股骨头坏死科普知识 查看全部
股骨头坏死早期能多走路吗？不行股骨头坏死是引起髋关节疼痛，活动受限的常见原因，很多是因为激素、酗酒、外伤等原因造成的。股骨头坏死早期很多人症状不明显，有时候感觉坐着起来的几步疼，然后接着走就不疼了。很多时候是休息的时候不疼，变换姿势起身走路就有点疼，有时候还疼得有点明显，但是再走就又不疼了。那么这样是不是就可以多走？尽量站着，不坐下变换体位不就不疼了吗？这样理解是不对的！股骨头坏死以后股骨头因为有坏死，股骨头强度就会有下降。尤其是负重区的坏死，在负重情况下会引起股骨头局部压力增加，会加速股骨头坏死的进展。负重不是说自己扛什么重东西，站立时候的体重压在股骨头上就是在增加股骨头的负重了。所以这样就可以理解，为什么股骨头坏死早期，即使不疼或者疼痛不严重，医生也是建议一定要扶拐保护。扶拐保护的目的就是尽量减轻体重对股骨头的压力，避免坏死的股骨头由于局部压力的增加，使得早期坏死治疗的效果大打折扣。所以股骨头坏死早期一定要重视，好好扶拐保护，减轻股骨头负重，保护好自己的股骨头北京积水潭医院矫形骨科科普号
谈谈股骨头坏死（二）一、什么样的外来因素会导致股骨头坏死，也就是股骨头内骨细胞的非正常死亡激素、酒精和创伤（绝大数多是股骨颈骨折）是临床观察到的导致股骨头坏死最常见的三大原因。教科书和文献中还记载着一些少见的原因，如镰状细胞贫血、减压病（如潜水、飞行等），但本人门诊还没有遇到过，也许是因为不太认识的缘故。还有一些是自身免疫性疾病的患者，因为大多接受过糖皮质激素的治疗，所以也无法确定是自身免疫性疾病本身还是激素治疗导致的股骨头坏死。印象中，继发股骨头坏死的自身免疫病患者中，系统性红斑狼疮似乎是最多见的。除了这些有比较明确原因的患者，还有少部分股骨头坏死患者则找不到目前已知的病因。激素是导致股骨头坏死的第一大病因。由于激素在国内使用的广泛性，包括必要的和不必要的，尤其是在国内各种不规范医疗还普遍存在的现状下，激素性股骨头坏死的数量在国内非常之多，估算年度新增病例数达10余万例。激素性股骨头坏死的患者人群中年轻人比例较高，十几、二十几岁的也有不少，这也给治疗带来了更大的挑战。从股骨头坏死这个疾病的角度来看，激素无疑是头号杀手。但另一个事实也同样存在，那就是从激素使用者的角度来看，大多数激素使用者并没有发生股骨头坏死。根据一些报道，激素使用后发生股骨头坏死的概率大概是10-30%。因此，对于那些确实有必要使用激素治疗的严重疾病，也不必因为股骨头坏死这样的并发症而影响治疗决策。自然，不必要的激素治疗应该避免。激素使用后是否发生股骨头坏死，教科书和一些文献通常认为，与使用的剂量和时长相关。但临床确实也见到一些股骨头坏死患者只有非常短暂的激素使用史,比如因为荨麻疹或者发热什么的。再结合很多长时间、大剂量使用激素的患者并没有发生股骨头坏死的事实，因此我们推测激素使用与发生股骨头坏死之间也可能存在的首先是一种“有”或“无”的关系，在“有”的群体里，才进一步存在坏死严重程度和激素使用剂量的相关性。酒精大概是导致股骨头坏死的第二大病因。酒精性股骨头坏死的患病人群大多为中年男性，具有喝酒频率高，剂量大，年头长的普遍特点。这一点与激素性股骨头坏死有较大区别，酒精性股骨头坏死与酒精摄入量之间存在更为明显的剂量依赖关系，只有足够剂量的酒精才可能导致股骨头坏死的发生。饮酒者自然远远要多于激素使用者，因此虽然说酒精是导致股骨头坏死的第二大病因，但饮酒者发生股骨头坏死还是一个很小概率的事件，具体的发生率难以统计，相比激素使用后10-30%的发生率，酒精摄入者的股骨头坏死发生率应该会小几个数量级。对于长期大量饮用酒精的人群，比股骨头坏死发生率更高、危害程度更大的并发症还有很多，比如神经系统损害、消化系统损害等等。酒精爱好者当然也不用太在意这些，因为喝酒的目的从来也不是为了健康。创伤是导致股骨头坏死的另一重大原因。创伤性股骨头坏死和酒精性股骨头坏死的患者数量哪个更多，不太确定，可能与不同的地区、不同的医疗机构有关。从本人门诊接触的患者来看，感觉两者数量似乎差不多。创伤性股骨头坏死的患者中，绝大多数是继发于股骨颈骨折，其它比较少见的原始损伤类型还包括股骨头骨折、髋关节脱位和髋臼骨折、股骨粗隆部和股骨干骨折髓内钉手术后。股骨颈骨折是一种常见骨折类型，发生人数众多，约占成人骨折总数的1.68%（据张英泽，临床创伤骨科流行病学）。GBD（国际骨折联盟）报告2019年全球新发骨折例数1.78亿，据此估算股骨颈骨折的年发生例数，全球约300万，中国约60万。而且，股骨颈骨折多发生于老年人，由于国内人口老龄化的加剧，股骨颈骨折的发生率还在持续快速增长中。接骨、外固定或内固定是骨折治疗的常规方法，但是股骨颈骨折的内固定治疗几乎是人体主要骨折中并发症最多、治疗结果最差的。也正因如此，人工关节置换已经成为了老年人股骨颈骨折的首要治疗方法。股骨颈骨折内固定治疗的并发症主要有三个，一是比其它部位骨折更高的骨折不愈合的发生率；二是股骨颈短缩，即使骨折愈合了，很多也是股骨颈短缩后的愈合。股骨颈短缩的后果是肢体变短和臀中肌力矩变小，进而导致步态异常、走路摇摆；三是股骨头坏死，这是股骨颈骨折最麻烦而又最常见的并发症，内固定后股骨头坏死的总体发生率约30%。股骨颈骨折后的股骨头坏死还存在一个非常奇怪的临床现象，那就是年轻患者的股骨头坏死发生率要高于年长患者，越年轻的股骨颈骨折患者发生股骨头坏死的可能性越大。这与常识不符，与年轻患者骨折愈合能力更强的事实也不一致，但这又是临床观察到的确实现象，也是教科书里早就记载的。对于这个现象，目前还没有一个具有足够说服力的解释。按照年龄分布，大概有10%的股骨颈骨折患者年龄小于40岁，这还不包括14岁以下儿童。因为股骨颈骨折病例基数巨大，这10%的部分也是一个数量庞大的人群。对于这部分年轻股骨颈骨折患者，其继发的股骨头坏死是一个更加严重的临床问题乃至社会问题，青少年移位型股骨颈骨折的股骨头坏死发生率可高达40-50%，治疗更是棘手，充满挑战。还有一类少见、隐蔽且易于与股骨头坏死混淆的损伤类型，是股骨头骨软骨损伤。损伤当时的常规X线片往往都是阴性表现，但是患者的疼痛很久都不能缓解。因为MRI（磁共振）影像上的表现与股骨头坏死非常相似，股骨头骨软骨损伤的患者常被认作是股骨头坏死而来就诊。这种骨软骨损伤更常见发生于距骨这个部位，也经常被认作是距骨骨坏死，这可能与我们的教科书对于骨软骨损伤和骨坏死这两个疾病的命名和阐述不够清晰有关。这两种疾病的性质显然是不一样的，骨软骨损伤实际是一种特殊类型的骨折，由关节两端的骨骼直接撞击导致，骨折块以软骨为主体、带少部分骨质。（待续）毛玉江医生的科普号
激素使用对股骨头骨坏死发生和进展的影响:一项全国性巢式病例对照研究（2024）激素使用对股骨头骨坏死发生和进展的影响:一项全国性巢式病例对照研究（2024）EffectofCorticosteroidUseontheOccurrenceandProgressionofOsteonecrosisoftheFemoralHead:ANationwideNestedCase-ControlStudy?KwonHM,HanM,LeeTS,JungI,SongJJ,YangHM,LeeJ,LeeSH,LeeYH,ParkKK.EffectofCorticosteroidUseontheOccurrenceandProgressionofOsteonecrosisoftheFemoralHead:ANationwideNestedCase-ControlStudy[J].JArthroplasty,2024转载文章的原链接1：https://pubmed.ncbi.nlm.nih.gov/38830431/转载文章的原链接2：https://www-clinicalkey-com-443.vpnm.ccmu.edu.cn/#!/content/journal/1-s2.0-S0883540324004595?AbstractBackgroundAlthoughitisverywellknownthatcorticosteroidscauseosteonecrosisofthefemoralhead(ONFH),itisunclearastowhichpatientsdevelopONFH.Additionally,therearenostudiesontheassociationbetweencorticosteroiduseandfemoralheadcollapseinONFHpatients.WeaimedtoinvestigatetheassociationbetweencorticosteroiduseandtheriskofONFHamongthegeneralpopulationandwhatfactorsaffectONFHoccurrence.Additionally,weaimedtodemonstratewhichfactorsaffectfemoralheadcollapseandtotalhiparthroplasty(THA)afterONFHoccurrence.