Objective: The purpose of this study was to elucidate the relationship between fasting plasma glucose (FPG), development of diabetes, and incident heart failure (HF) in a large, community sample of nondiabetic subjects.
Research design and methods: From Kaiser Permanente Northwest medical records, we identified 10,113 subjects with an FPG level of 100-125 mg/dl in 1997 or 1998 who were free of diabetes and HF and matched them to an equal number of subjects with an FPG level of <100 mg/dl on sex and 5-year age groups. Subjects were followed until a new diagnosis of HF was entered into the medical record, death, termination of health plan membership, or December 31, 2005, whichever came first.
Results: After controlling for known HF risk factors, each 10 mg/dl increase in FPG was independently associated with an 8% increase in the risk of HF over a mean follow-up of 79 months [hazard ratio (HR)=1.08, 95% confidence interval (CI) 1.03-1.13, P=.003]. However, in a subsequent analysis that included only those HF cases that occurred prior to diabetes onset and censored follow-up at the time of diabetes development, FPG was not a significant predictor of HF risk (HR=1.01, 95% CI 0.96-1.07, P=.621). Age, male sex, body mass index, smoking, and cardiovascular disease were highly predictive of HF incidence.
Conclusions: Although the risk of HF is increased among subjects with higher FPG, the increased risk is explained by greater likelihood of developing diabetes. Risk factors other than FPG are much stronger independent predictors of incident HF.