不同机器学习算法的社区老年人认知衰弱风险预测模型比较

目的 基于不同机器学习算法构建社区老年人认知衰弱风险预测模型，优选最佳模型，为防范社区老年人认知衰弱提供适宜评估工具。方法 选取苏州市3个社区卫生服务中心体检的1 105名老年人，随机分为训练集773人和验证集332人。基于训练集单因素logistic回归分析结果，使用logistic回归、伯努利朴素贝叶斯、随机森林、极端梯度提升、K邻近算法和支持向量机6种机器学习算法构建6种认知衰弱风险预测模型，并在验证集中进行评价和比较。基于最优算法构建社区老年人认知衰弱评分表，并对评分表进行验证。结果 训练集单因素logistic回归分析共识别出13个危险因素，6种模型ROC曲线下面积0.824～0.889,敏感度0.721～0.849,特异度0.700～0.837,约登指数0.498～0.674;随机森林模型为最佳模型，基于此模型构建的老年人认知衰弱评分表得分范围0～180分，内、外部验证ROC曲线下面积为0.858、0.831,最佳截断值为75分。结论 基于随机森林算法建立的预测模型预测性能最好，基于logistic回归建立的模型效果最差。基于随机森林算法构建的社区老年人认知衰弱评分表可用于...     

产后抑郁风险预测模型的构建和验证

目的 构建产后抑郁风险预测模型，并识别预测因子。 方法 选取住院分娩产妇835人为研究对象，按照时间段分为训练集722人及测试集113人，以产后6周是否发生产后抑郁为结局指标。利用logistic回归、支持向量机和随机森林3种监督学习算法建立风险预测模型，采用序列前向选择法筛选特征，通过网格搜索法调整模型参数。将训练好的模型在训练集上进行十折交叉验证，在测试集上进行外部验证。 结果 产妇产后6周抑郁发生率为22.6%（189/835）。经筛选，最终纳入14个预测因子。3种监督学习模型中，随机森林模型预测性能最佳，在测试集上的受试者工作特征曲线下面积、Brier得分、准确率、精确度、召回率和F1得分分别为0.943、0.073、0.903、0.684、0.722和0.703。 结论 基于随机森林的产后抑郁风险模型预测性能最佳，能够辅助医护人员识别高风险人群。     

苏州地区老年人疾病史与骨质疏松的相关性研究

目的通过流行病学调查苏州市社区老年人疾病史与骨质疏松的相关性。方法采用现场问卷调查了解受试者的基本资料(包括性别、年龄、身高、体重)及疾病史(包括高血脂及用药、胃/十二指肠切除、甲亢/甲低、肾功能不全、贫血、恶性肿瘤、皮质激素使用和其他疾病),并用双能X线骨密度仪(Hologic Wi)测量5 527例大于65岁老年人群的髋部、腰椎和股骨颈骨密度(bone mineral density,BMD)T值,然后以骨密度T值为因变量进行一元回归分析及全子集回归变量选择,并使用广义线性回归模型基于Elastic Net进行多元因变量选择。结果切除胃/十二指肠、其他骨科疾病及是否服用降血脂药纳入以髋部骨密度T值为因变量的最优模型(调整的判定系数最优)。高血脂、甲亢/甲低、肾功能不全、贫血纳入以腰椎BMD T值为因变量的最优模型(调整的判定系数最优);切除胃/十二指肠、肾功能不全、贫血纳入以股骨颈骨密度T值为因变量的最优模型(调整的判定系数最优); Elastic Net变量筛选,提示高血脂、长期服用降血脂药、切除胃/十二指肠、肾功能不全、贫血、恶性肿瘤、长期服用皮质激素及其他骨科疾病与骨密度T值具有相关性,短期服用皮质激素与骨密度T值没有相关性。结论在老年群体中,多种疾病的病史与骨质疏松密切相关,提示患有所列疾病的65岁以上老年人需定期进行BMD测量和疾病筛查,以有效预防骨质疏松的发生。     

去卵巢大鼠骨质疏松检测指标的实验分析

目的：分析大鼠去势后不同时间体重、骨载荷、骨密度的变化及其相关性，为骨质疏松治疗药物研究中动物模型和实验指标的选择提供参考。方法：5月龄雌性Wistar大鼠50只，随机分为模型组和正常组。两组分别于造模后3、9、13个月检测体重、骨密度和骨载荷。结果进行统计学分析。结果：模型组体重明显高于正常组；两组腰椎载荷和骨密度出现显著性差异的时间均早于股骨；股骨骨密度出现明显下降的时间早于载荷；模型组骨密度低于正常组并与体重的增长呈负相关，模型组载荷低于正常组但与体重的增长呈正相关。结论：去卵巢大鼠体重的变化可对载荷和骨密度产生不同影响，且不同部位的骨组织和不同年龄大鼠，骨量丢失存在差异.     

经伤椎内固定联合椎体成形术治疗骨质疏松性椎体骨折后活动功能低下的影响因素

目的 探讨骨质疏松性椎体骨折(osteoporotic vertebral fractures,OVF)内固定术后日常生活活动功能(activities of daily living,ADL)低下的影响因素.方法 选择2012年3月～2019年3月在本院就诊的OVF患者作为研究对象,收集患者人口学资料、临床特征和手术情况等资料,于术后6个月采用ADL量表评估患者的生活活动能力,以生活活动功能是否良好(ADL≤60分为不良,>60分为良好)将患者分为两组,观察两组患者临床特征是否存在差异.结果 患者6个月时ADL≤60分者共31例,占35.63％.ADL良好组和ADL不良组患者的年龄、类固醇药物史、神经功能分级、骨密度值、椎体前缘压缩率、椎管占位、骨水泥渗漏、术前ODI等指标差异存在统计学意义(P<0.05).二分类多因素Logistic回归分析显示,年龄≥75岁、骨水泥渗漏、骨密度<-3.5 SD是导致ADL不良的风险因素(P<0.05),神经功能分级是其保护性因素(P<0.05).结论 OVF患者内固定术后ADL不良发生率较高,应引起临床重视,并根据患者风险因素给予相应防治措施.     

非线性回归拟合正常人群前臂远端骨密度变化

目的采用非线性回归中的最大值函数拟合正常人群前臂远端骨密度（BMD）变化。方法用pDEXA OSTEOMETER DTX-200骨密度仪测量603例5～89岁健康人前臂远端的BMD，对BMD随年龄的变化趋势采用非线性回归模型拟合并与线性回归模型作比较；根据最大值函数回归模型方程，以极大值作为峰值BMD计算出峰值BMD时相应的年龄值以及BMD下降13％和25％时相应的年龄值。结果以方程的决定系数R^2来评价，非线性回归中最大值函数回归模型的拟合效果与线性回归模型中拟合效果最佳的三次回归模型相近。结论在BMD随年龄变化的研究中可试用非线性回归中的最大值函数进行拟合分析。     

低氟饮水对饮水型地氟病区氟骨症患者骨密度的影响分析

目的了解饮水型地氟病区改水后低氟饮水对氟骨症患者骨密度的影响,为氟骨症患者早期预防和治疗骨质疏松提供科学依据。方法在陕西省大荔县随机选择50～60岁氟骨症患者43例(男20例、女23例),并随机选择在当地生活的健康居民50名(男25名,女25名)作为对照组,采用双能X线吸收法测定腰椎正位1～4椎体、左髋部包括股骨颈、大转子、转子间、髋部总体的骨密度,并比较两组骨密度和骨质疏松检出率的差异;使用多因素Logistic回归分析探讨罹患氟骨症的危险预测因素。结果饮水型地氟病区改水后,50～60岁轻、中度氟骨症患者腰椎和髋部骨密度与当地健康居民相比差异无显著统计学意义(P0.05),骨质疏松的检出率与对照组相似;多因素Logistic回归分析显示,地氟病区居住时间是罹患氟骨症的独立危险预测因素。结论该县改水降氟工作数十年来成效明显,50～60岁年龄段的轻、中度氟骨症患者的骨密度和骨质疏松的检出率与当地正常人群相仿,提示长年饮用低氟水源对轻、中度氟骨症患者的骨密度有积极作用。     