虽然众所周知，皮质类固醇会导致股骨头骨坏死(ONFH)，但目前尚不清楚哪些患者会发生ONFH。此外，没有关于皮质类固醇使用与ONFH患者股骨头塌陷之间关系的研究。我们的目的是调查普通人群中皮质类固醇使用与ONFH风险之间的关系，以及影响ONFH发生的因素。此外，我们旨在证明哪些因素影响ONFH发生后股骨头塌陷和全髋关节置换术(THA)。?MethodsAnationwide,nestedcase-controlstudywasconductedwithdatafromtheNationalHealthInsuranceServicePhysicalHealthExaminationCohort(2002to2019)intheRepublicofKorea.WedefinedONFH(N=3,500)usingdiagnosisandtreatmentcodes.PatientswhohadONFHwerematched1:5toformacontrolgroupbasedonthevariablesofbirthyear,sex,andfollow-upduration.Additionally,inpatientswhohaveONFH,welookedforriskfactorsforprogressiontoTHA.使用韩国国民健康保险服务机构身体健康检查队列(2002年至2019年)的数据进行了一项全国性的巢式病例对照研究。我们使用诊断和治疗代码定义ONFH(N=3,500)。根据出生年份、性别、随访时间等变量对ONFH患者进行1:5匹配，形成对照组。此外，在患有ONFH的患者中，我们寻找进展为THA的危险因素。?ResultsComparedwiththecontrolgroup,ONFHpatientshadalowhouseholdincomeandhadmorediabetes,hypertension,dyslipidemia,andheavyalcoholuse(drinkingmorethan3to7drinksperweek).Systemiccorticosteroiduse(≥1,800mg)wassignificantlyassociatedwithanincreasedriskofONFHincidence.However,lipidprofiles,corticosteroidprescription,andcumulativedosesofcorticosteroiddidnotaffecttheprogressiontoTHA.与对照组相比，ONFH患者家庭收入较低，糖尿病、高血压、血脂异常和重度饮酒(每周饮酒超过3至7天)的发生率更高。全身使用皮质类固醇(≥1800mg)与ONFH发生率增加显著相关。然而，脂质谱、皮质类固醇处方和皮质类固醇累积剂量对进展到THA没有影响。?ConclusionsTheONFHriskincreasedrapidlywhencumulativeprednisoloneusewas≥1,800mg.However,oralorhigh-doseintravenouscorticosteroiduseandcumulativedosedidnotaffecttheprognosisofONFH.SincetheoccurrenceandprognosisofONFHarecomplexandmultifactorialprocesses,furtherstudyisneeded.当累积使用泼尼松龙≥1800mg时，ONFH风险迅速增加。然而，口服或大剂量静脉注射皮质类固醇和累积剂量不影响股骨头坏死的预后。由于股骨头坏死的发生和预后是一个复杂的多因素过程，需要进一步研究。?Osteonecrosisofthefemoralhead(ONFH)maycauseseverehippain,andcanultimatelyleadtocollapseofthefemoralheadandsubsequentrapiddestructionofthehipjointinrelativelyyoungpatients[1-3].Intractablepaincausedbyfemoralheadcollapseeventuallyleadstototalhiparthroplasty(THA)atayoungage[4,5].Althoughtrauma,theuseofcorticosteroid,andexcessivealcoholintakearewell-knownriskfactorsforONFH,theetiologyofONFHremainsmultifactorial,andnoconsensusexistsonthecommonpathophysiology[6-9].Impairedperfusionofbloodtobone,abnormalcellularreparativeprocesses,andgeneticpointmutationsarepresumedtoaffectONFHoccurrence,butithasnotyetbeenclarified[10-12].股骨头坏死(Osteonecrosisoffemoralhead,ONFH)可引起严重的髋关节疼痛，在相对年轻的患者中，最终可导致股骨头塌陷并随后迅速破坏髋关节[1-3]。股骨头塌陷引起的顽固性疼痛最终导致在年轻时进行全髋关节置换术(THA)[4,5]。虽然创伤、皮质类固醇的使用和过量饮酒是众所周知的ONFH的危险因素，但ONFH的病因仍然是多因素的，对共同的病理生理尚未达成共识[6-9]。据推测，骨血流灌注受损、细胞修复过程异常以及基因点突变可能影响ONFH的发生，但尚未明确[10-12]。AlthoughitisverywellknownthatcorticosteroidscauseONFH,itisunclearastowhichpatientsdevelopONFH.Thereareseveralstudiesreportingthathigh-dosecorticosteroiduseof2grams(g)ormorewithin3monthsisassociatedwithONFHdevelopment[8,13-15].However,thesewerestudieswitharelativelysmallnumberofpatients,andthereareheterogeneitiesinpatientdemographicsandepidemiologicvariabilities.OnceONFHoccurs,theprognosisisknowntobepoor,andtherearestudiesstatingthatfemoralheadcollapseprogressesfromabout24.6to42.8%withnaturalevolutionwithin2years[16,17].Inparticular,whenfemoralheadcollapseoccurs,manypatientsoftenreceiveTHAatayoungage,sopredictingtheprognosisforfemoralheadcollapseisasimportantaspredictingtheoccurrenceofONFH.SeveralstudieshavedemonstratedthatfemoralheadcollapseinONFHpatientsisassociatedwithsize,volume,locationofnecroticlesions,andanatomicalparameters[1,2,18,19].However,therearenostudiesontheassociationbetweencorticosteroiduseandfemoralheadcollapseinONFHpatients.虽然众所周知，皮质类固醇会引起ONFH，但尚不清楚哪些患者会发生ONFH。