检验医学 ›› 2025, Vol. 40 ›› Issue (8): 735-741.DOI: 10.3969/j.issn.1673-8640.2025.08.003

• 论著 • 上一篇    下一篇

基于实验室指标的老年人骨质疏松症列线图诊断模型的建立和验证

柯文才1, 刘洁2, 林勇3, 鲁广建1, 焦路阳1()   

  1. 1.新乡医学院第一附属医院,河南 新乡 453100
    2.上海市嘉定区中心医院检验科,上海 201899
    3.复旦大学附属上海市第五人民医院检验科,上海 200240
  • 收稿日期:2024-03-14 修回日期:2024-12-04 出版日期:2025-08-30 发布日期:2025-08-28
  • 通讯作者: 焦路阳,E-mail:jiaoluyang2009@163.com
  • 作者简介:柯文才,男,1984年生,学士,副主任技师,主要从事临床检验工作。
  • 基金资助:
    国家重点研发计划项目(2018YFC2000200)

Establishment and verification of nomogram diagnostic model of osteoporosis in elders based on laboratory indicators

KE Wencai1, LIU Jie2, LIN Yong3, LU Guangjian1, JIAO Luyang1()   

  1. 1. The First Affiliated Hospital of Xinxiang Medical College,Xinxiang 453100,Henan,China
    2. Department of Clinical Laboratory,Jiading Central Hospital of Shanghai,Shanghai 201899,China
    3. Department of Clinical Laboratory,Shanghai Fifth People's Hospital,Fudan University,Shanghai 200240,China
  • Received:2024-03-14 Revised:2024-12-04 Online:2025-08-30 Published:2025-08-28

摘要:

目的 探讨老年患者骨质疏松症(OP)的影响因素,并建立老年人OP发生风险的列线图模型。方法 选取2022年6月—2023年12月复旦大学附属上海市第五人民医院268名行健康体检的老年人作为训练集,根据是否发生OP分为OP组(60例)和非OP组(208例)。另选取2023年1—8月上海市嘉定区中心医院老年患者106例作为外部验证集。收集所有研究对象相关临床资料和实验室检测结果。采用LASSO回归分析进行变量筛选。采用多因素Logistic回归分析评估OP发生的影响因素,基于危险因素建立列线图模型。采用受试者工作特征(ROC)曲线、决策曲线、校准曲线和Hosmer-Lemeshow拟合优度检验对建立的列线图模型进行验证。结果 与非OP组比较,OP组高龄和女性所占比例升高(P<0.05),血清骨钙素(OCN)、β胶原降解产物(β-CTX)、碱性磷酸酶(ALP)、甲状旁腺激素(PTH)水平升高(P<0.05),血红蛋白(Hb)、游离三碘甲状腺原氨酸(FT3)、游离甲状腺素(FT4)、白蛋白(Alb)水平降低(P<0.05)。筛选出13个变量,其中性别、年龄、β-CTX、FT4是老年人发生OP的独立危险因素[比值比(OR)值分别为12.130、1.050、1.002、0.874,95%可信区间(CI)分别为4.633~31.755、1.005~1.097、1.001~1.003、0.765~0.999,P<0.05]。在训练集和验证集中,列线图模型诊断OP的曲线下面积(AUC)分别为0.842、0.763。列线图模型具有较好的校准度(Hosmer-Lemeshow拟合优度检验P值分别为0.873、0.733)和较高的临床效用。结论 基于性别、年龄、β-CTX和FT4建立的老年人OP发生风险列线图模型具有较高的临床应用效能,有利于判断老年人发生OP的风险。

关键词: 列线图模型, 实验室指标, 骨质疏松症, 老年人

Abstract:

Objective To investigate the influencing factors of osteoporosis(OP)in elders,and to establish and verify nomogram diagnostic model of OP in elders. Methods A total of 268 elders undergoing physical examination admitted to Shanghai Fifth People's Hospital Affiliated to Fudan University from June 2022 to December 2023 were enrolled as training group (60 cases of OP and 208 cases of non-OP),and 106 elders admitted to Jiading Central Hospital of Shanghai from January to August 2023 were enrolled as validation set. The clinical data and clinical laboratory determination results were collected. LASSO regression analysis was used to screen the influencing factors of OP. The influencing factors were found out by multi-variate Logistic regression analysis,and the nomogram diagnostic model was established based on risk factors. The established nomogram model was verified by receiver operating characteristic (ROC) curve,decision curve,calibration curve and Hosmer-Lemeshow goodness-of-fit test. Results Compared with non-OP group,the proportions of elders and females in OP group were increased (P<0.05),and the levels of serum osteocalcin (OCN),beta-collagen degradation products (β-CTX),alkaline phosphatase (ALP) and parathyroid hormone (PTH) were increased (P<0.05). The levels of hemoglobin (Hb),free triiodothyronine (FT3),free thyroxine (FT4) and albumin (Alb) were decreased (P<0.05). Totally,13 variates were screened out. Among them,gender,age,β-CTX and FT4 were independent risk factors for OP in elders [odds ratios (OR) were 12.130,1.050,1.002 and 0.874,95% confidence intervals (CI) were 4.633-31.755,1.005-1.097,1.001-1.003 and 0.765-0.999,respectively,P<0.05]. In the training set and validation set,the areas under curves (AUC) for diagnosing OP by the nomogram model were 0.842 and 0.763,respectively. The nomogram model had good calibration (the P values of Hosmer-Lemeshow goodness-of-fit test were 0.873 and 0.733,respectively) and high clinical utility. Conclusions The nomogram diagnostic model established for OP in elders has good accuracy,and it can provide auxiliary diagnosis for the occurrence of OP in elders.

Key words: Nomogram model, Laboratroy indicator, Osteoporosis, Elder

中图分类号: