医学
四分位数
逻辑回归
体质指数
内科学
肌萎缩
可能性
优势比
物理疗法
瘦体质量
肌肉团
糖尿病
曲线下面积
接收机工作特性
质量指数
全国健康与营养检查调查
横断面研究
临床决策
风险评估
回归分析
心脏病学
线性回归
血压
队列研究
梅德林
试验预测值
骨骼肌
生物电阻抗分析
Lasso(编程语言)
疾病严重程度
作者
Honglei Wang,Yefeng Shen
标识
DOI:10.1177/15347346251409496
摘要
ObjectiveTo examine the association between skeletal muscle mass index (SMI) and lower extremity ulcer (LEU) risk, and to develop and externally validate a prediction model incorporating SMI.MethodsWe analyzed NHANES 1999-2004 data and validated findings in a hospital-based cohort. LEUs included diabetic foot, venous, arterial, and pressure ulcers. SMI was calculated from DXA-derived appendicular lean mass. Logistic regression assessed SMI-LEU associations; restricted cubic splines tested non-linearity. LASSO identified predictors for a nomogram, evaluated using AUC, calibration, and decision curve analysis (DCA).ResultsAmong 7194 participants, lower SMI quartiles were associated with higher LEU odds (Q1: OR = 1.74, 95% CI 1.06-2.86; Q2: OR = 1.76, 95% CI 1.03-3.03 vs Q4). No non-linear association was observed. The final model (SMI group, marital status, smoking, diabetes, CKD, RBC) showed acceptable discrimination (AUC: 0.618 training; 0.682 validation) and acceptable calibration, although its clinical applicability remains limited.ConclusionLower SMI showed an independent association with increased LEU risk, and the exploratory model incorporating SMI and clinical factors provides preliminary evidence for its potential use in future risk assessment research.
科研通智能强力驱动
Strongly Powered by AbleSci AI