9.5% respectively compared with 1.2% for those with normal glucose tolerance at baseline. After application of the new criteria the corresponding incidence rates were 24.8, 6.5 and 5.9% vs. 0.7%. In multivariate logistic regression analysis, the odds ratio for incident diabetes was 1, 6.5[2.5-16.7], 7.7[4.7-12.6] and 28.9[15.6-53.5] in subjects with normal glucose, isolated IFG, isolated IGT and both IFG and IGT respectively using ADA 1997 criteria. The corresponding odds ratios after application of the new criteria were 1, 4.6[2.5-9.0], 6.8[3.6-12.9] and 27.1[15.5-47.5] respectively. In addition to fasting and 2-hour glucose [p<0.001], the waist-to-hip ratio was an important risk factor for developing diabetes [p<0.01]. Both isolated IFG and isolated IGT, and especially combined IFG and IGT, based on either the new or previous ADA criteria are strong predictors for development of type 2 diabetes