Elevated but nondiabetic glucose levels, including impaired fasting glucose (IFG), 2-hour impaired glucose tolerance (IGT), and glycosylated hemoglobin A1c (HbA1c), are associated with increased risk for all-cause mortality. However, the glucose measures may differ in their abilities to predict adverse outcomes. The World Health Organization (WHO) and the American Diabetes Association (ADA) also differ in their definitions of IFG, and it remains unclear which definition best predicts future adverse outcomes. The objective of this study was to evaluate the prognostic value of different cut-points of prediabetic glucose measures for predicting all-cause mortality in individuals without diabetes. In databases including MEDLINE, PubMed, Embase, Clinicaltrials.gov, WHO International Clinical Trials Registry Platform, and Cochrane, we searched prospective cohort studies with minimum 3 years of follow-up, in adults without diabetes at enrollment. Data from eligible studies were pooled to synthesize results for each glucose measure. Random effect regression model was used to calculate pooled hazard ratio or relative risk for all-cause mortality. We screened over 4,000 abstracts and identified 170 eligible studies with mean follow-up of 10.2 years. Compared with individuals with normal glycemia, those with IFG defined by the WHO criteria had 1.14 times increased risk for all-cause mortality. There was no significant increase in mortality risk among those with IFG diagnosed by the ADA criteria. Individuals with IGT had 1.17 times higher risk for all-cause mortality compared with individuals with normal glycemia. HbA1c as low as 5.5% was associated with an increased risk for mortality. The WHO criteria for IFG seems to have better predictive value for all-cause mortality than the ADA criteria. Furthermore, individuals with HbA1c levels at 5.5% may have higher risk for mortality.


M. Huang: None. U. Gujral: None. S. He: None. R. Jagannathan: None. J. Wei: None. L.R. Staimez: None. K. Narayan: None.

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at http://www.diabetesjournals.org/content/license.