目的：探讨2型糖尿病（type 2 diabetes mellitus, T2DM）老年男性患者的甲状腺激素水平与其左心室舒张功能不全（left ventricular diastolic dysfunction, LVDD）间的相关性。方法：选取2020年1月至2021年11月在我院老年病科住院的208例老年男性T2DM患者（≥60岁），根据超声心动图组织多普勒及血流多普勒结果，参照2020年《超声心动图评估心脏收缩和舒张功能临床应用指南》将其分为LVDD组（112例）和左心室舒张功能正常对照组（对照组，96例），比较2组间的基线资料、甲状腺激素水平、超声心动图指标。再根据游离三碘甲状腺原氨酸（free triiodothyronine， FT₃）四分位数水平，将全部患者分为F1组（2.43 pmol/L<FT₃≤3.30 pmol/L，52例）、F2组（3.31 pmol/L<FT₃≤3.78 pmol/L， 52例）、F3组（3.79 pmol/L<FT₃≤4.65 pmol/L， 52例）、F4组（4.66 pmol/L<FT₃≤6.01 pmol/L，52例），比较4组间的基线资料和超声心动图指标。采用Pearson相关分析观察FT₃与超声指标间的相关性，用多因素logistic回归分析FT₃与LVDD间的关系，绘制受试者工作特征（receiver operating characteristic， ROC）曲线，评估FT₃和logistic回归模型诊断LVDD的效能。结果：老年男性T2DM患者中，LVDD组较对照组FT₃水平降低[（3.51±0.24） pmol/L比(4.72±0.33) pmol/L, P<0.01]。根据FT₃水平四分位数分组，4组间的LVDD百分比（P<0.01）、超声心动图参数二尖瓣口舒张早期血流峰值速度E峰与二尖瓣环舒张早期运动峰值速度e’峰比值（E/e’）（P<0.01）差异均有统计学意义。Pearson相关分析显示，FT₃水平与E/e’、左房最大容积指数（left atrial volume index，LAVI）呈显著负相关（P均<0.001）。多因素Logistic回归分析显示，随着老年男性T2DM患者的FT₃水平递减，其发生LVDD的风险逐渐增加（F3组OR=2.29；F2组OR=3.36；F1组OR=4.95，P均<0.05）。年龄、糖化血红蛋白(glycosylated hemoglobin，HbA1c)水平、糖尿病病程、FT₃水平进入Logistic回归分析模型，ROC曲线分析显示，FT₃和Logistic回归分析模型诊断LVDD的曲线下面积为0.742和0.858；FT₃的临界值取3.37 pmol/L时，其提示患者出现LVDD的灵敏度为66.2%，特异度为75.0%。结论：老年男性T2DM患者的FT₃水平降低时，LVDD风险上升，检测FT₃可能有助于早期筛查LVDD。

Object To explore the relationship between free triiodothyronine (FT₃) level and left ventricular diastolic dysfunction(LVDD) in elderly male patients with type 2 diabetes mellitus. Methods A total of 208 elderly male (≥60 years) hospitalized patients with T2DM in Ruijin Hospital Shanghai Jiao Tong University School of Medicine were enrolled from January 2020 to November 2021. According to the Echocardiographic Assessment of Heart Disease for Systolic and Diastolic Function 2020, patients were divided into LVDD group (112 cases) and normal left ventricular fuction (control group, 96 cases). Baseline data, echocardiographic indicators, thyroid hormone indicators and thyroid hormone levels between 2 groups were compared. Meanwhile, all the patients were also divided into 4 groups according to the quartile level of FT₃: F1 group (2.43 pmol/L<FT₃≤3.30 pmol/L, n=52); F2 group (3.31 pmol/L<FT₃≤3.78 pmol/L, n=52); F3 group(3.78 pmol/L<FT₃≤4.65 pmol/L, n=52) and F4 group (4.66 pmol/L<FT₃≤6.01 pmol/L, n=52). Baseline data and echocardiographic indicators were compared between the four groups. Pearson correlation was used to analyze the correlation between FT₃ and echocardiographic indicators. Multivariate logistic regression analysis was used to explore the relationship between FT₃ level and LVDD. Receiver operating characteristic (ROC) curve was used to evaluate the efficacy of FT₃ and multivariate logistic regression model for diagnosing LVDD. Results Compared with the control group, the level of FT₃ were significantly lower in LVDD group [（3.51±0.24）pmol/L vs (4.72±0.33) pmol/L, P<0.01]. Echocardiographic indicators showed that the percentage of LVDD, ratio of peak flow velocity E peak in early diastole in mitral orifice and peak flow velocity E peak in early diastole in mitral annulus (E/e’) were significantly different (P<0.01) in the four groups by the quartile level of FT₃. Pearson correlation analysis showed that FT₃ was negatively correlated with E/e’ (P<0.001) and left atrial volume index (P<0.001). Multivariate logistic regression analysis showed that as the quartile level of FT₃ decreased, the risk of LVDD gradually increased (F3: OR=2.29; F2: OR=3.36; F1: OR=4.95; all P<0.05). Age, glycosylated hemoglobin (HbA1c), duration of diabetes mellitus and FT₃ were entered into the logistic regression analysis model. ROC results showed the AUC of FT₃ and multivariate logistic regression model for diagnosing LVDD were 0.742 and 0.858,respectively.When cutoff of FT₃ was taken as <3.37 pmol/L,the sensitivity and specificity for diagnosing LVDD were 66.2% and 75.0%,respectively. Conclusions The risk of LVDD increases as FT₃ level decreases in elderly male patients with T2DM, and FT₃ may serve as a biomarker for early diagnosing LVDD.