切换至 "中华医学电子期刊资源库"

中华关节外科杂志(电子版) ›› 2024, Vol. 18 ›› Issue (05) : 596 -603. doi: 10.3877/ cma.j.issn.1674-134X.2024.05.006

临床论著

类风湿关节炎患者骨质疏松症风险预测列线图模型构建
蒲彦婷1,(), 吴翠先1, 兰玉梅1   
  1. 1.621701 江油市第二人民医院
  • 收稿日期:2023-12-19 出版日期:2024-10-01
  • 通信作者: 蒲彦婷

Nomograph model for predicting risk of osteoporoticpatients with rheumatoid arthritis

Yanting Pu1,(), Cuixian Wu1, Yumei Lan1   

  1. 1.Jiangyou Second People's Hospital, Jiangyou 621701, China
  • Received:2023-12-19 Published:2024-10-01
  • Corresponding author: Yanting Pu
引用本文:

蒲彦婷, 吴翠先, 兰玉梅. 类风湿关节炎患者骨质疏松症风险预测列线图模型构建[J]. 中华关节外科杂志(电子版), 2024, 18(05): 596-603.

Yanting Pu, Cuixian Wu, Yumei Lan. Nomograph model for predicting risk of osteoporoticpatients with rheumatoid arthritis[J]. Chinese Journal of Joint Surgery(Electronic Edition), 2024, 18(05): 596-603.

目的

探讨类风湿关节炎(RA)患者发生骨质疏松症的危险因素并构建预测模型。

方法

选取2021年1月至2022年6月江油市第二人民医院收治的520例RA患者为研究对象,患者均进行骨密度检查且自愿参加本研究。排除有肾功能和甲状腺功能障碍、有其他恶性肿瘤及内分泌系统疾病、骨关节炎、孕妇等患者。按照7∶3的比例随机分为建模组(n=364)及验证组(n=156),根据是否发生骨质疏松症将患者分为发生组和未发生组;收集患者年龄、性别、病史、吸烟、饮酒、病程、实验室指标等临床资料,分别采用单因素和多因素logistic回归分析影响RA患者骨质疏松症发生的危险因素;构建预测RA患者骨质疏松症发生的列线图模型,绘制受试者工作特征(ROC)曲线验证模型的区分度,采用校正曲线、Hosmer-LemeshowH-L)检验评价列线图模型的校准度。

结果

520例患者骨质疏松症发生率为44.8%(233/520);建模组骨质疏松症发生率为45.1%(164/364),验证组骨质疏松症发生率为44.2%(69/156)。多因素logistic回归分析结果显示,摄入钙[比值比(OR)=2.439]、使用糖皮质激素史(OR=2.552)、年龄(OR=1.151)、身体质量指数(BMI)(OR=0.882)、病程(OR=1.071)、红细胞沉降率(ESR)水平(OR=1.057)、28个关节的疾病活动评分(DAS28)(OR=2.386)均为骨质疏松症发生的独立影响因素(均为P<0.05)。基于以上危险因素建立预测RA患者骨质疏松症发生风险的列线图模型,并进行内外部验证,结果显示建模组和验证组校正曲线和理想曲线拟合度均较好,表明模型预测骨质疏松症发生风险与实际发生风险基本一致;ROC曲线下面积分别为0.907[95%置信区间(CI)(0.876,0.939)]、0.899[95%CI0.850,0.948)],表明预测模型具有良好的预测能力。

结论

RA患者发生骨质疏松症的影响因素包括年龄、BMI、病程、有无糖皮质激素使用史、摄入钙水平、ESR水平、DAS28评分,基于以上因素构建的预测模型可有效预测RA患者发生骨质疏松症的风险,有助于临床医师早期识别RA骨质疏松症患者。

Objective

To explore the risk factors of osteoporosis in patients with rheumatoid arthritis(RA) and build a prediction model.

Methods

A total of 520 RA patients admitted to the Second People's Hospital of Jiangyou City from January 2021 to June 2022 were selected as the study objects. All the patients underwent bone mineral density examination and volunteered to participate in the study. Exclusion criteria:renal or thyroid dysfunction, malignant tumors, endocrine diseases, osteoarthritis, pregnancy. The patients were randomly divided into modeling group (n=364) and verification group (n=156) according to the ratio of 7∶3. According to the occurrence of osteoporosis, the patients were divided into developing group and nondeveloping group. Clinical data such as age, gender, medical history, smoking, drinking, course of disease and laboratory indicators were collected. Univariate and multivariate logistic regression were used to analyze the risk factors affecting the occurrence of osteoporosis in RA patients. The calibration curve and Hosmer-LemeshowH-L) test were used to evaluate the calibration degree of the calibration model, and the receiver operating characteristic (ROC) curve was drawn to verify the differentiation of the model.

Results

The incidence of osteoporosis in 520 patients was 44.8% (233/520). The incidence of osteoporosis in the modeling group was 45.1% (164/364). The incidence of osteoporosis in the verification group was 44.2% (69/156). The results of multivariate logistic regression analysis showed that calcium intake[odds ratio (OR)=2.439], history of glucocorticoid use (OR=2.552), age (OR=1.151), body mass index (BMI) (OR=0.882), duration of disease(OR=1.071), erythrocyte sedimentation rate (ESR) level (OR=1.057), disease activity in 28 joints DAS28 score(OR=2.386) was an independent factor for the occurrence of osteoporosis (P<0.05). Based on the above risk factors, a nomogram model was established to predict the risk of osteoporosis in RA patients, and internal and external verification was carried out. The results showed that the calibration curve and ideal curve fit well in both the modeling group and the verification group, indicating that the risk of osteoporosis predicted by the model was basically consistent with the actual risk. The areas under ROC curve were 0.907[ 95% confidence interval (CI)(0.876, 0.939)]and 0.899[ 95%CI (0.8850, 0.948)]respectively, indicating that the prediction model had good prediction ability.

Conclusions

The factors influencing the development of osteoporosis in patients with RA include age, BMI, course of disease, history of glucocorticoid use, calcium intake level, ESR level, and DAS28 score. The prediction model based on the above risk factors can effectively predict the risk of osteoporosis in RA patients, which is helpful for clinicians to identify RA patients with early osteoporosis.

