中国全科医学 ›› 2026, Vol. 29 ›› Issue (31): 4582-4588.DOI: 10.12114/j.issn.1007-9572.2025.0581

• 论著 • 上一篇    

脂肪肝指数与健康人群代谢功能障碍相关脂肪性肝病发病风险关系的回顾性队列研究

张盼盼1, 张梓鑫2, 吴欣雨2, 韩全乐2,*(), 赵莉1, 杨静2, 李雷2, 李旭阳2, 张伯亨2, 吴寿岭3   

  1. 1.063000 河北省唐山市工人医院消化内科
    2.063000 河北省唐山市工人医院心内科
    3.063000 河北省唐山市,开滦总医院心内科
  • 收稿日期:2025-12-13 修回日期:2026-06-09 出版日期:2026-11-05 发布日期:2026-10-10
  • 通讯作者: 韩全乐

  • 作者贡献:

    张盼盼、韩全乐提出研究思路,设计方案,撰写论文;张梓鑫、吴欣雨负责研究的可行性分析、数据收集与整理、统计学处理;赵莉、杨静、李雷、李旭阳、张伯亨、吴寿岭为研究方案提供意见和指导;韩全乐对文章的知识性内容做批评性审阅,负责文章的质量控制及审校,对论文负责。

  • 基金资助:
    河北省医学课题重点研究计划项目(20231775)

A Retrospective Cohort Study on the Fatty Liver Index and Metabolic Dysfunction-associated Fatty Liver Disease in Healthy Populations

ZHANG Panpan1, ZHANG Zixin2, WU Xinyu2, HAN Quanle2,*(), ZHAO Li1, YANG Jing2, LI Lei2, LI Xuyang2, ZHANG Boheng2, WU Shouling3   

  1. 1. Department of Gastroenterology, Tangshan Gongren Hospital, Tangshan 063000, China
    2. Department of Cardiology, Tangshan Gongren Hospital, Tangshan 063000, China
    3. Department of Cardiology, Kailuan Hospital, Tangshan 063000, China
  • Received:2025-12-13 Revised:2026-06-09 Published:2026-11-05 Online:2026-10-10
  • Contact: HAN Quanle

摘要: 背景 代谢功能障碍相关脂肪性肝病(MAFLD)作为代谢障碍相关的慢性肝病,其发病率在全球范围内迅速上升。相关研究表明,脂肪肝指数(FLI)与MAFLD发病有相关性,但并没有大样本人群的流行病学研究。 目的 探讨FLI与健康人群MAFLD发病风险的关系。 方法 本研究为回顾性队列研究,选取2006年6月—2007年10月开滦研究健康体检的101 510例参与者作为研究对象。排除相关数据缺失者3 386例、严重肝肾功能不全者307例、存在MAFLD病史者29 809例、既往有恶性肿瘤者219例,最终67 789例参与者纳入本研究。根据体检FLI水平四分位数将研究对象分为4组,Q1组(n=16 947):FLI<0.21,Q2组(n=16 947):FLI 0.21~0.44,Q3组(n=16 948):FLI >0.44~0.94,Q4组(n=16 947):FLI>0.94。随访时间截至2021-12-31,每年记录1次MAFLD发病情况。计算4组MAFLD发病密度,根据Kaplan-Meier法绘制4组MAFLD累积发病率曲线,并通过Log-rank检验进行比较,采用多因素Cox回归分析探讨FLI对MAFLD发病的影响。为验证结果的稳健性,本研究排除随访时间≤2年者进行敏感性分析。绘制FLI预测MAFLD发病的受试者工作特征(ROC)曲线,并计算ROC曲线下面积(AUC)。 结果 本研究随访时间中位数为14.53(10.21,15.08)年,共29 626例(43.70%)患者发生MAFLD。Q1~Q4组MAFLD发生例数分别为3 891、6 651、8 640、10 444例,发病密度分别为17.28/千人年、31.98/千人年、44.24/千人年、56.78/千人年,累积发病率分别为22.22%、38.26%、50.08%、60.76%。Log-rank检验结果显示,4组MAFLD累积发病率比较,差异有统计学意义(χ2=5 257.65, P<0.001)。多因素Cox回归分析结果显示,与Q1组相比,Q2组(HR=1.97,95%CI=1.89~2.05)、Q3组(HR=2.83,95%CI=2.72~2.95)、Q4组(HR=3.66,95%CI=3.51~3.81)MAFLD发病风险均升高(P<0.05)。排除2年之内发病者(n=10 021)进行多因素Cox回归分析,结果显示,与Q1组比较,Q2组、Q3组、Q4组MAFLD发病风险与主分析结果相似,结果较稳健。ROC曲线结果显示,FLI预测MAFLD发病的AUC为0.674。 结论 FLI与健康人群MAFLD发病风险呈正相关,FLI可以作为健康人群早期预测MAFLD发病风险的优选指标。

