基于八种胰岛素抵抗替代指数预测 中国中老年糖尿病人群肝病 发病风险——一项基于CHARLS的队列研究

Assessment of Eight Insulin Resistance Surrogate Indexes for Predicting Liver Disease Incidence in Chinese Middle-Aged and Elderly Populations with Diabetes—A Cohort Study Based on CHARLS

摘要: 背景：本研究旨在探讨八种胰岛素抵抗替代指数，包括eGDR、METS-IR、TyG指数及其变体(TyG-BMI、TyG-WHtR、TyG-WC、TyG-CVAI和CTI)，与糖尿病患者肝脏疾病风险之间的关系。研究旨在评估并比较这些指数对肝脏疾病风险的预测能力，以弥补当前该领域比较性研究的不足。方法：本研究分析了中国健康与养老追踪调查(CHARLS)的数据。采用多变量Cox比例风险回归模型分析胰岛素抵抗替代指数与肝脏疾病风险的关系。通过限制性立方样条图探索胰岛素抵抗替代指数与肝脏疾病风险之间的剂量–反应关系。利用受试者工作特征曲线分析计算各胰岛素抵抗替代指数的曲线下面积。结果：在调整混杂因素后，eGDR每增加一个单位与肝脏疾病风险降低相关(调整后HR：0.418，95% CI：0.222~0.787，P = 0.007)。相反，METS-IR、CTI、TyG、TyG-BMI、TyG-WHtR、TyG-WC和TyG-CVAI的增加则与较高的肝脏疾病风险相关，其调整后HR (95% CI)分别为1.271 (1.047~1.542)、1.345 (1.077~1.679)、1.353 (1.105~1.657)、1.337 (1.114~1.606)、1.400 (1.104~1.776)、1.652 (1.278~2.136)和1.534 (1.196~1.968)。剂量–反应分析显示，这八种胰岛素抵抗替代指数与肝脏疾病风险之间存在线性关系。ROC分析表明，在糖尿病人群中，TyG-WC对肝脏疾病的预测价值最高(AUC: 0.651)。结论：这八种胰岛素抵抗替代指数与糖尿病患者发生肝脏疾病的高风险相关。其中，TyG-WC在预测中国中老年糖尿病人群肝脏疾病风险方面表现出中等水平的预测能力。

Abstract: Background: This study examined the relationships between eight insulin resistance (IR) surrogate indexes, including eGDR, METS-IR, TyG index and its variants (TyG-BMI, TyG-WHtR, TyG-WC, TyG-CVAI, and CTI), and liver disease risk in individuals with diabetes. It aimed to assess and compare the predictive abilities of these indexes for liver disease risk, addressing the current lack of comparative studies in this area. Methods: Data from the China Health and Retirement Longitudinal Study (CHARLS) were analyzed in this study. Multivariate Cox regression models were applied to analyse the relationships of IR surrogate indexes with liver disease risk. The dose-response relationships between IR surrogate indexes and liver disease risk were explored using restricted cubic splines. The areas under the curve (AUCs) of IR surrogate indexes were calculated by receiver operating characteristic (ROC) analysis. Results: After adjusting for confounders, a unit increase in eGDR was linked to a lower risk of liver disease (adjusted HR: 0.418, 95% CI: 0.222~0.787, P = 0.007). Conversely, increases in METS-IR, CTI, TyG, TyG-BMI, TyG-WHtR, TyG-WC, and TyG-CVAI were linked to higher liver disease risk, with adjusted HRs (95% CIs) of 1.271 (1.047~1.542), 1.345 (1.077~1.679), 1.353 (1.105~1.657), 1.337 (1.114~1.606), 1.400 (1.104~1.776), 1.652 (1.278~2.136), and 1.534 (1.196~1.968), respectively. Dose-response analyses indicated a linear relationship between eight IR surrogate indexes and liver disease risk. ROC analysis showed TyG-WC had the highest predictive value for liver disease in diabetics (AUC: 0.651). Conclusion: The eight IR surrogate indexes were associated with high risk of liver disease in individuals with diabetes. Among these, TyG-WC demonstrated a moderate predictive ability for liver disease risk in Chinese middle-aged and elderly populations with diabetes.