表1 建模组和验证组的一般资料
Table 1 General data of the modeling group and the validation group
因素Factors 建模组Modeling group 验证组Validation group 统计值Statistical value P
例数Number of cases 364 156
性别[例(%)]Gender
男Male 166(45.6) 63(40.4) χ 2=1.207 >0.05
女Female 198(54.4) 93(59.6)
骨折史[例(%)]History of bone fracture
有Yes 68(18.7) 22(15.4) χ 2=1.600 >0.05
无No 296(81.3) 134(84.6)
骨质疏松家族史[例(%)]Family history of osteoporosis
有Yes 107(29.4) 34(21.8) χ 2=3.192 >0.05
无No 257(70.6) 122(78.2)
累计关节数[例(%)]Cumulative joint number
<3个 120(33.0) 53(34.0) χ 2=0.050 >0.05
≥3个 244(67.0) 103(66.0)
绝经[例(%)]Menopause
有Yes 130(65.7) 60(64.5) χ 2=0.036 >0.05
无No 68(34.3) 33(35.3)
吸烟[例(%)]Smoking
有Yes 135(37.1) 46(29.5) χ 2=2.780 >0.05
无No 229(62.9) 110(70.5)
饮酒[例(%)]Tipple
有Yes 129(35.4) 47(30.1) χ 2=1.376 >0.05
无No 235(64.6) 109(69.9)
钙摄入[例(%)]Calcium intake
≥600 mg 312(85.7) 126(80.8) χ 2=2.010 >0.05
<600 mg 52(14.3) 30(19.2)
使用糖皮质激素史[例(%)]History of glucocorticoid
有Yes 101(27.7) 36(23.1) χ 2=1.227 >0.05
无No 263(72.3) 120(76.9)
年龄[岁,(x¯±s)]Age(years) 56.8±6.2 55.9±6.3 t=1.586 >0.05
BMI[kg/m2,(x¯±s)] 23.3±3.4 23.5±3.3 t=0.492 >0.05
病程[个月,(x¯±s)]Course of disease (month) 40.3±7.9 41.3±7.7 t=1.441 >0.05
抗CCP抗体[U/ml,(x¯±s)]Anti-CCP antibody 371.5±58.9 368.6±55.7 t=0.528 >0.05
RF[IU/ml,(x±s)] 75.2±20.9 73.47±19.9 t=0.862 >0.05
25(OH)D[μg/L,(x¯±s)] 14.6±4.2 14.1±4.3 t=1.034 >0.05
ESR[mm/h,(x¯±s)] 58.4±11.0 56.89±10.8 t=1.414 >0.05
CRP[mg/L,(x¯±s)] 15.6±4.5 15.1±4.5 t=1.111 >0.05
亮氨酸氨基肽酶[U/L,(x¯±s)]Leucine aminopeptidase 25.5±6.9 24.7±6.9 t=1.231 >0.05
DAS28评分[分,(x¯±s)] 5.9±1.3 5.8±1.3 t=0.732 >0.05
Sharp评分[分,(x¯±s)]Sharp score 66.7±17.1 68.8±16.9 t=1.164 >0.05
HAQ评分[分,(x¯±s)]HAQ score 1.5±0.5 1.5±0.5 t=0.407 >0.05
表2 建模组骨质疏松症发生的单因素分析
Table 2 Univariate analysis of osteoporosis occurrence in the modeling group
因素Factors 例数Number of cases 骨质疏松组Osteoporosis group 未发生组Non-osteoporosis group 统计值Statistical value P
例数Number of cases 164 200
性别[例(%)]Gender
男Male 166 76(45.8) 90(54.2) χ 2=0.065 >0.05
女Female 198 88(44.4) 110(55.6)
骨折史[例(%)]History of bone fracture
有Yes 68 33(48.5) 35(51.5) χ 2=0.408 >0.05
无No 296 131(44.3) 165(55.7)
骨质疏松家族史[例(%)]Family history of osteoporosis
有Yes 107 51(47.7) 56(52.3) χ 2=0.417 >0.05
无No 257 113(44.0) 144(56.0)
累计关节数[例(%)]Cumulative joint number
<3个 120 38(31.7) 82(68.3) χ 2=12.962 <0.001
≥3个 244 126(51.6) 118(48.4)
绝经[例(%)]Menopause
有Yes 130 70(53.9) 60(46.1) χ 2=13.551 <0.001
无No 68 18(6.5) 50(73.5)
吸烟[例(%)]Smoking
有Yes 135 58(43.0) 77(57.0) χ 2=0.379 >0.05
无No 229 106(46.3) 123(53.7)
饮酒[例(%)]Tipple
有Yes 129 55(42.6) 74(57.4) χ 2=0.472 >0.05
无No 235 109(46.4) 126(53.6)
钙摄入[例(%)]Calcium intake
≥600 mg 312 131(42.0) 181(58.0) χ 2=8.303 0.004
<600 mg 52 33(63.5) 19(36.5)
使用糖皮质激素史[例(%)]History of glucocorticoid
