绝经后妇女不同部位骨密度预测椎体自发性骨折的效果分析  

Analysis of the effect of bone mineral density at different sites of postmenopausal women in predicting spontaneous vertebral fractures

作　　者：朱玉平 王大寿 淦勇 陆冬玲 ZHU Yuping;WANG Dashou;GAN Yong;LU Dongling(Department of Pain Management,Beijing Jishuitan Hospital Guizhou Hospital,Guiyang 550014;Internal Medicine WardⅠ,Beijing Jishuitan Hospital Guizhou Hospital,Guiyang 550014)

机构地区：[1]北京积水潭医院贵州医院疼痛科,贵阳550014 [2]北京积水潭医院贵州医院内科一病区,贵阳550014

出　　处：《颈腰痛杂志》2025年第2期256-261,共6页The Journal of Cervicodynia and Lumbodynia

基　　金：北京积水潭医院贵州医院国家临床重点专科建设项目(黔财社[2023]186号)。

摘　　要：目的通过Logistic回归分析不同部位的骨密度(BMD)、多部位骨密度联合对椎体自发性骨折的影响,研究其预测椎体自发性骨折的效果。方法将符合研究标准的患者分为骨折组和非骨折组,采用t检验分析两组各部位骨密度的特点,运用方差分析组内各部位BMD的差异。将不同部位的BMD分别纳入Logistic回归分析和受试者工作特征曲线(ROC)曲线分析,研究单个部位BMD预测椎体自发性骨折的效果。将不同部位BMD组合和年龄联合敏感部位的骨密度依次纳入Logistic回归模型中,分别计算其危险评分和预测概率,再根据预测概率与患者真实情况进行ROC曲线分析,最后将不同的ROC曲线进行对比分析。结果骨折组各部位BMD分别低于非骨折组(P<0.05),骨折组内不同部位BMD的方差分析提示差异有统计学意义(P<0.05),非骨折组内不同部位BMD的方差分析差异无统计学意义(P>0.05)。对骨折组内不同部位的骨密度进行两两比较分析结果显示:腰椎骨密度(LSBMD)分别低于股骨颈骨密度(FNBMD)、全髋部骨密度(THBMD),差异有统计学意义(P<0.05)。各部位BMD分别进行ROC曲线分析,结果显示均能预测椎体自发性骨折且效果显著(P<0.05),其AUC值分别为LSBMD-0.775(95%CI:0.686~0.864)、FNBMD-0.737(95%CI:0.640~0.835)、THBMD-0.753(95%CI:0.658~0.848),LSBMD骨折阈值T=-2.95,FNBMD骨折阈值T=-2.55,THBMD骨折阈值T=-2.05。将不同部位BMD联合(MSBMD)、不同年龄段联合腰椎BMD(age_LSBMD)进行Logistic回归,结果显示LSBMD、50~59岁、60~69岁是影响椎体自发性骨折的独立因素(P<0.05)。将不同部位BMD组合、age_LSBMD分别进行ROC曲线分析,其AUC值分别为:LS_FNBMD-0.769(95%CI:0.676~0.862)、LS_THBMD-0.777(95%CI:0.686~0.867)、FN_THBMD-0.757(95%CI:0.661~0.852)、MSBMD-0.768(95%CI:0.675~0.862)、age_LSBMD-0.803(95%CI:0.716~0.891)。将不同的ROC曲线进行两两比较,各ROC曲线之间的差异无统计学意义(P>0.05)。结论骨密度与椎体�Objective To examine the effectiveness of predicting vertebral spontaneous fracture by examining the impact of bone mineral density(BMD)at various sites and combined multi-site BMD through Logistic regression.Methods Patients who met the study’s eligibility requirements were split into groups according to whether they had fractures.The features of BMD at each site in the two groups were evaluated using a t-test,and the variability of BMD at each site within the groups was analyzed using an ANOVA.Logistic regression and ROC curve analysis were used to determine how well BMD at certain places could predict spontaneous vertebral fractures.The ROC curve analysis was then carried out based on the predicted probabilities and the patients’actual conditions.The BMD of various site BMD combinations and age-combined sensitive sites were sequentially included in the Logistic regression model to calculate their hazard scores and predicted probabilities,respectively.Results Each site’s BMD was lower in the fracture group than in the non-fracture group(P<0.05),and the ANOVA of BMD at various sites within the fracture group suggested that the difference was statistically significant(P<0.05),in contrast to the non-fracture group,where the ANOVA of BMD at various sites was not statistically significant(P>0.05).In the fracture group,the findings of a two-by-two comparative study of BMD at various sites revealed that the lumbar spine′s BMD(LSBMD)was lower than the femoral neck’s(FNBMD)and the total hip’s(THBMD),respectively,and these differences were statistically significant(P<0.05).ROC curve analysis was performed for each site of BMD separately,and the results showed that all predicted vertebral spontaneous fractures significantly(P<0.05),with AUC values of LSBMD-0.775(95%CI:0.686-0.864),FNBMD-0.737(95%CI:0.640-0.835),THBMD-0.753(95%CI:0.658~0.848),LSBMD fracture threshold T=-2.95,FNBMD fracture threshold T=-2.55,THBMD fracture threshold T=-2.05.According to a logistic regression analysis of combined BMD at various s

关 键 词：绝经后妇女 不同部位 骨密度 椎体自发性骨折 骨质疏松 

分 类 号：R445.9[医药卫生—影像医学与核医学]