骨质疏松症老年患者发生腰背痛的影响因素分析及预测模型的构建

Analysis of the influencing factors of lower back pain in elderly patients with osteoporosis and construction of prediction model

目的分析骨质疏松症老年患者发生腰背痛的影响因素,并构建预测模型。方法将352例老年骨质疏松症患者依据是否发生腰背痛分为疼痛组(127例,VAS评分>3分,疼痛持续时间≥3个月)、非疼痛组(225例,VAS评分≤3分)。比较两组患者一般临床资料、血常规、骨代谢指标[血清骨钙素(BGP)、β-胶原特殊序列(β-CTX)、甲状旁腺激素(PTH)、25羟维生素D3(25-(OH)D3)]、炎性因子[肿瘤坏死因子-α(TNF-α)、白细胞介素-6(IL-6)、白细胞介素-1β(IL-1β)、降钙素(PCT)、中性粒细胞/淋巴细胞比值(NLR)]、免疫功能指标[CD3^(+)、CD4^(+)、CD8^(+)、CD4^(+)/CD8^(+)]水平,采用单因素、多因素Logistic回归分析骨质疏松症老年患者发生腰背痛的影响因素,构建预测模型,ROC曲线分析预测模型的预测价值。结果与非疼痛组比,疼痛组患者年龄≥65岁占比高,β-CTX、PTH及NLR水平高,骨质疏松症病程长,腰椎骨密度值低、BGP水平低(P<0.05)。多因素Logistic回归分析显示,年龄≥65岁、骨质疏松症病程长、高水平β-CTX、高水平NLR是老年骨质疏松症患者发生腰背痛的危险因素(P<0.05),高水平腰椎骨密度、高水平25-(OH)D3是保护因素(P<0.05)。预测模型Y=10.603+0.752×年龄-0.417×腰椎骨密度+0.682×骨质疏松症病程+2.018×β-CTX-1.406×25-(OH)D3+6.409×NLR,预测患者发生腰背痛的ROC曲线下面积为0.892,敏感性为81.82%,特异性为92.06%;该模型预测发生腰背痛的预测值与实际值的一致率为98.01%。结论骨质疏松症老年患者年龄、病程、腰椎骨密度、β-CTX、25-(OH)D3及NLR水平与腰背痛的发生密切相关,由影响因素建立的预测模型可早期有效地预测骨质疏松症患者腰背痛的发生。

Objective To analyze the influencing factors of lower back pain in elderly patients with osteoporosis and construct a prediction model.Methods 352 elderly patients with osteoporosis were divided into pain group(127 cases,VAS score>3,pain duration≥3 months)and non-pain group(225 cases,VAS score≤3)according to the occurrence of low back pain.The two groups were compared in general clinical data,blood routine,bone metabolism indicators[serum bone gla protein(BGP),β-collagen specific sequence(β-CTX),parathyroid hormone(PTH)and 25-hydroxyvitamin D3(25-(OH)D3)],inflammatory factors[tumor necrosis factor-α(TNF-α),interleukin-6(IL-6),interleukin-1β(IL-1β),procalcitonin(PCT)and neutrophil to lymphocyte ratio(NLR)],and immune function indicators[CD3^(+),CD4^(+),CD8^(+)and CD4^(+)/CD8^(+)].Univariate analysis and multivariate logistic regression analysis were conducted to identify the influencing factors of lower back pain in elderly patients with osteoporosis.A prediction model was constructed,and ROC curve was used to analyze the predictive value of the model.Results Compared with the non-pain group,the proportion of patients aged 65 or older,the levels ofβ-CTX and PTH,and NLR in the pain group were higher.The course of osteoporosis in the pain group was longer.Lumbar bone mineral density and BGP level in the pain group were lower(P<0.05).Multivariate logistic regression analysis found that age≥65 years old,long course of osteoporosis,high β-CTX level,and high NLR were risk factors for lower back pain in elderly patients with osteoporosis(P<0.05),while high lumbar bone mineral density and high 25-(OH)D3 level were protective factors(P<0.05).A prediction model was constructed as follows:Y=10.603+0.752×age-0.417×lumbar bone mineral density+0.682×course of osteoporosis+2.018×β-CTX-1.406×25-(OH)D3+6.409×NLR.The AUC,sensibility and specificity for predicting low back pain were 0.892,81.82% and 92.06%.The consistency rate between the predicted and actual values was 98.01%.Conclusion Age,course of osteoporosis,lumbar bone mineral density,β-CTX level,25-(OH)D3 level and NLR level are closely related to the occurrence of low back pain in elderly patients with osteoporosis.The prediction model constructed based on influencing factors can effectively help to predict the occurrence of low back pain in patients with osteoporosis in the early stage.