血浆色素上皮衍生因子与初发2型糖尿病患者骨密度的相关性

目的 研究初发2型糖尿病（type 2 diabetes mellitus，T2DM）患者血浆色素上皮衍生因子（pigment epithelium-derived factor，PEDF）水平与骨密度的关系。方法 纳入50例初发T2DM患者，通过问诊及查体获取其一般临床资料，采集其空腹血，检测其临床检验指标、骨代谢指标以及血浆PEDF水平，利用双能X线测量骨密度。根据骨密度是否降低分为骨量减少组和骨量正常组，比较两组人群一般临床资料、临床检验指标、骨代谢指标、骨密度、血浆PEDF水平，利用Pearson相关分析研究血浆PEDF水平与腰1～4、股骨颈以及全髋骨密度的相关性，利用单因素以及多因素线性回归研究腰1～4、股骨颈以及全髋骨密度的独立影响因素。结果 初发T2DM患者骨量减少组较骨量正常组血浆PEDF水平更低，血浆PEDF水平与腰1～4、股骨颈、全髋骨密度正相关，血浆PEDF是初发T2DM患者骨密度的独立影响因素。结论 PEDF可能作为初发T2DM患者骨密度变化的重要分子发挥作用，其有望成为治疗糖尿病性骨质疏松的新因子。     