有几项研究报道，在3个月内使用2克或更多的高剂量皮质类固醇与ONFH的发生有关[8,13-15]。然而，这些研究的患者数量相对较少，并且在患者人口统计学和流行病学变异方面存在异质性。一旦发生ONFH，预后很差，有研究表明股骨头塌陷在2年内自然演化，从24.6%发展到42.8%[16,17]。特别是股骨头塌陷发生时，许多患者往往在年轻时接受THA，因此预测股骨头塌陷的预后与预测ONFH的发生同样重要。多项研究表明，ONFH患者股骨头塌陷与坏死灶的大小、体积、位置和解剖学参数有关[1,2,18,19]。然而，目前还没有关于ONFH患者使用皮质类固醇与股骨头塌陷之间关系的研究。TheNationalHealthInsuranceService(NHIS)ofSouthKoreaisanationalinsurancecoverageservicethatcoversabout97%ofcitizensinSouthKorea,includingdemographicdataandaclaimdatabaseincludingdiagnosesandprescriptions[20].Additionally,theNHISoffershealthscreeningexaminationsforinsuredKoreansaged40yearsorolder,includingbasicbloodtests,smoking,drinking,andphysicalactivitiesusingself-reportedquestionnaires.BecausetheNationalHealthInsuranceService–NationalSampleCohort(NHIS-NSC)isahighlyrepresentativepopulation-basedcohortoftheentireKoreanpopulation,itcanbeusedasapopulation-based,nationwidestudyforepidemiologicresearchonvariousdiseases[20-22].韩国国民健康保险服务(NationalHealthInsuranceService,NHIS)是一项覆盖韩国约97%公民的国民保险服务，包括人口统计数据和包括诊断和处方在内的索赔数据库[20]。此外，国民健康保险制度还为40岁或40岁以上的参保韩国人提供健康检查，包括基本的血液检查、吸烟、饮酒和使用自我报告问卷的体育活动。由于国民健康保险服务-国民样本队列(NationalHealthInsuranceService-NationalSampleCohort,NHIS-NSC)是韩国人口中极具代表性的基于人群的队列，因此可以作为一项基于人群的全国性研究，用于各种疾病的流行病学研究[20-22]。Therefore,inthepresentstudy,weaimedtoinvestigatetheassociationbetweencorticosteroiduseandtheriskofONFHdevelopmentamongthewholegeneralpopulationandwhatfactorsaffectONFHdevelopmentusinganationwidelargesamplefromlongitudinalnestedcase-controldatainalong-termcohortstudy.Inaddition,weaimedtodemonstratewhichfactorsaffectTHAafterthedevelopmentofONFH.因此，在本研究中，我们旨在通过一项长期队列研究，在全国范围内采用纵向嵌套病例对照数据的大样本，调查皮质类固醇使用与整个普通人群中ONFH发生风险之间的关系，以及影响ONFH发生的因素。此外，我们旨在证明哪些因素会影响ONFH发展后的THA。?MaterialsandMethodsStudyPopulationWeuseda2002to2019datasetfromtheNHISPhysicalHealthExaminationCohort(NHIS-NSC)2.0databaseintheRepublicofKorea[23].TheNHIS-NSC2.0databasewascomprisedofatotalof1,137,861participantswhowerestratifiedandrandomlysampledsuchthattheage,sex,region,healthinsurancetype,andhouseholdincomesofthesamplesremainedproportionaltothoseofthe2006Koreanpopulationof50million.Weexcluded223,474participantswhowerealreadyprescribedintravenous(IV)ororalcorticosteroidstoreducethebiasintheanalysisoftheeffectoftheprescribedcorticosteroiddoseonONFH.Fromtheremainingcohort,weexcluded241participantswhowerealreadydiagnosedwithONFHin2002toobtainthenewdevelopmentofONFH(washoutperiod).Therefore,theindexdatewasthedatethatthepatientwasdiagnosedwithONFH.Weincluded914,146participantsinthefinalanalysis.Theinformationinthedatasetincludedallinpatientandoutpatientmedicalclaimsdata,includingprescriptiondruguse,diagnosticandtreatmentcodes,primaryandsecondarydiagnosticcodes,andhealthexaminationdata.AlltheindividualsincludedintheNHIS-NSCwerefolloweduntil2019,unlesstherewasadeathordisqualificationforNationalHealthInsurance,suchasemigration.TheprotocolofthisstudywasapprovedbytheInstitutionalReviewBoardofourinstitution(4-2022-0304)andtheNHIS(NHIS-2022-2-232).我们使用了2002年至2019年的数据集，这些数据集来自韩国NHIS身体健康检查队列(NHIS-nsc)2.0数据库[23]。国民健康保险制度-国家安全保障制度2.0数据库共包括1137861名参与者，这些参与者的年龄、性别、地区、健康保险类型和家庭收入与2006年韩国5000万人口的比例保持一致，并进行了分层和随机抽样。我们排除了223,474名已经静脉注射或口服皮质类固醇的参与者，以减少处方皮质类固醇剂量对ONFH影响分析的偏倚。从剩余队列中，我们排除了241名在2002年已经诊断为ONFH的参与者，以获得ONFH的新进展(洗脱期)。因此，索引日期为患者被诊断为ONFH的日期。我们在最终分析中纳入了914146名参与者。数据集中的信息包括所有住院和门诊医疗索赔数据，包括处方药使用、诊断和治疗代码、初级和二级诊断代码以及健康检查数据。纳入国家健康保险制度-国家安全保障制度的所有个人都被跟踪到2019年，除非死亡或丧失参加国家健康保险的资格，例如移民。