中国全科医学 ›› 2025, Vol. 28 ›› Issue (23): 2861-2869.DOI: 10.12114/j.issn.1007-9572.2024.0527

• 论著·慢性病共病专题·糖肝共管 • 上一篇    下一篇

基于体检队列的代谢相关脂肪性肝病与高血糖关联及联合预测模型构建研究

吴莎1, 张代义2, 李晋1, 宣勤考3, 钱晓东3, 朱传武1, 浦剑虹2, 朱莉1,*()   

  1. 1.215131 江苏省苏州市,苏州大学苏州医学院附属传染病医院 苏州市第五人民医院传染病科
    2.215006 江苏省苏州市,苏州大学附属第一医院健康管理中心
    3.215006 江苏省苏州市,苏州大学附属第一医院心血管内科
  • 收稿日期:2024-11-19 修回日期:2025-02-16 出版日期:2025-08-15 发布日期:2025-06-17
  • 通讯作者: 朱莉
  • 吴莎和张代义共同为第一作者


    作者贡献:

    吴莎负责数据整理分析、论文撰写及修改;张代义负责数据整理及分析;李晋负责数据分析和论文修改;宣勤考、钱晓东负责数据收集;朱传武、浦剑虹负责研究思路指导;朱莉负责研究设计构思、研究思路指导、数据整体分析、论文修改指导。

  • 基金资助:
    江苏省社会发展面上项目(BE2022734); 苏州市科技计划项目(SKY2022061,SKY2023221,SYW2024034,SYW2024036); 王宝恩肝纤维化研究基金(CFHPC2025053); 2024年度苏州市"姑苏医星"青年科技人才托举项目(苏科协[2024]59号)

Correlation Analysis and Model Construction of Metabolic Associated Fatty Liver Disease and Hyperglycemia Based on a Health Examination Cohort

WU Sha1, ZHANG Daiyi2, LI Jin1, XUAN Qinkao3, QIAN Xiaodong3, ZHU Chuanwu1, PU Jianhong2, ZHU Li1,*()   

  1. 1. Department of Infectious Diseases, the Affiliated Infectious Disease Hospital, Suzhou Medical College, Soochow University/the Fifth People's Hospital of Suzhou, Suzhou 215131, China
    2. Health Management Center, the First Affiliated Hospital of Soochow University, Suzhou 215006, China
    3. Department of Cardiovascular Medicine, the First Affiliated Hospital of Soochow University, Suzhou 215006, China
  • Received:2024-11-19 Revised:2025-02-16 Published:2025-08-15 Online:2025-06-17
  • Contact: ZHU Li
  • About author:

    WU Sha and ZHANG Daiyi are co-first authors

摘要: 背景 代谢相关脂肪性肝病(MAFLD)与2型糖尿病(T2DM)共存的全球流行率和发生率持续上升,这种共存增加了肝脏相关不良结局的风险。临床实践中需要对MAFLD合并高血糖的高风险患者进行筛查和早期诊断,以延缓其进展。 目的 基于T2DM与MAFLD之间的关系,利用现有大规模体检数据探究高血糖对MAFLD肝脂肪变性和肝纤维化的影响,并分析影响MAFLD合并高血糖发生的关键因素。 方法 收集2024年3—7月苏州大学附属第一医院18 286名健康体检者的基本信息、既往史、腹部超声检查、生化指标和血常规项目等数据。根据腹部超声检查结果及MAFLD诊断标准,筛选出5 258例MAFLD患者,将MAFLD患者根据基于4因子的纤维化指数(FIB-4)水平分为T1组(FIB-4<1.30,n=4 275)、T2组(1.30≤FIB-4≤2.67,n=924)和T3组(FIB-4>2.67,n=59),比较三组间各临床指标差异。进一步根据有无糖尿病史、空腹血糖(FBG)≥7.0 mmol/L、糖化血红蛋白(HbA1c)≥6.5%(满足任1项)将MAFLD患者分为MAFLD合并高糖组(n=752)和MAFLD合并无糖组(n=4 506),比较两组肝脂肪变性和肝纤维化相关指标差异。采用单因素及多因素Logistic回归分析探究影响MAFLD合并高血糖发生的关键因素。采用受试者工作特征(ROC)曲线评估联合预测模型对MAFLD合并高血糖发生的预测价值。 结果 T1组、T2组、T3组吸烟、高血压、糖尿病、高脂血症、高尿酸血症、冠心病、年龄、BMI、FBG、HbA1c、血小板计数(PLT)、白细胞计数(WBC)、红细胞计数(RBC)、血红蛋白(Hb)、红细胞分布宽度(RDW)、中性粒细胞计数(NEUT)、淋巴细胞计数(LYM)、单核细胞计数(MONO)、丙氨酸氨基转移酶(ALT)、天冬氨酸氨基转移酶(AST)、总胆固醇(TC)、三酰甘油(TG)、高密度脂蛋白胆固醇(HDL-C)、低密度脂蛋白胆固醇(LDL-C)、白蛋白(ALB)、γ-谷氨酰转肽酶(GGT)、尿酸(UA)、肌酐(Cr)、尿素氮(BUN)、总胆红素(TBIL)、直接胆红素(DBIL)、间接胆红素(IBIL)、碱性磷酸酶(ALP)、预估葡萄糖处理率(eGDR)比较,差异均有统计学意义(P<0.05)。MAFLD合并高糖组脂肪肝指数(FLI)、肝脂肪变性指数(HSI)、ZJU、FIB-4、AST/PLT比值(APRI)、非酒精性脂肪肝纤维化评分(NFS)和BMI、AST/ALT及糖尿病评分(BARD)均高于MAFLD合并无糖组(P<0.05)。将MAFLD合并高糖组和MAFLD合并无糖组分别按1∶1随机纳入训练集和测试集,训练集中所有指标进行单因素及多因素Logistic回归分析,结果显示,年龄、腰围(WC)、高血压、高脂血症、TG、GGT、UA、BUN是MAFLD合并高血糖发生的影响因素(P<0.05)。进一步将以上指标进行ROC曲线分析,结果显示年龄、WC、高血压、高脂血症、TG、GGT、UA、BUN作为独立预测因子判断MAFLD合并高血糖发生具有一定的预测准确度[0.53≤ROC曲线下面积(AUC)≤0.75]。联合这8项关键预测因子构建预测模型并绘制ROC曲线,结果显示,联合模型预测准确度达到0.805(95%CI=0.781~0.828),灵敏度为75.8%,特异度为72.6%。在测试集中对该联合模型进行效能验证,结果显示阳性预测值为70.5%,阴性预测值为73.1%,预测准确率为72.7%。 结论 基于FIB-4分组的MAFLD患者,高血压、FBG、HbA1c、PLT、WBC、RBC、LYM、AST、eGDR水平在3组间有显著差异。机体高血糖能够加重MAFLD的肝脂肪变性和肝纤维化程度。此外,年龄、WC、高血压、高脂血症、TG、GGT、UA、BUN是导致MAFLD向MAFLD合并高血糖进展的影响因素。联合上述8项指标构建预测模型能够增强MAFLD人群合并高血糖发生的预测准确率,可能为临床上MAFLD人群合并高血糖的早期鉴别提供参考依据。

关键词: 代谢相关脂肪性肝病, 高血糖, 肝脂肪变性, 肝纤维化, 预测模型

Abstract:

Background

The global prevalence and incidence of metabolic associated fatty liver disease (MAFLD) co-occurring with type 2 diabetes mellitus (T2DM) are increasing, significantly elevating the risk of liver-related adverse outcomes. In clinical practice, early screening and diagnosis of high-risk MAFLD patients with hyperglycemia are crucial to slowing disease progression.

Objective

Based on the relationship between T2DM and MAFLD, this study evaluates the impact of hyperglycemia on hepatic steatosis and liver fibrosis in MAFLD using large-scale health examination data and aims to identify key factors influencing the development of MAFLD with hyperglycemia.