机器学习算法和COX列线图在肝细胞癌术后生存预测中的应用价值

目的探讨机器学习算法和COX列线图在肝细胞癌术后生存预测中的应用价值。方法采用回顾性描述性研究方法。收集2012年1月至2017年1月中国医学科学院北京协和医学院肿瘤医院收治的375例肝细胞癌行根治性肝切除术患者的临床病理资料;男304例,女71例;中位年龄为57岁,年龄范围为21~79岁。375例患者通过计算机产生随机数方法以8∶2比例分为训练集300例和验证集75例,应用逻辑回归、支持向量机、决策树、随机森林、人工神经网络机器学习算法构建肝细胞癌患者术后生存的预测模型,筛选性能最优的机器学习算法预测模型;构建肝细胞癌患者术后生存预测的COX列线图预测模型;比较最优机器学习算法预测模型和COX列线图预测模型预测肝细胞癌患者术后生存的性能。观察指标:(1)训练集与验证集患者临床病理资料分析。(2)训练集与验证集患者随访及生存情况。(3)机器学习算法预测模型构建及验证。(4)COX列线图预测模型构建及验证。(5)随机森林机器学习算法预测模型与COX列线图预测模型预测性能评价。采用门诊或电话方式进行随访,了解患者生存情况。随访时间截至2019年12月或患者死亡。正态分布的计量资料以±s表示,组间比较采用配对t检验。偏态分布的计量资料以M(P25,P75)或M(范围)表示,组间比较采用Mann-Whitney U检验。计数资料以绝对数表示,当Tmin≥5,N≥40时,组间比较采用χ2检验;当1≤Tmin≤5,N≥40时,采用校正χ2检验;当Tmin<1或N<40时,采用Fisher确切概率法。采用Kaplan-Meier法计算生存率和绘制生存曲线。采用COX比例风险模型进行单因素分析,将P<0.2的变量纳入Lasso回归分析,根据λ值筛选影响预后的变量,最后将变量纳入COX比例风险模型进行多因素分析。结果(1)训练集与验证集患者临床病理资料分析:训练集和验证集患者微血管侵犯(无、有),肝硬化(无、有)分别为292、8例,105、195例和69、6例,37、38例,两组患者比较,差异均有统计学意义(χ2=4.749,5.239,P<0.05)。(2)训练集与验证集患者随访及生存情况:训练集与验证集患者均获得随访。训练集300例患者随访时间为1.1~85.5个月,中位随访时间为50.3个月。验证集75例患者随访时间为1.0~85.7个月,中位随访时间为46.7个月。375例肝细胞癌患者术后1、3年总体生存率分别为91.7%、79.5%。训练集和验证集患者术后1、3年总体生存率分别为92.0%、79.7%和90.7%、81.9%。两组患者术后生存情况比较,差异无统计学意义(χ2=0.113,P>0.05)。(3)机器学习算法预测模型构建及验证。①筛选最优机器学习算法预测模型:根据变量对预测肝细胞癌术后3年生存的信息增益度,应用逻辑回归、支持向量机、决策树、随机森林和人工神经网络5种机器学习算法对肝细胞癌临床病理因素进行变量综合排名。筛选主要预测因素为乙型肝炎e抗原(HBeAg)、手术方式、肿瘤最大直径、围术期输血、肝被膜侵犯、肝脏Ⅳ段侵犯。将预测因素前3、6、9、12、15、18、21、24、27、29个变量依次引入5种机器学习算法。其结果显示:当引入9个变量时,逻辑回归、支持向量机、决策树、随机森林机器学习算法预测模型受试者工作特征曲线的曲线下面积(AUC)趋于稳定。当引入变量>12个时,人工神经网络机器学习算法预测模型AUC波动明显,逻辑回归、支持向量机机器学习算法预测模型AUC稳定性可继续改善,而随机森林机器学习算法预测模型AUC接近0.990,说明随机森林机器学习算法预测模型为最优机器学习算法预测模型。②随机森林机器学习算法预测模型优化和验证:将预测因素29个变量依次引入随机森林机器学习算法预测模型中,构建训练集最佳随机森林机器学习算法预测模型。其结果显示:当引入变量=10个时,网格搜索法示最佳决策树结点个数=4,最佳决策树数目=1000;当引入变量≥10个时,随机森林机器学习算法预测模型AUC稳定在0.990左右。其中当引入变量=10个时,随机森林机器学习算法预测模型预测训练集术后3年总体生存AUC为0.992,灵敏度为0.629,特异度为0.996,预测验证集术后3年总体生存AUC为0.723,灵敏度为0.177,特异度为0.948。(4)COX列线图预测模型构建及验证。①训练集患者术后生存因素分析。单因素分析结果显示:HBeAg、甲胎蛋白、围术期输血、肿瘤最大直径、肝被膜侵犯、肿瘤分化程度是影响肝细胞癌患者术后生存的相关因素(风险比=1.958,1.878,2.170,1.188,2.052,0.222,95%可信区间为1.185~3.235,1.147~3.076,1.389~3.393,1.092~1.291,1.240~3.395,0.070~0.703,P<0.05)。将P<0.2的临床病理因素纳入Lasso回归分析,其结果显示:性别,HBeAg,甲胎蛋白,手术方式,围术期输血,肿瘤最大直径,肿瘤位置在肝脏Ⅴ段和肝脏Ⅷ段,肝被膜侵犯,肿瘤分化程度(高分化、中高分化、中分化、中低分化)是影响肝细胞癌患者术后生存的相关因素。进一步将上述临床病理因素纳入多因素COX回归分析,其结果显示:HBeAg、手术方式、肿瘤最大直径是肝细胞癌患者术后生存的独立影响因素(风险比=1.770,8.799,1.142,95%可信区间为1.049~2.987,1.203~64.342,1.051~1.242,P<0.05)。②COX列线图预测模型的构建和验证:将训练集COX多因素分析结果中P≤0.1的临床病理因素引入Rstudio软件及其rms软件包,构建训练集COX列线图预测模型。COX列线图预测模型预测术后总体生存的C-index为0.723(se=0.028),预测训练集术后3年总体生存AUC为0.760,预测验证集术后3年总体生存AUC为0.795。训练集校准图验证显示COX列线图预测模型对术后生存有较好预测效果。COX列线图回归函数=0.62706×HBeAg(正常=0,异常=1)+0.13434×肿瘤最大直径(cm)+2.10758×手术方式(腹腔镜=0,开腹手术=1)+0.54558×围术期输血(无输血=0,输血=1)-1.42133×高分化(非高分化=0,高分化=1)。计算所有患者COX列线图风险评分,应用Xtile软件寻找COX列线图风险评分最佳阈值,风险评分≥2.9分为高危组,风险评分<2.9分为低危组。Kaplan-Meier总体生存曲线结果显示:训练集低危组和高危组患者术后总体生存比较,差异有统计学意义(χ2=33.065,P<0.05)。验证集低危组和高危组患者术后总体生存比较,差异有统计学意义(χ2=6.585,P<0.05)。进一步采用决策曲线分析结果显示:联合HBeAg、手术方式、围术期输血、肿瘤最大直径和肿瘤分化程度因素的COX列线图预测模型预测性能优于单一因素的预测性能。(5)随机森林机器学习算法预测模型和COX列线图预测模型预测性能评价:通过对2种模型中共同含有的重要变量(肿瘤最大直径)进行分析,并将2种模型通过预测误差曲线进行比较,观察2种模型的预测差异。其结果显示:肿瘤最大直径为2.2 cm时,随机森林机器学习算法和COX列线图预测模型预测患者术后3年生存率分别为77.17%和74.77%(χ2=0.182,P>0.05);肿瘤最大直径为6.3 cm时,随机森林机器学习算法和COX列线图预测模型预测患者术后3年生存率分别为57.51%和61.65%(χ2=0.394,P>0.05);肿瘤最大直径为14.2 cm时,随机森林机器学习算法和COX列线图预测模型预测患者术后3年生存率分别为51.03%和27.52%(χ2=12.762,P<0.05)。随着肿瘤最大直径增加,2种模型预测患者生存率差异增大。验证集中,随机森林机器学习算法预测模型预测患者术后3年总体生存AUC为0.723,COX列线图预测模型预测患者术后3年总体生存AUC为0.795,两者比较,差异有统计学意义(t=3.353,P<0.05)。采用Bootstrap交叉验证结果显示:随机森林机器学习算法预测模型和COX列线图预测模型预测3年生存的整合Brier得分分别为0.139、0.134,COX列线图预测模型预测误差低于随机森林机器学习算法预测模型。结论与机器学习算法预测模型比较,COX列线图预测模型预测肝细胞癌术后3年生存性能更佳,且其变量少,易于临床使用。     