本研究的方案经我院机构审查委员会(4-2022-0304)和NHIS(NHIS-2022-2-232)批准。?CaseandControlSelectionWeusedtheInternationalStatisticalClassificationofDiseasesandRelatedHealthProblems,10thRevision(ICD-10)toidentifycasepatients.WedefinedONFH(N=3,500)usingthediagnosticcodeandtreatmentcodeICD-10codesforosteonecrosisofthefemur(SupplementaryTable1).Toavoidlengthbias,anestedcase-controlanalysiswasdesignedtoinvestigatetheeffectoftheprescribedcorticosteroiddoseonONFH.PatientswhohadONFHwerematched1:5toformacontrolgroupbasedonthevariablesofbirthyear,sex,andfollow-upduration.Identicalexclusioncriteriawereappliedtothecontrolgroup(N=12).Finally,casepatients(N=3,488)andcontrolpatients(N=17,440)werematchedbasedonsexandageattheindexdateandcompared(Figure1).我们使用国际疾病和相关健康问题统计分类第十次修订版(ICD-10)来确定病例患者。我们使用股骨骨坏死的诊断代码和治疗代码ICD-10来定义ONFH(N=3,500)(补充表1)。为避免长度偏倚，设计了巢式病例对照分析，以调查处方皮质类固醇剂量对ONFH的影响。根据出生年份、性别、随访时间等变量对ONFH患者进行1:5匹配，形成对照组。对照组(N=12)采用相同的排除标准。最后，根据索引日期的性别和年龄对病例患者(N=3,488)和对照患者(N=17,440)进行匹配和比较(图1)。??Fig.1Studyflow:selectionofparticipantsofcasesandmatchedcontrolsONFH,osteonecrosisoffemoralhead.??DefinitionofParametersFromthehealthexaminations,participantswereclassifiedinto4groupsbasedonthebodymassindex(BMI)accordingtotheWorldHealthOrganizationWesternPacificRegionguidelineandtheKoreanpracticalguidelineasfollows:low(<18.5),normal(18.5to23),overweight(23to25),andobese(>25).UnderlyingcomorbiditiesweredefinedbyatleastoneadmissionoroutpatienttreatmentupontheprimaryorfirstsecondarydiagnosisusingtheICD-10codeforthepastyear(hypertension:I10-I15,diabetesmellitus:E10-E14,dyslipidemia:E78).Thefrequencyofdrinkingalcoholwascategorizedas“non-alcoholuse”(drinkinglessthan1drinkperweek),“mildalcoholuse”(drinking1to2drinksperweek),or“heavyalcoholuse”(drinkingmorethan3to7drinksperweek).Smokingstatuswascategorizedasnonsmoker,pastsmoker,orcurrentsmoker.Afterovernightfasting,bloodsampleswereobtained,andserumhigh-densitylipoproteincholesterol,low-densitylipoprotein(LDL)cholesterol,andtriglyceride(TG)resultscouldbeobtained.根据健康检查结果，根据世界卫生组织西太平洋地区指南和韩国实践指南，将参与者的身体质量指数(BMI)分为4组:低(<18.5)、正常(18.5~23)、超重(23~25)和肥胖(>25)。潜在的合并症是在过去一年中使用ICD-10代码进行原发性或首次继发性诊断时至少进行一次住院或门诊治疗(高血压:I10-I15，糖尿病:E10-E14，血脂异常:E78)。饮酒频率分为“不饮酒”(每周饮酒少于1杯)、“轻度饮酒”(每周饮酒1至2杯)或“重度饮酒”(每周饮酒超过3至7杯)。吸烟状况分为不吸烟者、过去吸烟者和现在吸烟者。禁食过夜后采集血样，得到血清高密度脂蛋白胆固醇、低密度脂蛋白胆固醇和甘油三酯(TG)结果。?SystemicCorticosteroidPrescriptionWedefinedsystemiccorticosteroidprescriptionsasoralcorticosteroidprescriptionsandIVcorticosteroidprescription.Systemiccorticosteroidusewasdefinedastheprescriptionofmedicationcodesduringadmissionsandoutpatientvisitsfrom2003totheindexdate(Table1).Bothgroupshadidenticalobservationperiods.Allcorticosteroidformulationswereconvertedintoadailydosebasedonprednisoneequivalentdoses[24](1mgprednisolone=4mghydrocortisone=0.8mgmethylprednisolone=0.8mgtriamcinolone=0.16mgbetamethasone=1.2mgdeflazacort),andwecalculatedcumulativedosesofexposureforcorticosteroidswiththesumofthedosesforalltheprescribeddays,whichwasexpressedasthecumulativedefineddailydose(cDDD)accordingtotheWorldHealthOrganizationdefinition.我们将系统性糖皮质激素处方定义为口服糖皮质激素处方和静脉糖皮质激素处方。系统性糖皮质激素使用定义为从2003年到基线日期（见表1）的住院和门诊就诊期间的药物代码处方。两组观察期相同。将所有糖皮质激素制剂转换为基于泼尼松等效剂量（1毫克泼尼松龙等于4毫克氢化可的松等于0.8毫克甲基泼尼松龙等于0.8毫克曲安奈德等于0.16毫克倍他米松等于1.2毫克德夫氯喹）的每日剂量，并根据世界卫生组织的定义计算糖皮质激素的累积暴露剂量（累积定义每日剂量，cDDD），即将所有处方天数的剂量相加。?Table1PrescriptionDrugCodesandDosageofSystemicCorticosteroidBasedontheHealthInsuranceClaimsPaymentCodingSystem.Drug???????Code??????CorticosteroidandDosageOralcorticosteroids116401ATB???betamethasone0.5mg116501ATB???betamethasonesodiumphosphate0.5mg296900ATB???betamethasone0.25mg140801ATB???deflazacort6mg140802ATB???deflazacort30mg141901ATB???dexamethasone0.5mg141903ATB???dexamethasone0.75mg141904ATB???dexamethasone4mg170901ATB???hydrocortisone10mg170905ATB???hydrocortisone20mg170906ATB???hydrocortisone5mg193301ATB???methylprednisolone16mg193302ATB???methylprednisolone4mg193303ATB???methylprednisolone8mg193304ATB???methylprednisolone2mg193305ATB???methylprednisolone1mg217001ATB???prednisolone5mg243201ATB???triamcinolone1mg243202ATB???triamcinolone2mg243203ATB???triamcinolone4mgHigh-doseIVcorticosteroids?171201BIJ???????hydrocortisonesodiumsuccinate100mg171202BIJ?????hydrocortisonesodiumsuccinate250mg171203BIJ?????hydrocortisonesodiumsuccinate40mg193601BIJ?????methylprednisolonesodiumsuccinate125mg193602BIJ?????methylprednisolonesodiumsuccinate250mg193603BIJ?????methylprednisolonesodiumsuccinate40mgmethylprednisolonesodiumsuccinate500mg193604BIJ?????methylprednisolonesodiumsuccinate1000mg193605BIJ?????prednisolonesodiumsuccinate1000mg217301BIJ?????prednisolonesodiumsuccinate250mg217302BIJ???????SubgroupAnalysisATHAafteranONFHdiagnosiswasusedasasurrogateforfemoralheadcollapse.Therefore,wedividedallcasesinto2groups:agroupthatreceivedTHAandanothergroupthatdidnotreceiveTHAusingtreatmentcodes(N0711.N7020).Amongatotalof3,430patients,weexcluded70patientswhohadaTHAtreatmentcodepriortodiagnosis,1,175patientswhoreceivedTHAduringtheobservationperiod,and2,255patientswhodidnot.Wecomparedthe2groupsandanalyzedthemusingtheCoxproportionalhazardsmodel,whichfactorsaffectdiseaseprogressionleadingtoTHA(Figure2).诊断为ONFH后进行THA作为股骨头塌陷的替代。因此，我们使用治疗代码(N0711)将所有病例分为两组:接受THA治疗的组和未接受THA治疗的组。N7020)。在总共3430例患者中，我们排除了70例在诊断前有THA治疗代码的患者，1175例在观察期间接受THA治疗的患者和2255例未接受THA治疗的患者。我们对两组患者进行比较，并使用Cox比例风险模型进行分析，分析影响THA病情进展的因素(图2)。??Fig.2Studyflow:comparisonofpatientswhoreceivedanddidnotreceivetotalhiparthroplastyafterdiagnosisONFH,osteonecrosisoffemoralhead;THA,totalhiparthroplasty.??DataAnalysisChi-squaredtestsforcategoricalvariableswereperformedtocomparethebaselinecharacteristicsbetweencaseandcontrolpatients.AconditionallogisticregressionanalysiswasperformedtoevaluatetheassociationbetweensystemiccorticosteroiduseandtheriskofONFH.Forsubgroupanalysis,weusedCoxproportionalhazardmodelstoestimateadjustedhazardratios(aHRs)and95%confidenceintervals(CIs).APvalue<.05wasconsideredsignificant.AllstatisticalanalyseswereperformedusingSASEnterpriseGuideversion7.1(SASInstitute,Cary,NorthCarolina).?ResultsDemographicDataforONFHCasesandMatchedControlsTable2showsthebaselinecharacteristicsofcasesandmatchedcontrols.The2groupshadevendistributionsinthematchingvariables,includingsexandbirthyear.Subjectswerepredominantlymen(61.1%),andthemostcommonagegroupwasthoseborninthe1950s(age63to72asof2022).Comparedwiththecontrolgroup,ONFHcasepatientshadlowhouseholdincomeandhadmorediabetes,hypertension,dyslipidemia,andheavyalcoholuse(drinkingmorethan3to7drinksperweek).Inaddition,thecasepatientshadnormalserumLDLlevels(<100mg/dL),butconversely,thereweremanycasesofhypertriglyceridemiawithserumTGlevelsof200mg/dLormore.??Table2BaselineCharacteristicofOsteonecrosisofFemoralHeadCasesandMatchedControls.totalN=20,928(%)?????CasesN=3,488(%)ControlsN=17,440(%)?????PSex????????????????????????Men12,792(61.1)?2,132(61.1)???10,660(61.1)?Women???8,136(38.9)???1,356(38.9)???6,780(38.9)???????Birthyear??????????????????????????????<1930????852(4.1)142(4.1)710(4.1)1930to1939?3,024(14.4)???504(14.4)??????2,520(14.4)?????1940to1949?4,074(19.5)???679(19.5)??????3,395(19.5)?????1950to1959?5,112(24.4)???852(24.4)??????4,260(24.4)?????1960to1969?3,786(18.1)???631(18.1)??????3,155(18.1)?????1970to1979?2,184(10.4)???364(10.4)??????1,820(10.4)?????1980to1989?1,080(5.2)?????180(5.2)900(5.2)1990to1999?558(2.7)93(2.7)??465(2.7)2000to2009?210(1.0)35(1.0)??175(1.0)2010to2019?48(0.2)??8(0.2)????40(0.2)??Age(atindexdate)55.2±16.9????55.2±16.9???????55.2±16.9????Householdincome????????????????????????Low4,535(21.7)???896(25.7)??????3,639(20.9)???????<.001Middle???6,500(31.0)???1,090(31.2)???5,410(31.0)???????High???????9,893(26.5)???1,502(43.1)???8,391(48.1)???????Comorbidities????????????????????????Hypertension?6,631(31.7)???1,301(37.3)???5,330(30.6)?????