Introduction: Discordance among definitions of prediabetes has been well documented, however less is known about which definitions are most indicative of long-term risk of clinical outcomes.
Hypothesis: Prediabetes defined by Hemoglobin A1c (HbA1c) will be more strongly associated with risk of future outcomes than prediabetes defined by fasting glucose (FG) or 2-hour glucose.
Methods: We included 10,692 ARIC participants who attended visit 2 (1990-92), when HbA1c and FG were measured, and 6,521 participants who attended visit 4 (1996-98), when FG and 2-hour glucose were measured, without prevalent diabetes, cardiovascular disease, heart failure, or chronic kidney disease at the relevant baseline. Participants were followed for up to 20 years for all-cause mortality and incident diabetes, cardiovascular disease/heart failure, and chronic kidney disease. Cox proportional hazards models were used to estimate adjusted hazard ratios of incident events for each prediabetes definition compared to normal glycemic levels. Seemingly unrelated regression was used to compare prediabetes definitions.
Results: Prediabetes defined by FG and 2-hour glucose were similarly associated with risk of all-cause mortality, cardiovascular disease/heart failure, and chronic kidney disease (all p-values > 0.1). Prediabetes defined by FG was more strongly associated with risk of diabetes than the 2-hour glucose definition (p < 0.01). Prediabetes defined by HbA1c was more strongly associated with risk of all-cause mortality, diabetes, cardiovascular disease/heart failure, and chronic kidney disease than with prediabetes defined by FG (all p-values < 0.001; Table).
Conclusion: The HbA1c-based definition of prediabetes was more strongly associated with all-cause mortality, and incident cardiovascular disease/heart failure, and chronic kidney disease as compared to FG or 2-hour glucose-based definitions. Screening with HbA1c to identify persons with prediabetes identifies a group of persons at high risk for clinical outcomes.