关键词: 脂肪肝指数, 代谢功能障碍相关脂肪性肝病, 健康人群, 队列研究

Abstract:

Background

Metabolic dysfunction-associated fatty liver disease (MAFLD), a chronic liver condition linked to metabolic disorders, is experiencing a rapid rise in global incidence. Relevant studies indicate a correlation between the fatty liver index (FLI) and MAFLD, however, these studies lack large-scale population-based epidemiological research.

Objective

To explore the correlation between the FLI and the incidence of MAFLD in a health examination population.

Methods

This retrospective cohort study selected 101 510 participants who underwent physical examinations in the Kailuan Study from June 2006 to October 2007 as the research subjects. After excluding 3 386 participants with missing data, 307 with severe liver or kidney dysfunction, 29 809 with a history of MAFLD, and 219 with a history of malignant tumors, 67 789 participants were ultimately included in this study. Based on the quartiles of FLI levels, the participants were divided into four groups: Q1 group (n=16 947): FLI<0.21; Q2 group (n=16 947): FLI 0.21-0.44; Q3 group (n=16 948): FLI >0.44-0.94; and Q4 group (n=16 947): FLI>0.94. Follow-up continued until December 31, 2021, with annual records of MAFLD incidence. The incidence density of MAFLD was calculated for the four groups, and the cumulative incidence of MAFLD was determined using the Kaplan-Meier method and compared by Log-rank test. The impact of FLI on the risk of new-onset MAFLD was analyzed using multivariate Cox regression analysis. To verify the robustness of the results, sensitivity analyses were conducted by excluding participants with follow-up times ≤2 years. Receiver operating characteristic (ROC) curves for FLI in predicting MAFLD onset were plotted, and the area under the ROC curve (AUC) was calculated.

Results

During a median follow-up of 14.53 (10.21, 15.08) years, 29 626 patients (43.70%) developed MAFLD. The number of MAFLD cases in groups Q1 through Q4 was 3 891, 6 651, 8 640, and 10 444, The incidence density of MAFLD in the Q1-Q4 groups was 17.28/1 000 person-years, 31.98/1 000 person-years, 44.24/1 000 person-years, and 56.78/1 000 person-years, Log-rank test results showed statistically significant differences in cumulative MAFLD incidence among the four groups (χ2=5 257.65, P<0.001). Multivariable Cox regression analysis showed that, compared with the Q1 group, the risk of MAFLD onset was significantly increased in the Q2 group (HR=1.97, 95%CI=1.89-2.05), Q3 group (HR=2.83, 95%CI=2.72-2.95), and Q4 group (HR=3.66, 95%CI=3.51-3.81) (P<0.05). After excluding participants who developed MAFLD within 2 years (n=10 021), multivariable Cox regression analysis yielded results similar to the primary analysis, indicating robust findings. ROC curve analysis showed that the AUC of FLI for predicting MAFLD risk was 0.674.

Conclusion

There is a positive correlation between FLI and the risk of newly diagnosed MAFLD in a physical examination population. FLI can serve as an early predictive indicator for MAFLD in such populations.

Key words: Fatty liver index, Metabolic dysfunction-associated fatty liver disease, Healthy populations, Cohort study