<.001Diabetesmellitus???4,567(21.8)???908(26.0)???????3,659(21.0)???<.001Dyslipidemia??5,555(26.5)???1,120(32.1)???4,435(25.4)?????<.001Frequencyofdrinkingalcohol??????????????????????????????<1d/wk??10,044(48.0)?1,613(46.2)???8,431(48.3)???????<.0011to2d/wk?????3,987(19.1)???560(16.1)??????3,427(19.7)?????3to7d/wk?????2,794(13.3)???621(17.8)??????2,173(12.5)?????Missing??4,103(19.6)???694(19.9)??????3,409(19.5)???????Smokingstatus?????????????????????????????Non-smoker???9,455(45.2)???1,539(44.1)???7,916(45.4)?????.199Ex-smoker?????2,865(13.7)???462(13.3)??????2,403(13.8)?????Currentsmoker??????4,488(21.4)???790(22.6)???????3,698(21.2)???Missing??4,120(19.7)???697(20.0)??????3,423(19.6)???????BMI,kg/m2??24.0±3.3??????23.9±3.4??????24.0±3.3??.547<18·5?????608(2.9)113(3.2)495(2.8).09018·5to22·9???6,126(29.3)???1,029(29.5)???5,097(29.2)?????23·0to24·9???4,239(20.3)???651(18.7)??????3,588(20.6)?????≥25·0?6,113(29.2)???1,049(30.1)???5,064(29.0)???????Missing??3,842(25.1)???646(18.5)??????3,196(18.3)???????HDL,mg/dL??54.2±20.5????55.2±22.7????54.0±20.0.015<40?2,113(10.1)???339(9.7)1,774(10.2%).03340to59??8,856(42.3)???1,411(40.5)???7,445(42.7%)??≥60???????4,700(22.5)???830(23.8)????3870(22.2%)???????Missing??5,259(25.1)???908(26.0)??????4,351(24.9%)??LDL,mg/dL???114.8±77.2??110.6±56.9??115.7±80.6<.001<100??????5,627(26.9)???1,045(30.0)???4,582(26.3)???????<.001100to159?????8,493(40.6)???1,297(37.2)???7,196(41.3)?????≥160?????1,528(7.3)?????234(6.7)1,294(7.4)?????Missing??5,280(25.2)???912(26.1)??????4,368(25.0)???????TG,mg/dL?????144.5±108.7154.8±125.0142.5±105.1??????<.001<150??????10,395(49.7)?1,651(47.3)???8,744(50.1)???????<.001150to199?????2,447(11.7)???390(11.2)??????2,057(11.8)?????≥200?????2,829(13.5)???540(15.5)??????2,289(13.1)???????Missing??5,257(25.1)???907(26.0)??????4,350(24.9)???????BMI,bodymassindex;HDL,highdensitylipoprotein;LDL,lowdensitylipoprotein;TG,triglyceride.??Table3showsthecomparisonofprescribedsystemiccorticosteroidsincasepatientsandcontrolpatients.Bothgroupshadidenticalobservationperiods(meanobservationperiodwas8.26years).Systemiccorticosteroidwasprescribedatleastoncemoreinthecasegroupthaninthecontrolgroup(71.8versus62.9%,P<.001).EvenwhenoralcorticosteroidsandIVcorticosteroidswerecompared,respectively,thenumberofcasesprescribedcorticosteroidsatleastoncewashigherinthepatientswhohadONFH(oralcorticosteroids:71.2versus62.4%,P<.001/IVcorticosteroids:10.7versus4.9%,P<.001).??Table3DifferenceinCorticosteroidUseBetweenCasesandMatchedControlsFrom2003UntilDiagnosis.TotalN=20,928(%)?????CasesN=3,488(%)ControlsN=17,440(%)?????PCorticosteroid(oralorhigh-doseIV)???????????????????????????Neveruse???????7,453(35.6)???983(28.2)??????6,470(37.1)?????<.001Everuse?13,475(64.4)?2,505(71.8)?10,970(62.9)???????Oralcorticosteroid????????????????????????Neveruse???????7,562(36.1)???1,004(18.8)???6,558(37.6)?????<.001Everuse?13,366(63.9)?2,484(71.2)?10,882(62.4)???????Neveruse???????7,562(36.1)???1,004(18.8)???6,558(37.6)?????<.001<180cDDDs??12,667(60.5)?2,192(62.8)???10,475(60.1)?????180to364cDDDs?397(1.9)146(4.2)251(1.4)365to729cDDDs?179(0.9)84(2.4)??95(0.5)??≥730cDDDs123(0.6)62(1.8)??61(0.4)??High-doseIVcorticosteroid??????????????????????????Neveruse???????19,703(94.2)?3,116(89.3)???16,587(95.1)?????<.001Everuse?1,225(5.8)?????372(10.7)??????853(4.9)Neveruse???????19,703(94.2)?3,116(89.3)???16,587(95.1)?????<.001<180cDDDs??988(4.7)266(7.6)722(4.1)180to364cDDDs?120(0.6)46(1.3)??74(0.4)??365to729cDDDs?77(0.4)??40(1.2)??37(0.2)??