影响安徽安庆地区农村男性骨密度相关因素研究

目的研究影响安庆地区农村男性股骨骨密度相关因素。方法采用与骨质疏松相关的生活方式情况调查表,用双能X线吸收仪对377名男性研究对象进行股骨扫描,用多元线性回归模型分析男性一般特征及劳动强度、吸烟、饮酒、饮茶与股骨骨密度关系。结果年龄、身高、体重和腰围与股骨骨密度相关;在调整相关危险因素后,劳动强度和饮酒可增加股骨骨密度(P=0.002、0.010),而吸烟、饮茶与股骨骨密度未显示出相关性。结论在安庆地区农村男性人群中年龄、身高、体重、腰围、劳动强度和饮酒与股骨骨密度有显著相关性。     

Bone mineral density in patients on maintenance dialysis

Disorders of bone and mineral metabolism affect almost all patients with advanced chronic kidney disease (CKD). High prevalence of decreased bone mineral density has been reported in this population; however, the role and diagnostic utility of bone density measurements are not well established. The incidence of bone fractures is high in patients with ESRD, but the association between fractures and bone density is not obvious. A recent meta-analysis suggested that decreased density at the radius might be associated with higher overall fracture risk. Changes in bone mineral density reflect several underlying pathological processes, such as vitamin D deficiency, estrogen deficiency and changes in bone turnover. The response of bone to these factors and processes is not uniform: it can vary in different compartments of the same bone or in different bones of the skeleton. Therefore, it is important to differentiate between the various types of bone. This may be possible by proper selection of the measurement site or using methods such as quantitative bone computed tomography. Previous studies used different methods and measured bone mineral density at diverse sites of the skeleton, which makes the comparison of their results very difficult. The association between changes in bone mineral metabolism and cardiovascular mortality is well known in ESRD patients. Studies also suggest that low bone density itself might be an indicator for high risk of cardiovascular events and poor overall outcome in this population. Some of the risk factors of low bone mineral density, such as vitamin D or estrogen deficiency, are potentially modifiable. Further studies are needed to elucidate if interventions modifying these risk factors will have an impact on clinical outcomes. In this review, we discuss the options for and problems of assessment of bone density and summarize the literature about factors associated with low bone density and its link to clinical outcomes in patients on maintenance dialysis.     

