Abstract
In this analysis, the association of baseline glycated hemoglobin (HbA1c) levels with all-cause mortality in cardiovascular disease (CVD) patients with diabetes was investigated using data from NHANES 1999-2014. Under examination were 845 CVD patients who had diabetes and were followed for a median follow-up of 7.3 years, and an all-cause mortality rate of 22.60% was observed. To examine the association between HbA1c and mortality, multivariable Cox proportional hazard models using spline models determined the non-linear association. HbA1c as a continuous variable was not associated with mortality. However, a significant association was observed when HbA1c was classified according to quartiles. Particularly, after adjustment for potential confounders, in comparison to participants with HbA1c levels below 6.2%, patients with HbA1c levels of 6.2-6.8% and 6.9-7.6% had lower risks of all-cause mortality (hazard ratio: 0.49, 95% CI: 0.30-0.80 and hazard ratio: 0.64, 95% CI: 0.39-1.03, respectively). Using restricted cubic splines, further testing confirmed the lack of a linear association and instead suggested a U-shaped relationship between HbA1c and mortality, with an optimal HbA1c target value of 6.9%. A 1-unit increase in HbA1c with HbA1c less than or equal to 6.9% was predictive of a 55% reduction in all-cause mortality compared to HbA1c levels above 6.9%, which exhibited an elevation in risk. All told, these data suggest that the relationship between HbA1c and all-cause mortality in CVD patients with diabetes is non-linear and U-shaped, and therefore may suggest that individualization of glycemic control may be beneficial for this patient population.