≥730cDDDs40(0.2)??20(0.6)??20(0.1)??cDDDs,cumulativedefineddailydoses(prednisolone10mg);IV,intravenous.??CumulativeCorticosteroidDoseandONFHDevelopmentConditionallogisticregressionanalysiswasperformedtoanalyzetheeffectofthecumulativedoseofsystemiccorticosteroidsfor1yearbeforediagnosisontheincidenceofONFH(Table4).SystemiccorticosteroidusewassignificantlyassociatedwithanincreasedriskofONFHincidencecomparedtothatofnonusers.Specifically,higherdosesoforalcorticosteroidwereassociatedwithincreasedriskofONFH(oddsratiosof1.61(95%CI1.47to1.78,P<.001),4.39(95%CI3.47to5.56,P<.001),6.42(95%CI4.65to8.87,P<.001),and5.44(95%CI3.65to8.13,P<.001)forpatientswhohavecorticosteroiduseof<180,180to365,365to720,and>720cDDDs(valuesconvertedtodailyprednisolone10mg),respectively.TheriskofONFHincreasedrapidlywhenthecDDDwas180ormore,thatis,whencumulativeprednisoloneusewas1,800mgormore.Inaddition,atrendtowardriskincreasewithcumulativedosesofIVcorticosteroidusewasshownwithadjustedoddsratio(AOR)of1.63(95%CI1.39to1.90,P<.001),2.34(95%CI1.57to3.49,P<.001),3.77(95%CI2.33to6.10,P<.001),and2.76(95%CI1.41to5.39,P=.003)forpatientswhohavecorticosteroiduseof<180,180to365,365to720,and>720cDDDs(valuesconvertedtodailyprednisolone10mg),respectively.??Table4AdjustedRiskofOsteonecrosisofFemoralHeadOccurrence.ConditionalLogisticRegressionAnalysisforONFHDevelopmentCrudeOR??????95%CI???P?????AdjustedOR??95%CI???????PHouseholdincome????????????????????????????????????????Low1.381.26to1.52???<.001?????1.381.25to1.51???????<.001Middle???1.131.03to1.23???.0071.131.03to1.23???????.008High???????1.00ref.?????????1.00ref.??Comorbidities???????????????????????????????????????Hypertension?1.471.35to1.60???<.001?????1.21???????1.06to1.37???.004Diabetesmellitus???1.381.26to1.51???<.001???????1.070.95to1.20???.289Dyslipidemia??1.471.35to1.60???<.001?????1.18???????1.04to1.34???.012Frequencyofdrinkingalcohol?????????????????????????????????????????????<1d/wk??1.00ref.?????????1.00ref.??1to2d/wk?????0.870.78to0.97???.0150.900.80to1.01.0713to7d/wk?????1.541.38to1.73???<.001?????1.54???????1.36to1.74???<.001Missing??1.080.97to1.20???.1801.120.59to2.14???????.731Smokingstatus?????????????????????????????????????????????Non–smoker??1.00ref.?????????1.00ref.??Ex–smoker?????1.010.89to1.15???.8670.930.82to1.07.314Currentsmoker??????1.131.01to1.26???.0321.01???????0.90to1.14???.855Missing??1.060.95to1.18???.2791.030.55to1.94???????.923BMI,kg/m2?????????????????????????????????????????<18·5?????1.130.91to1.41???.2621.150.92to1.44???????.23218·5to22·9???1.00ref.?????????1.00ref.??23·0to24·9???0.900.81to0.99???.0490.900.80to1.01.062≥25·0?1.020.93to1.13???.6280.990.89to1.09???????.784Missing??1.000.90to1.13???.9450.800.55to1.17???????.254HDL,mg/dL?????????????????????????????????????????<40?1.010.88to1.15???.9250.950.83to1.09???????.44740to59??1.00ref.?????????1.00ref.??≥60???????1.131.03to1.25???.0091.100.99to1.21???????.069Missing??1.121.01to1.23???.0312.910.14to60.46???????.490LDL,mg/dL??????????????????????????????????????????<100??????1.00ref.?????????1.00ref.??100to159?????0.790.72to0.86???<.001?????0.86???????0.78to0.94???.001≥160?????0.790.67to0.92???.0030.84071to0.98???????.031Missing??0.930.83to1.03???.1520.840.28to2.56???????.762TG,mg/dL????????????????????????????????????????????<150??????1.00ref.?????????1.00ref.??150to199?????1.010.89to1.14???.9170.990.88to1.13.970≥200?????1.261.13to1.40???<.001?????1.171.04to1.32.009Missing??1.121.01to1.23???.0250.460.03to7.70???????.585Medication????????????????????????????????????????????Steroid(oralorhigh–doseIV)?????????????????????????????????????????????Neveruse???????1.00ref.?????????1.00ref.??Everuse?1.751.59to1.92???