Abstract P278: Association of Hemoglobin A1C with Cardiovascular Disease and All-Cause Mortality in Individuals without Diabetes - The Multi-Ethnic Study of Atherosclerosis

Abstract

Background: Previous research suggests that hemoglobin A1C (HbA1C) is a cardiovascular risk marker. Currently, ACCF/AHA recommends HbA1c for cardiovascular risk assessment in adults without diabetes. However, the predictive ability of HbA1C, especially in different ethnic groups, remains unclear. We examined the hypothesis that HbA1c predicts cardiovascular disease (CVD), coronary heart diseae (CHD), and all-cause mortality in a multi-ethnic population free of diabetes and CVD, and assessed whether this association varies by race/ethnicity, in Multi-Ethnic Study of Atherosclerosis (MESA). Methods: During a median follow-up of 5.9 years (maximum: 7.1), 178 had an adjudicated CHD event defined as MI, definite angina, probable angina if followed by PTCI/CABG, resuscitated cardiac arrest and CHD death; 244 had CVD events defined as CHD, stroke, stroke death or CVD death; and 228 died. Cox regression was used to estimate the association of 1% increases in HbA1c with CVD/CHD/all-cause mortality after adjusting for age, gender, race/ethnicity, total household income, systolic blood pressure, hypertensive medications, total and HDL cholesterol, triglycerides, BMI, and smoking. Receiver Operating Curve analysis and Net Reclassification Improvement were used to examine the predictive value of HbA1C for cardiovascular risk assessment. Similar models were used to evaluate these associations stratified by ethnicity. Results: Among 5069 participants, 53% were women, 43% Caucasian, 25% African American, 20% Hispanic, and 11.6% Chinese. Mean age was 63 years and HbA1c was 5.4%. Hemoglobin A1C increased the hazards for CVD and CHD by 32% and 45%, respectively, although the 95% CI overlapped the null slightly [HR (95%CI): 1.32 (0.92-1.89), 1.45 (0.95-2.21)]. Interaction between ethnicity and HbA1c was significant (p<0.001). In stratified analysis, HbA1C was independently associated with CVD only in Caucasians [2.09 (1.25-3.49)] but not in other ethnicities [African Americans: 0.67 (0.35-1.26), Hispanics: 0.89 (0.37-2.14), Chinese: 0.95 (0.17-5.42)]. Overall, addition of HbA1C to Framingham Risk Score did not improve area under the curve for incident CVD (0.703 vs. 0.703; p= 0.18) or net reclassification index (-0.020). However, in Caucasians, the addition of HbA1C to the Framingham Risk Score significantly improved both the area under the curve (0.700 vs. 0.715; p<0.001) and net reclassification index (0.089) for incident CVD. Similar results were obtained for CHD. Conclusion: Associations between HbA1c and CHD/CVD observed in predominately Caucasian populations (and confirmed in this study) may not extend to other ethnicities. More data are needed on ethnic-specific variations in the relationship between glucose homeostasis and cardiovascular disease.