A nomogram for predicting lifetime hip fracture risk from radius bone mineral density and age

A statistical model for predicting a woman's lifetime risk of hip fracture using her bone mineral density at menopause has been proposed by Black et al. (1992b). We made an additional assumption concerning the correlation of bone mineral density between any two ages among postmenopausal women and applied the modified model to baseline ages between 50 and 85 years and any bone mineral density level likely to be observed in the population. The results are displayed in a form more convenient for application of this model in the clinical setting.     

Prediction Models of Prevalent Radiographic Vertebral Fractures Among Older Women

It is unknown how well prediction models incorporating multiple risk factors identify women with radiographic prevalent vertebral fracture (PVFx) compared with simpler models and what their value might be in clinical practice to select older women for lateral spine imaging. We compared 4 regression models for predicting PVFx in women aged 68 y and older enrolled in the Study of Osteoporotic Fractures with a femoral neck T-score ≤ −1.0, using area under receiving operator characteristic curves (AUROC) and a net reclassification index. The AUROC for a model with age, femoral neck bone mineral density, historical height loss (HHL), prior nonspine fracture, body mass index, back pain, and grip strength was only minimally better than that of a more parsimonious model with age, femoral neck bone mineral density, and historical height loss (AUROC 0.689 vs 0.679, p values for difference in 5 bootstrapped samples <0.001–0.35). The prevalence of PVFx among this older population of Caucasian women remained more than 20% even when women with low probability of PVFx, as estimated by the prediction models, were included in the screened population. These results suggest that lateral spine imaging is appropriate to consider for all Caucasian women aged 70 y and older with low bone mass to identify those with PVFx.     

Low bone mineral density in COPD patients related to worse lung function,low weight and decreased fat-free mass

Summary Low bone mineral density is frequently seen in COPD patients. Advanced COPD, low BMI and muscle depletion are risk factors for developing low bone mineral density (BMD). Low bone mineral density is seen in 75% of the GOLD stage IV patients. Introduction We set out to investigate the prevalence of low bone mineral density (BMD) in chronic obstructive pulmonary disease (COPD) as well as the predictors of abnormal bone mineral density. Methods A cross-sectional design was used to evaluate 115 subjects with COPD (GOLD stages II–IV). Bone mineral density (BMD) was measured using an ultrasound densitometer. The forced expiratory volume in 1 s (FEV₁) was assessed and fat-free mass was measured using bioelectrical impedance analysis. Chi-square tests and logistic regression were used for analysis. Results The prevalence of a T-score < −1.0 SD and > −2.5 SD was 28.6% in GOLD stage II, 40.3% in GOLD stage III and 57.1% in GOLD stage IV. The prevalence of a T-score ≤−2.5 SD was 0% in GOLD stage II, 9.6% in GOLD stage III and 17.9% in GOLD stage IV. In a logistic model FFM, BMI and FEV₁ were significant predictors of abnormal bone mineral density. Patients in GOLD stage IV have a 7.6 times greater risk of abnormal bone mineral density than patients in GOLD stage II. Conclusions Low bone mineral density is frequently present in COPD patients. Low FFM, BMI and FEV₁ are risk factors for developing a low T-score. A low FFM or BMI in GOLD stage IV strongly suggests loss of BMD and warrants further examination. An erratum to this article can be found at     

