Background: 
Hemoglobin A1c (HbA1c) reflects long-time glycemic control and is associated with an increased risk of cardiovascular events among diabetic and non-diabetic patients. The precise impact of HbA1c itself on the all-cause mortality in addition to the control of other cardiovascular risk factors remains uncertain.

Purpose: 
We tested the hypothesis that HbA1c levels associate with long-term mortality.

Methods: 
The present analysis is based on the longitudinal ECAD registry of consecutive patients undergoing coronary angiography at the West German Heart and Vascular Center between 2004 and 2019. HbA1c was quantified at hospital admission using standardized enzymatic methods. The incidence of death due to any cause was evaluated during follow-up. Cox regression analysis was used to determine the association of HbA1c with incident mortality, adjusting for age, sex, systolic blood pressure, low-density lipoprotein cholesterol, smoking status, and family history of premature cardiovascular disease. In addition to the analysis on HbA1c as continuous variable, the association of HbA1c-groups (≤10th percentile, >10th-<25th percentile, 25th-<50th percentile, 50th-<75th percentile, 75th-<90th percentile, and ≥90thpercentile) with incident mortality was determined using HbA1c

<10th percentile as reference.

Results: 
Among 11,476 patients, mean age was 65.4 ± 12.2 years and 70.3% were men. Mean HbA1c was 6.2±1.14%. During a median follow-up 3.0 years, 1,609 patients (14.0%) died. In multivariable analysis, higher HbA1c levels were independently associated with higher all-cause mortality [hazard ratio (95% confidence interval): 1.21 (1.16, 1.26) per 1 standard deviation change in HbA1c, p<0.001. Using HbA1c ≤5.2% (≤10^thpercentile) as reference, we observed a decline in event rate for low-to intermediate HbA1c levels (5.3-5.5%: 0.75 [0.61-0.92], p=0.007; 5.6-5.9%: 0.65 [0.54-0.78], p<0.001) with an increase of the hazard ratio in patients with HbA1c levels above the median (6.0-6.5: 0.95 [0.79-1.14], p=0.6; 6.6-7.5%: 1.19 (0.97-1.46), p=0.09; ≥7.6: 1.48 [1.21-1.83], p<0.001, figure 1).

Conclusions: 
In a large longitudinal registry cohort of patients undergoing invasive coronary angiography, we observe a U-shaped association of HbA1c with long-term mortality with best prognosis of patients in the range of HbA1c levels between 5.3 and 5.9%. 

Figure: Association of HbA1c with incident all-cause mortality in subgroups of HbA1c (≤10th percentile, >10th-<25th percentile, 25th-<50th percentile, 50th-<75th percentile, 75th-<90th percentile, and ≥90th percentile)