胆固醇–高密度脂蛋白胆固醇–葡萄糖指数与中老年人心血管–肾脏–代谢共病发生风险的关联：基于CHARLS的队列研究

Association of the Cholesterol-HDL-Glucose Index with the Risk of Cardiovascular-Kidney-Metabolic Multimorbidity in Middle-Aged and Older Adults: A Cohort Study Based on CHARLS

摘要: 目的：探讨胆固醇–高密度脂蛋白胆固醇–葡萄糖指数(CHG指数)及其长期暴露模式与中国中老年人新发心血管–肾脏–代谢共病(CKM共病)风险的关联。方法：基于中国健康与养老追踪调查(CHARLS)数据，以2015年调查为基线，纳入4583名≥45岁且基线未发生CKM共病的研究对象。利用2011~2012年和2015年血液检测资料计算基线CHG指数、累积CHG指数并识别CHG在两个时间点的变化模式。采用Cox比例风险模型评估关联，并结合限制性立方样条、K-means聚类、Kaplan-Meier曲线、时间依赖ROC曲线、风险重分类分析、敏感性分析和亚组分析进行补充分析。结果：随访期间共发生648例新发CKM共病。完全调整模型中，基线CHG指数每升高1个单位，CKM共病风险升高36% (HR = 1.36, 95% CI: 1.08~1.71, P = 0.009)；累积CHG指数每升高1个单位，风险升高16% (HR = 1.16, 95% CI: 1.08~1.26, P < 0.001)。与累积CHG最低四分位组相比，Q3和Q4组风险分别升高46%和54%；与低CHG变化模式组相比，中等和高CHG变化模式组风险分别升高28%和52%。限制性立方样条提示累积CHG与CKM共病风险近似线性正相关(总体关联P = 0.002，非线性P = 0.879)。各亚组交互作用检验均无统计学意义，3年和5年时间依赖AUC为0.622~0.628；加入CHG相关指标后，风险重分类指标总体提示一定增量信息。结论：较高CHG指数，尤其是较高累积CHG水平和长期高CHG变化模式，与中老年人新发CKM共病风险升高相关。CHG指数可能为CKM共病风险分层提供补充信息，但其单独预测能力有限，仍需在独立人群中进一步验证。

Abstract: Objective: To investigate the associations of the cholesterol-HDL-glucose (CHG) index and its long-term exposure patterns with the risk of new-onset cardiovascular-kidney-metabolic multimorbidity (CKM multimorbidity) among middle-aged and older Chinese adults. Methods: Based on data from the China Health and Retirement Longitudinal Study (CHARLS), 4583 participants aged 45 years or older who were free of CKM multimorbidity at the 2015 baseline and had complete CHG exposure and follow-up outcome data were included. Blood biomarker data from 2011~2012 and 2015 were used to calculate the baseline CHG index, cumulative CHG index, and two-time-point CHG patterns of change. Cox proportional hazards models were used to evaluate associations, with restricted cubic splines, K-means clustering, Kaplan-Meier curves, time-dependent receiver operating characteristic curves, risk reclassification analysis, sensitivity analysis, and subgroup analyses as complementary analyses. Results: During follow-up, 648 participants developed CKM multimorbidity. In the fully adjusted model, each 1-unit increase in baseline CHG was associated with a 36% higher risk of CKM multimorbidity (HR = 1.36, 95% CI: 1.08~1.71, P = 0.009), and each 1-unit increase in cumulative CHG was associated with a 16% higher risk (HR = 1.16, 95% CI: 1.08~1.26, P < 0.001). Compared with the lowest cumulative CHG quartile, the Q3 and Q4 groups had 46% and 54% higher risks, respectively. Compared with the low-CHG change-pattern group, the intermediate- and high-CHG change-pattern groups had 28% and 52% higher risks, respectively. Restricted cubic splines indicated an approximately linear positive association between cumulative CHG and CKM multimorbidity risk (P for overall association = 0.002; P for nonlinearity = 0.879). No significant interactions were observed in subgroup analyses. The 3- and 5-year time-dependent AUCs ranged from 0.622 to 0.628; risk reclassification metrics generally suggested incremental information after adding CHG-related indicators. Conclusion: Higher CHG index, particularly higher cumulative CHG levels and a persistently high CHG change pattern, was associated with an increased risk of new-onset CKM multimorbidity among middle-aged and older adults. The CHG index may provide supplementary information for CKM multimorbidity risk stratification; however, its discriminatory ability as a stand-alone predictor was limited and requires further validation in independent populations.