Nonresponders to Osteoporosis Therapy

The goal of osteoporosis therapy is reduction of fracture risk. In randomized controlled trials, relative risk of fracture is determined by comparing the absolute fracture rate of a treatment group to a control group. Fracture risk cannot be measured in individual patients being treated for osteoporosis. Since osteoporosis is a silent disease, and some patients may not respond to therapy, a surrogate test for reduction of fracture risk is often used-most commonly a bone density test. A proposed definition of nonresponse is: A decrease in bone mineral density greater than the Least Significant Change at the 95% level of confidence. The Least Significant Change is a value based on bone density measurements in patients and calculated according to well-established standards. There are other candidates for measuring responsiveness to therapy, most notably biochemical markers of bone metabolism, but none is as well validated or standardized as bone density testing. Causes of nonresponse include poor adherence, co-morbid conditions, calcium and vitamin D deficiency, malabsorption, metabolic factors, wrong dose, wrong dosing interval, and lack of efficacy. A bone density increase or stability of bone density is associated with fracture risk reduction in approved osteoporosis therapies, while a bone density decrease is cause for clinical concern. The proposed definition of nonresponse identifies a subset of patients who may require a change of therapy and/or additional medical intervention. More data are needed to develop guidelines for clinicians. Further study is suggested.     

骨质疏松性胸腰椎骨折椎体强化术后发生邻近椎体骨折的危险因素分析

目的 探讨骨质疏松性椎体压缩性骨折(osteoporotic vertebral compression fracture,OVCF)病人椎体强化术治疗后发生邻近椎体骨折的高危因素.方法 回顾性研究2012年3月至2014年8月苏北人民医院骨科收治的OVCF病人200例(263椎),收集病人的年龄、性别、椎体高度恢复、Cobb角、脊柱侧凸畸形、骨折病史、骨水泥量、骨水泥渗漏、骨密度等资料,应用单因素分析观察每种因素与椎体再骨折发生的相关性,筛查出可疑的相关因素,然后采用多因素Logistic回归分析得出影响椎体强化术后发生邻近椎体骨折的高危因素.结果 所有病人均获2年以上随访,平均随访时间为2.5年.共35例(45椎)发生再骨折,再骨折率为17.5%.单因素统计分析发现对椎体成形术后邻近节段再发骨折有影响的变量有:年龄、椎体高度恢复、Cobb角恢复、脊柱侧凸畸形、骨折病史、骨水泥渗漏、骨密度.多因素Logistic回归分析结果显示,年龄(OR:1.08,95%CI:1.04~1.13)、椎体高度恢复(OR:1.06,95%CI:1.01~1.11)、Cobb角(OR:4.03,95%CI:1.21~13.40)、脊柱侧凸畸形(OR:2.56,95%CI:1.12~5.85)和发生骨水泥渗漏(OR:6.25,95%CI:0.04~0.73)是发生再骨折的危险因素,而骨密度(OR:0.37,95%CI:0.22~0.65)是发生再骨折的保护因素.结论 年龄越大、椎体高度恢复越高、Cobb角越大、骨密度越低、有脊柱侧凸畸形和骨水泥渗漏的病人更容易发生术后邻近椎体再骨折.     

Prediction of lymphatic invasion/lymph node metastasis, recurrence, and survival in patients with gastric cancer by cDNA array-based expression profiling

BACKGROUND: We assessed the predictability of various classes of gastric carcinoma defined by clinicopathological parameters, such as invasiveness and clinical outcomes, using cDNA array data obtained from 54 cases. MATERIALS AND METHODS: We searched an optimal combination of genes to discriminate the classes defined with the clinicopathological parameters by using a feature subset selection algorithm, which was applied to a set of genes preselected on the basis of statistical difference in expression (two-sided t test, P < or = 0.05). With the selected features (gene set), we evaluated the predictability of each parameter in a leave-one-out cross-validation test. RESULTS: We successfully selected sets of genes for which the classifier predicted better versus worse overall survival (tumor-specific death) and tumor-free survival (recurrence), with respective classification rates of 94 and 92%. A contingency table analysis (chi2 test) and Cox proportional hazard model analysis revealed that lymph node metastasis is the most important factor (confounding factor) in patients' prognoses and risks of recurrence. The feature subset selection procedure successfully extracted expression patterns characteristic of lymph node metastasis and lymphatic vessel invasion, yielding 92 and 98% prediction accuracies for these respective factors. CONCLUSION: We conclude that expression profiling using feature subset selection provides a powerful means of stratification of gastric cancer patients in regard to the prognostic factors. Further studies should be warranted to apply this method to personalization of the treatment options.     

