目的 探讨体检人群甘油三酯/高密度脂蛋白胆固醇（triglyceride/high density lipoprotein cholesterol ratio，TG/HDL-C）、甘油三酯-葡萄糖乘积（riglyceride-glucose index，TyG）指数及其衍生指数水平与白蛋白尿异常的相关性，同时分析TG/HDL-C、TyG以及TyG-BMI（riglyceride-glucose index-body mass index，TyG-BMI）预测白蛋白尿的能力。 方法 获取2022年1月至12月体检中心年龄在18岁以上3 739例受检者进行回顾性分析，将研究对象分为白蛋白尿异常组（n=841）和对照人群（n=2 898），将TG/HDL-C、TyG以及TyG-BMI按四分位数进行分组，采用logistic回归模型分析与体检人群白蛋白尿异常风险之间的关系。采用限制性立方样条回归模型分析不同水平TG/HDL-C、BMI、TyG以及TyG-BMI与白蛋白尿的剂量反应关系。通过受试者工作特征曲线分析单一指标与联合指标对白蛋白尿的预测价值。 结果 除平均血红蛋白浓度、高密度脂蛋白胆固醇外，异常组年龄、中性粒细胞、单核细胞、血小板、血白细胞计数、平均红细胞体积、红细胞分布宽度SD、红细胞分布宽度、血小板压积、总蛋白、白蛋白、谷氨酰转肽酶、空腹血糖、糖化血红蛋白、总胆固醇、低密度脂蛋白胆固醇、甘油三酯（triglyceride，TG）、体质指数（body mass index，BMI）、收缩压、舒张压、脉率、碱性磷酸酶、TyG指数、TyG-BMI、TG/HDL-C均高于对照组。调整混杂因素后，逻辑回归结果显示，与第2、第3分位数相比，TG/HDL-C、TyG、TyG-BMI指数第4四分位数组发生风险最高，其分别为1.895倍（95%CI=1.355～2.653）、2.377倍（95%CI=1.657～3.417）、2.319（95%CI=1.666～3.237）。BMI、TyG-BMI指数与白蛋白尿异常呈明显的非线性剂量反应关系，曲线呈近似“U”型，TyG、TG/HDL-C与白蛋白尿异常之间的剂量反应关系呈逐渐递增趋势，曲线呈近似“J”型。受试者工作特征（receiver operating characteristic，ROC）曲线结果显示，TyG-BMI预测能力优于其他3个指标，其AUC值为0.640。除TyG-BMI+TG/HDL-C之外，其余联合预测能力均优于单一预测指标，TyG-BMI+TG/HDL-C+ TyG指数预测白蛋白尿的能力最优，AUC为0.6504。 结论 TyG指数、TG/HDL-C比值升高与白蛋白尿异常密切相关，均对白蛋白尿具有一定的预测价值。综合考虑，TyG-BMI+TG/HDL-C+TyG指数是较佳白蛋白尿预测联合指标。

Objective To investigate the association of triglyceride/high-density lipoprotein cholesterol ratio（TG/HDL-C），triglyceride-glucose index（TyG），and TyG-derived indices with abnormal albuminuria in a population undergoing physical examination，as well as the ability of TG/HDL-C，TyG，and triglyceride-glucose index-body mass index ratio（TyG-BMI） in predicting albuminuria. Methods A retrospective analysis was performed for 3 739 subjects，aged 18 years or older，who underwent physical examination in Physical Examination Center from January to December 2022，and the subjects were divided into abnormal albuminuria group with 841 subjects and control group with 2 898 subjects. The subjects were divided into groups based on the quantiles of TG/HDL-C，TyG，and TyG-BMI，and the logistic regression model was used to investigate their association with the risk of abnormal albuminuria in the physical examination population. The restricted cubic spline regression model was used to analyze the dose-response relationship between different levels of TG/HDL-C，BMI，TyG，and TyG-BMI and albuminuria. The receiver operating characteristic（ROC） curve was used to investigate the value of the indices used alone or in combination in predicting albuminuria. Results Except for mean corpuscular hemoglobin concentration and HDL-C，compared with the control group，the abnormal albuminuria group had significantly higher age，neutrophil count，monocytes，platelet count，leukocyte count，mean corpuscular volume，red blood cell distribution width-standard deviation，red blood cell distribution width，plateletcrit，total protein，albumin，glutamyl transpeptidase，fasting blood glucose，glycosylated hemoglobin，total cholesterol，low-density lipoprotein cholesterol，triglyceride，BMI，systolic pressure，diastolic pressure，pulse rate，alkaline phosphatase，TyG，TyG-BMI，and TG/HDL-C. After adjustment for confounding factors，the logistic regression analysis showed that the groups with the 4th quartiles of TG/HDL-C，TyG，and TyG-BMI indices had the highest risk of albuminuria compared with the groups with the 2nd and 3rd quartiles of these indices，with a risk of 1.895 times（95%CI=1.355-2.653），2.377 times（95%CI=1.657-3.417），and 2.319 times（95%CI=1.666-3.237），respectively. BMI and TyG-BMI index had a significant U-shaped non-linear dose-response relationship with albuminuria，while TyG and TG/HDL-C had a gradually increasing J-shaped dose-response relationship with albuminuria. The ROC curve analysis showed that TyG-BMI had a better predictive ability than the other three indices，with an area under the ROC curve（AUC） of 0.640. Except for the combination of TyG-BMI and TG/HDL-C，the other combinations of these indices had a better predictive ability than each index alone，among which the combination of TyG-BMI，TG/HDL-C，and TyG index had the best ability to predict albuminuria，with an AUC of 0.6504. Conclusion Increases in TyG index and TG/HDL-C ratio are closely associated with abnormal albuminuria，and both of them have a certain value in predicting albuminuria. The combination of TyG-BMI，TG/HDL-C，and TyG index is a relatively good indicator for predicting albuminuria.