有Yes 101 61(60.4) 40(39.6) χ 2=13.289 <0.001
无No 263 103(39.2) 160(60.8)
年龄[岁,(x¯±s)]Age(years) - 60.5±5.9 53.9±6.5 t=10.022 <0.001
BMI[kg/m2,(x¯±s)] - 22.4±3.4 24.01±3.5 t=4.383 <0.001
病程[个月,(x¯±s)]Course of disease (month) - 42.5±6.1 38.4±9.3 t=4.824 <0.001
抗CCP抗体[U/ml,(x¯±s)]Anti-CCP antibody - 354.3±52.9 385.6±63.8 t=5.036 <0.001
RF[IU/ml,(x¯±s)] - 93.1±23.7 60.4±18.6 t=14.742 <0.001
25(OH)D[μg/L,(x¯±s)] - 12.3±3.1 16.4±5.3 t=8.636 <0.001
ESR[mm/h,(x¯±s)] - 62.8±12.4 54.7±10.0 t=6.979 <0.001
CRP[mg/L,(x¯±s)] - 17.1±5.5 14.4±3.7 t=5.680 <0.001
亮氨酸氨基肽酶[U/L,(x¯±s)]Leucine aminopeptidase - 28.6±7.0 22.9±6.7 t=7.953 <0.001
DAS28评分[分,(x¯±s)] - 6.7±1.5 5.2±1.1 t=11.499 <0.001
Sharp评分[分,(x¯±s)]Sharp score - 87.0±19.2 50.3±15.4 t=20.180 <0.001
HAQ评分[分,(x¯±s)]HAQ score - 1.6±0.5 1.5±0.5 t=2.614 0.009
表3 多因素logistic回归分析影响RA患者骨质疏松症发生的危险因素
Table 3 Multivariatelogistic regression analysis of risk factors for osteoporosis in RA patients
图1 影响类风湿性关节炎患者骨质疏松症发生的预测列线图模型
Figure 1 Predictive nomogram model of influence factors in the osteoporosis development of rheumatoid arthritis patients
图2 建模组预测类风湿性关节炎患者骨质疏松症发生的校准曲线
Figure 2 Calibration curve of the modeling group predicting the osteoporosis onset in rheumatoid arthritis patients
图3 验证组预测类风湿性关节炎患者骨质疏松症发生的校准曲线
Figure 3 Calibration curve of the validation group predicting the osteoporosis onset f in rheumatoid arthritis patients
图4 建模组预测类风湿性关节炎患者骨质疏松症发生的ROC(受试者工作特征曲线)
Figure 4ROC of the modeling group predicting osteoporosis in rheumatoid arthritis patients
图5 验证组预测类风湿性关节炎患者骨质疏松症发生的ROC(受试者工作特征曲线)
Figure 5ROC of the validation group predicting osteoporosis in rheumatoid arthritis patients
[1]
Emery P,Vlahos B,Szczypa P,et al. Longterm drug survival of tumor necrosis factor inhibitors in patients with rheumatoid arthritis[J]. J Rheumatol,2020,47( 4 ): 493-501.
[2]
Adami G,Saag KG. Osteoporosis pathophysiology,epidemiology,and screening in rheumatoid arthritis[J/OL]. Curr Rheumatol Rep,2019,21( 7 ): 34. DOI: 10.1007/s11926-019-0836-7.
[3]
杨群政,章先锋,陈英,等. 类风湿关节炎与骨质疏松症的关系研究[J]. 中医正骨,2021,33( 2 ): 26-29.
[4]
梁美娥,张莉芸,许珂,等. 类风湿关节炎患者继发骨质疏松症高危因素的Meta分析[J]. 中国药物与临床,2017,17( 9 ):1265-1268.
[5]
中华医学会风湿病学分会. 2018中国类风湿关节炎诊疗指南[J]. 中华内科杂志,2018,57( 4 ): 242-251.
[6]
中华医学会骨质疏松和骨矿盐疾病分会. 原发性骨质疏松症诊疗指南( 2017 )[J]. 中国骨质疏松杂志,2019,25( 3 ): 281-309.
[7]
Takanashi S,Kaneko Y,Takeuchi T. CDAI and DAS28 in the management of rheumatoid arthritis in clinical practice[J]. Ann Rheum Dis,2020,79( 5 ): 671-674.
[8]
邓娟,裴必伟,徐胜前,等. 类风湿关节炎患者全身骨质疏松和局部骨侵蚀的关系研究[J]. 中华风湿病学杂志,2012,16( 10 ):674-678.
[9]
Pincus T,Summey JA,Soraci SA JR,et al. Assessment of patient satisfaction in activities of daily living using a modified Stanford Health Assessment Questionnaire[J]. Arthritis Rheum,1983,26( 11 ): 1346-1353.
[10]
Cush JJ. Rheumatoid arthritis: early diagnosis and treatment[J].Med Clin North Am,2021,105( 2 ): 355-365.
[11]
马尧,杨明义,许珂,等. 细胞焦亡与类风湿性关节炎的相关研究进展[J/CD]. 中华关节外科杂志( 电子版 ),2022,16( 5 ):586-591.
[12]
Roux C,Cortet B,Chapurlat R,et al. Management of bone fragility in patients with rheumatoid arthritis in France: an analysis of a national health insurance claims database[J/OL]. Joint Bone Spine,2022,89( 4 ): 105340. DOI: 10.1016/j.jbspin.2021.105340.
[13]
赵文娟,常志芳,赵剑波,等. 1061例类风湿关节炎患者骨密度及相关危险因素分析[J]. 中国骨质疏松杂志,2022,28( 9 ):1321-1325,1335.
[14]
Kareem R,Botleroo RA,Bhandari R,et al. The impact of rheumatoid arthritis on bone loss: links to osteoporosis and osteopenia[J/OL].Cureus,2021,13( 8 ): e17519. DOI: 10.7759/cureus.17519.
[15]
刘驰,石磊,何婧,等. 股骨颈骨折合并抑郁老年男性骨密度和骨代谢变化[J/CD]. 中华关节外科杂志( 电子版 ),2020,14( 4 ):435-439.
[16]
段梦子,张伟,秦丛丛. 老年骨质疏松症流行状况及影响因素[J]. 中国卫生工程学,2023,22( 2 ): 205-207,210.
[17]
杜秀琴,徐赛民. 绝经后类风湿关节炎患者发生骨质疏松症的危险因素分析[J]. 中国妇幼保健,2022,37( 3 ): 503-505.
[18]
孙志坚,舒强. 类风湿关节炎合并骨质疏松症及骨折的研究进展[J]. 山东第一医科大学( 山东省医学科学院 )学报,2021,42( 12 ): 947-954.
[19]
刘莺燕. 血沉参数变化与类风湿关节炎患者病情活动的关系[J]. 中国医学创新,2019,16( 21 ): 55-58.
[20]
蔡淑芬,邢其丹,丰吉南,等. 老年类风湿关节炎患者发生骨质疏松的危险因素分析[J]. 中国骨质疏松杂志,2018,24( 7 ):922-925,939.
[21]
颜小斌,吕国荣,徐振宏,等. 类风湿关节炎合并骨质疏松与7关节超声评分相关性及预测模型[J]. 中国骨质疏松杂志,2020,26( 12 ): 1737-1741.
[1] 王博冉, 乔春梅, 李春歌, 王欣, 王晓磊. 超声造影评估类风湿关节炎亚临床滑膜炎疾病进展的价值[J]. 中华医学超声杂志(电子版), 2024, 21(08): 802-808.
[2] 胡可, 鲁蓉. 基于多参数超声特征的中老年女性压力性尿失禁诊断模型研究[J]. 中华医学超声杂志(电子版), 2024, 21(05): 477-483.
[3] 许银峰, 盛璞义, 余世明, 张阳春. 偏心性髋臼旋转截骨术治疗发育性髋关节发育不良[J]. 中华关节外科杂志(电子版), 2024, 18(05): 568-574.
[4] 刘鹏, 罗天, 许珂媛, 邓红美, 李瑄, 唐翠萍. 八段锦对膝关节炎疗效的初步步态分析[J]. 中华关节外科杂志(电子版), 2024, 18(05): 590-595.
[5] 庄若语, 杭明辉, 李文华, 张霆, 侯炜. 膝骨关节炎半定量磁共振评分研究进展[J]. 中华关节外科杂志(电子版), 2024, 18(04): 545-552.
[6] 白香妮, 孙巨军, 谢鹤, 李宏斌. 急性胰腺炎患者血清微小RNA-142-3p和磷脂酰肌醇3-激酶水平变化及对并发腹腔感染风险预测[J]. 中华实验和临床感染病杂志(电子版), 2024, 18(04): 222-228.
[7] 屈勤芳, 束方莲. 盆腔器官脱垂患者盆底重建手术后压力性尿失禁发生的影响因素及列线图预测模型构建[J]. 中华腔镜泌尿外科杂志(电子版), 2024, 18(06): 606-612.
[8] 张瑜, 姜梦妮. 基于DWI信号值构建局部进展期胰腺癌放化疗生存获益预测模型[J]. 中华肝脏外科手术学电子杂志, 2024, 13(05): 657-664.
[9] 单良, 刘怡, 于涛, 徐丽. 老年股骨颈骨折术后患者心理弹性现状及影响因素分析[J]. 中华老年骨科与康复电子杂志, 2024, 10(05): 294-300.
[10] 高广涵, 张耀南, 石磊, 王林, 王飞, 郑子天, 王鸿禹, 郭民政, 薛庆云. 膝骨关节炎患者前交叉韧带功能影像学影响因素分析[J]. 中华老年骨科与康复电子杂志, 2024, 10(05): 301-307.
[11] 周锐, 罗飞. 骨质疏松椎体骨折的分型进展[J]. 中华老年骨科与康复电子杂志, 2024, 10(05): 315-320.
[12] 刘燚隆, 党荣广, 艾蓉, 张凯. 肝硬化合并静脉曲张出血患者内镜治疗后再出血风险的模型建立与验证[J]. 中华消化病与影像杂志(电子版), 2024, 14(04): 336-342.
[13] 颜世锐, 熊辉. 感染性心内膜炎合并急性肾损伤患者的危险因素探索及死亡风险预测[J]. 中华临床医师杂志(电子版), 2024, 18(07): 618-624.
[14] 于晓光, 秦永辉, 李佳, 贾国兴, 李军, 赵振栓, 刘国彬. 人工单髁置换术治疗膝关节内侧间室骨关节炎合并前交叉韧带功能不良的近期疗效[J]. 中华临床医师杂志(电子版), 2024, 18(04): 337-342.
[15] 魏贤杰, 黄河溯源, 张克石, 关振鹏. 山奈素通过抑制NF-κB减弱类风湿关节炎大鼠模型软骨细胞炎症及基质降解[J]. 中华临床医师杂志(电子版), 2024, 18(04): 375-382.
阅读次数
全文


摘要