Bone mineral density testing in healthy postmenopausal women. The role of clinical risk factor assessment in determining fracture risk.

The ease of measurement and the quantitative nature of bone mineral densitometry (BMD) is clinically appealing. Despite BMD's proven capability to stratify fracture risk, data indicate that clinical risk factors provide complementary information on fracture susceptibility that is independent of BMD. Methods to quantify fracture risk using both clinical and BMD variables would have great appeal for clinical decision-making. We describe a procedure for quantifying hip fracture risk (5-yr and remaining lifetime) based on (1) the individual's age alone (base model, assuming average clinical risk factors and bone density), (2) incorporation of multiple patient-specific clinical risk factor data in the base model, and (3) incorporation of both patient-specific clinical risk factor data and BMD results.     

Relative contributions of bone density, bone turnover, and clinical risk factors to long-term fracture prediction.

Long-term fracture prediction using bone mineral density remains controversial, as does the additional contribution from assessing bone turnover or clinical risk factors. We measured bone mineral density at various sites, along with biochemical markers of bone turnover, sex steroid levels, and over 100 clinical variables, at baseline on an age-stratified sample of 304 Rochester, MN women in 1980. The 225 postmenopausal women were subsequently followed for 3146 person-years (median, 16.2 years per subject), wherein they experienced 302 new fractures: 81% resulted from minimal or moderate trauma and 60% of these involved the proximal femur, thoracic or lumbar vertebrae, or distal forearm. Accounting for multiple fractures per subject, these osteoporotic fractures together were best predicted by baseline femoral neck bone mineral density (age-adjusted hazard ratio [HR] per SD decrease, 1.37; 95% CI, 1.10-1.70); 19 moderate trauma forearm fractures were best predicted by distal radius bone mineral content, whereas 28 hip fractures and 100 vertebral fractures were best predicted by femoral neck bone mineral density. Femoral neck bone mineral density performed comparably in predicting osteoporotic fracture risk within the first decade of follow-up (HR, 1.38; 95% CI, 1.10-1.74) as well as more than 10 years after baseline (HR, 1.39; 95% CI, 1.05-1.84). The older biochemical markers were not associated with fractures, but serum "free" estradiol index was independently predictive of short- and long-term fracture risk. Consistent clinical risk factors were not identified, but statistical power was limited. Identifying patients at increased long-term risk of fracture is challenging, but it is reassuring that femoral neck bone mineral density can predict osteoporotic fractures up to 20 years later.     

Using Risk Factors and Quantitative Ultrasound to Identify Postmenopausal Caucasian Women at Risk of Osteoporosis

There is a need to prescreen large numbers of individuals for osteoporosis due to current demands on clinical resources. Some previous attempts to predict individuals at risk have used simple indices based on patient information, or Quantitative Ultrasound (QUS) and have shown good sensitivity but also demonstrated low specificity, which means that many individuals with good bone mineral density were also selected. The aim of this study was to determine if a tool based on a combination of risk factors and QUS measurements could also be made to provide improved specificity. A risk factors measurement questionnaire was created and completed for a sample of Caucasian postmenopausal women (n = 235) who had undergone Dual-energy X-ray absorptiometry scanning. QUS measurements were also taken at various skeletal sites. Assessment tools were generated using stepwise regression to predict osteoporosis, evaluated by receiver operating characteristic curves, and assessed using area under the curve values. Specificity values were determined at a sensitivity of 0.90 to establish the comparative utility of each assessment tool. Using only a risk factors model the specificities were 0.28 at the lumbar spine, 0.45 for the femoral neck and 0.68 for the total hip. In a risk factors + QUS data model the specificities measured were 0.44 for the lumbar spine, 0.78 for the femoral neck, and 0.84 for the total hip. These novel assessment tools can identify those with low bone mineral density at a number of skeletal sites and help towards avoiding many unnecessary investigations in the future.