Diabetes mellitus has been a major concern as a common risk factor for cardiovascular disease. Glycated hemoglobin (HbA_1c) variability is an indicator of long-term blood glucose fluctuation. Therefore, it is of great clinical significance to explore the correlation between HbA_1c variability and new-onset atrial fibrillation (AF) in diabetic patients combined with heart failure with preserved ejection fraction (HFpEF) .

To investigate the correlation between HbA_1c variability and new onset AF in type 2 diabetes mellitus (T2DM) patients combined with HFpEF.

The clinical data of 317 T2DM patients combined with HFpEF diagnosed in the Department of Cardiology, the Second Affiliated Hospital of Zhengzhou University from January 2018 to January 2019 were retrospectively analyzed. The follow-up was performed until February 2022, with a mean follow-up time of 3.4 years. The included patients were divided into the AF group (34 cases) and non-AF group (283 cases) based on the presence of new-onset AF during the follow-up period. The HbA_1c variability was expressed as standard deviation of HbA_1c measurement (HbA_1c-SD) and HbA_1c coefficient of variation (HbA_1c-CV). Multivariate Cox regression analysis was used to explore the correlation between HbA_1c variability and new-onset AF in T2DM patients combined with HFpEF. The survival curves were plotted by the Kaplan-Meier (K-M) method. The receiver operating characteristic (ROC) curve of HbA_1c variability predicting new-onset AF in T2DM patients combined with HFpEF was plotted.

The HbA_1c-SD and HbA_1c-CV of patients in the AF group were higher than those in the non-AF group (P<0.05). The included patients were divided into the low HbA_1c variability (HbA_1c-SD≤0.34%, HbA_1c-CV≤4.74%) and high HbA_1c variability (HbA_1c-SD>0.34%, HbA_1c-CV>4.74%) groups according to the median of HbA_1c variability. Log-rank test results showed higher incidence of new-onset AF in patients with high HbA_1c variability (P_HbA1c-SD<0.001, P_HbA1c-CV=0.004). Multivariate Cox regression analysis showed that HbA_1c-SD〔HR=2.22, 95%CI (1.37, 3.61), P=0.001〕 and HbA_1c-CV〔HR=1.65, 95%CI (1.01, 2.67), P=0.001〕 were independent influencing factors for new-onset AF in T2DM patients combined with HFpEF. The AUC of HbA_1c-SD for predicting AF in T2DM patients combined with HFpEF was 0.784 〔95%CI (0.713, 0.855), P=0.001〕, with the optimum cutoff value of 0.36%, sensitivity and specificity of 79.4% and 73.1%, respectively. The AUC of HbA_1c-CV for predicting AF in patients with T2DM and HFpEF was 0.694 〔95%CI (0.591, 0.797), P<0.001〕, with the optimal cutoff value of 4.97%, sensitivity and specificity of 73.5% and 72.1%, respectively.

High HbA_1c variability (HbA_1c-SD>0.34%, HbA_1c-CV>4.74%) is independently associated with an increased risk of new-onset AF in T2DM patients combined with HFpEF, with significant clinical value in predicting AF.

糖尿病作为心血管疾病常见的危险因素一直备受关注。糖化血红蛋白（HbA_1c）变异性是反映长期血糖波动的指标。因此探讨HbA_1c变异性与糖尿病合并射血分数保留心力衰竭（HFpEF）新发心房颤动（AF）的关系具有重要临床意义。

探究在2型糖尿病（T2DM）合并HFpEF患者中，HbA_1c变异性与其新发AF的关系。

回顾性分析2018年1月—2019年1月于郑州大学第二附属医院心血管内科确诊为T2DM合并HFpEF患者317例的临床资料。随访截至2022年2月，平均随访时间3.4年。根据随访期内有无新发AF，将患者分为AF组（34例）与非AF组（283例）。HbA_1c变异性以HbA_1c测量标准差（HbA_1c-SD）和HbA_1c变异系数（HbA_1c-CV）表示。采用多因素Cox回归分析探究HbA_1c变异性与T2DM合并HFpEF患者新发AF的关系。采用K-M法绘制生存曲线。绘制HbA_1c变异性预测T2DM合并HFpEF患者新发AF的受试者工作特征曲线（ROC曲线）。

AF组患者HbA_1c-SD、HbA_1c-CV高于非AF组（P<0.05）。依据HbA_1c变异性中位数分为低HbA_1c变异性患者（HbA_1c-SD≤0.34%，HbA_1c-CV≤4.74%）和高HbA_1c变异性患者（HbA_1c-SD>0.34%，HbA_1c-CV>4.74%）。Log-rank检验结果显示，高HbA_1c变异性患者新发AF发生率高（P_HbA1c-SD<0.001、P_HbA1c-CV=0.004）。多因素Cox回归分析结果显示，HbA_1c-SD〔HR=2.22，95%CI（1.37，3.61），P=0.001〕、HbA_1c-CV〔HR=1.65，95%CI（1.01，2.67），P=0.001〕是T2DM合并HFpEF患者新发AF的独立影响因素。HbA_1c-SD预测T2DM合并HFpEF发生AF的ROC曲线下面积（AUC）为0.784〔95%CI（0.713，0.855），P=0.001〕，最佳截断值为0.36%，灵敏度为79.4%，特异度为73.1%。HbA_1c-CV预测T2DM合并HFpEF发生AF的AUC为0.694〔95%CI（0.591，0.797），P<0.001〕，最佳截断值为4.97%，灵敏度为73.5%，特异度为72.1%。

高HbA_1c变异性（HbA_1c-SD>0.34%，HbA_1c-CV>4.74%）与T2DM合并HFpEF患者新发AF的风险增加独立相关，并在预测该类患者发生AF中具有较大的临床价值。