Methods

Data from 18 286 individuals who underwent health examinations at the First Affiliated Hospital of Soochow University from March to July 2024 were analyzed. The dataset included demographic information, medical history, abdominal ultrasound results, biochemical markers, and routine blood tests. Individuals meeting the MAFLD diagnostic criteria were classified into the MAFLD group, which was further stratified into three subgroups according to the Fibrosis-4 index (FIB-4) scores: T1 (FIB-4<1.30, n=4 275), T2 (1.30≤FIB-4≤2.67, n=924), and T3 (FIB-4>2.67, n=59). Clinical indicators among these subgroups were compared. Additionally, the MAFLD group was divided into two subgroups: MAFLD with hyperglycemia (n=752) and MAFLD without hyperglycemia (n=4 506), based on a history of diabetes, fasting blood glucose (FBG) ≥7.0 mmol/L, or glycated hemoglobin A1c (HbA1c) ≥6.5% (meeting any one criterion). Differences in hepatic steatosis and liver fibrosis-related indicators between these subgroups were analyzed. Univariate and multivariate Logistic regression analyses were performed to identify key factors associated with MAFLD with hyperglycemia. The predictive performance of a combined model for MAFLD with hyperglycemia was evaluated using the receiver operating characteristic (ROC) curve analysis.

Results

Among the T1, T2, and T3 groups, significant differences (P<0.05) were observed in clinical indicators, including smoking, hypertension, diabetes, hyperlipidemia, hyperuricemia, coronary heart disease, age, BMI, FBG, HbA1c, platelet count (PLT), white blood cell count (WBC), red blood cell count (RBC), hemoglobin (Hb), red blood cell distribution width (RDW), neutrophil count (NEUT), lymphocyte count (LYM), monocyte count (MONO), alanine aminotransferase (ALT), aspartate aminotransferase (AST), total cholesterol (TC), triglycerides (TG), high density lipoprotein cholesterol (HDL-C), low density lipoprotein cholesterol (LDL-C), albumin (ALB), glutamyl transferase (GGT), uric acid (UA), creatinine (Cr), blood urea nitrogen (BUN), total bilirubin (TBIL), direct bilirubin (DBIL), indirect bilirubin (IBIL), alkaline phosphatase (ALP), and estimated glucose processing rate (eGDR). Moreover, the fatty liver index (FLI), hepatic steatosis index (HSI), and ZJU index were significantly higher in the MAFLD with hyperglycemia group compared to the MAFLD without hyperglycemia group (P<0.05). Additionally, the FIB-4, AST/PLT ratio index (APRI), non-alcoholic fatty liver disease fibrosis score (NFS), and BMI, AST/ALT and diabetes score (BARD) were also higher in the MAFLD with hyperglycemia group (P<0.05). The samples of MAFLD with hyperglycemia and MAFLD without hyperglycemia groups were randomly divided into training and validation sets at 1∶1 ratio respectively. In the training set, univariate and multivariate Logistic regression analyses identified age, waist circumference (WC), hypertension, hyperlipidemia, TG, GGT, UA and BUN as key influencing factors associated with MAFLD with hyperglycemia (P<0.05). Further ROC analysis of these factors demonstrated moderate predictive accuracy for MAFLD with hyperglycemia (0.53≤AUC≤0.75). A predictive model incorporating these eight key factors achieved an AUC of 0.805 (95%CI=0.781-0.828), with a sensitivity of 75.8% and specificity of 72.6%. Validation of this combined model in the validation set yielded a positive predictive value of 70.5%, a negative predictive value of 73.1%, and an overall predictive accuracy of 72.7%.

Conclusion

Among MAFLD patients stratified by FIB-4, significant differences in hypertension, FBG, HbA1c, PLT, WBC, RBC, LYM, AST, and eGDR were observed across the three subgroups. Hyperglycemia exacerbates hepatic steatosis and liver fibrosis in MAFLD. Furthermore, age, WC, hypertension, hyperlipidemia, TG, GGT, UA and BUN were identified as significant risk factors for the progression of MAFLD to MAFLD with hyperglycemia. The predictive model incorporating these eight indicators enhances the accuracy of assessing hyperglycemia risk in MAFLD, potentially providing a reference for the early differential diagnosis in clinical practice.

Key words: Metabolic associated fatty liver disease, Hyperglycemia, Hepatic steatosis, Liver fibrosis, Predictive model

中图分类号: