Comment on: Matsushita et al. (2010) The association of hemoglobin A1c with incident heart failure among people without diabetes: The atherosclerosis risk in communities study. Diabetes;59:2020-2026.

Matsushita et al. (1) studied associations between heart failure and prediabetic glycemia by categories of A1C and fasting plasma glucose (FPG) using Cox proportional hazards models from the Atherosclerosis Risk in Communities (ARIC) Study and reported that, after the adjustment for covariates including FPG, the hazard ratio of incident heart failure was higher in individuals with A1C 6.0–6.4% and 5.5–6.0% than in the reference group (A1C 5.0–5.4%), but that elevated FPG was not associated with heart failure after adjustment for covariates and A1C. They suggested that chronic hyperglycemia prior to the development of diabetes contributes to development of heart failure (1). It was also reported that A1C was significantly associated with risks of cardiovascular disease (CVD) and death from any cause when adjusting for FPG, but that FPG was not significantly associated with risks of CVD and death from any cause when adjusting for A1C in nondiabetic adults (2). A1C levels appear to increase with age, and any condition that changes red cell turnover will influence A1C levels. Other than age and iron deficiency, some cardiovascular risk factors including inflammation and oxidant stress may have independent associations with A1C. Bilirubin is a potent antioxidant in the human body, and increased serum levels of total bilirubin (TB) are reported to be associated with reduced risk for CVD in some epidemiological studies. In a previously published article (3), we hypothesized that the association of A1C with CVD beyond FPG in nondiabetic adults may at least partly be mediated by the association between A1C and bilirubin. Therefore, in apparently healthy Japanese men (n 1,803) and women (n 1,150), we performed multivariate linear regressions using A1C as a dependent variable and FPG, age, BMI, TB, hemoglobin, iron, white blood cell count (WBC), and high-sensitivity C-reactive protein (hs-CRP) as independent variables as well as logistic regressions using the highest quartile ( 5.6%) and the highest decile ( 5.8%) of A1C as a dependent variable and FPG, age, BMI, TB, hemoglobin, iron, WBC, hs-CRP, and smoking status as independent variables. We found that TB was significantly negatively associated with A1C independently of FPG, age, obesity, inflammation, hemoglobin, and iron in apparently healthy Japanese men and women. Therefore, the association of A1C with heart failure in nondiabetic adults may not only be related to hyperglycemia but also to other factors, including TB.