<.001?????1.771.60to1.94<.001POsteroid?????????????????????????????????????????????Neveruse???????1.00ref.?????????1.00ref.??Everuse?1.731.57to1.90???<.001?????1.681.53to1.85<.001Neveruse???????1.00ref.?????????1.00ref.??<180cDDDs??1.631.49to1.79???<.001?????1.61???????1.47to1.78???<.001180to364cDDDs?4.983.96to6.25???<.001???????4.393.47to5.56???<.001365to729cDDDs?7.515.50to10.27?<.001???????6.424.65to8.87???<.001≥730cDDDs8.966.16to13.02?<.001?????5.44???????3.65to8.13???<.001High–doseIVsteroid????????????????????????????????????????????Neveruse???????1.00ref.?????????1.00ref.??Everuse?2.422.12to2.76???<.001?????2.201.92to2.51<.001Neveruse???????1.00ref.?????????1·00?????ref.??<180cDDDs??2.041.76to2.38???<.001?????1.63???????1.39to1.90???<.001180to364cDDDs?3.402.35to4.92???<.001???????2.341.57to3.49???<.001365to729cDDDs?5.863.74to9.18???<.001???????3.772.33to6.10???<.001≥730cDDDs5.232.81to9.73???<.001?????2.76???????1.41to5.39???.003BMI,bodymassindex;cDDDs,cumulativedefineddailydoses(prednisolone10mg);CI,confidenceinterval;HDL,highdensitylipoprotein;IV,intravenous;LDL,lowdensitylipoprotein;ONFH,osteonecrosisoffemoralhead;OR,oddsratio;PO,peroral;TG,triglyceride.??RiskFactorsforTHAinPatientswhoHaveONFHFortheONFHcasepatients,asubgroupanalysiswasperformedon1,175patientsintheTHAgroupand2,255patientsinthenon-THAgroupduringtheobservationperiod.Thefollow-updurationoftheTHAgroup(1,175patients)was1.03±2.25years,andthefollow-updurationofthenon-THAgroup(2,255patients)was7.93±5.15years.Table5showsthebaselinecharacteristicsofTHAcasesandnon-THAcases.IntheTHAgroup,therewasagreaterproportionofmen(63.9versus60.0%,P=.028),higherBMI(24.1versus23.8,P=.027),heavyalcoholuseofthosewhoconsumedalcoholmorethan3to7daysaweek(22.9versus15.2%,P<.001),currentsmokers(26.4versus20.8%,P<.001),andhigherTG(162.3mg/dLversus150.0,P≤.001).However,whentheeffectondiseaseprogressionwasanalyzedusingCoxproportionalhazardsmodelforsurvivalanalysisafterONFHdiagnosis,heavyalcoholuseandlongercorticosteroidusesweretheriskfactorsaffectingdiseaseprogression,anddiabeteswasafactorthatslowedprogression.Moreover,otherlipidprofilesaswellascorticosteroidsusedandcumulativedosesofcorticosteroidsdidnotaffecttheincidenceofTHA(Table6).Inaddition,whenthesurvivalanalysiswasperformedafterthediagnosisofONFHbydividingthepatientgroupintomalesandfemales,heavyalcoholuseinvolvingthosewhoconsumedalcoholmorethan3to7daysaweekincreasedtheriskofdiseaseprogressionafterONFHdiagnosisinmalesonly(aHR:1.38,95%CI1.16to1.65,P<.001),andaBMIof25orhigherinfemalesonlyincreasedtheriskofincidenceofTHAafterONFHdiagnosis(aHR:1.52,95%CI1.18to1.96,P<.001).??Table5Comparisonofthe2GroupsAccordingtoWhetherorNotTotalHipArthroplastywasPerformedDuringtheFollow-UpPeriodAfterOsteonecrosisofFemoralHeadDiagnosis.TotalN=3,430(%)??????THAafterDiagnosisN=1,175(%)?????NoTHAafterDiagnosisN=2,255(%)???????PSex????????????????????????Men2,104(61.3)???751(63.9)??????10,660(61.1)???????.028Women???1,326(38.7)???424(36.1)??????6,780(38.9)???????Birthyear??????????????????????????????<1930????141(4.1)16(1.4)??125(5.5)<.0011930to1939?492(14.3)??????153(13.0)??????339(15.0)?????1940to1949?665(19.4)??????246(20.9)??????419(18.6)?????1950to1959?837(24.4)??????344(29.3)??????493(21.9)?????1960to1969?621(18.1)??????229(19.5)??????392(17.4)?????1970to1979?359(10.5)??????118(10.0)??????241(10.4)?????1980to1989?180(5.3)56(4.8)??124(5.2)1990to1999?92(2.7)??13(1.1)??79(3.5)??2000to2009?35(1.0)??0(0.0)????35(1.6)??2010to2019?8(0.2)????0(0.0)????8(0.3)????Age(atindexdate)55.1±17.0????56.41±13.8???????54.6±18.4????.007Householdincome????????????????????????Low880(25.7)??????283(24.1)??????597(26.5)???????.008Middle???1,077(31.4)???409(34.8)??????668(29.6)???????High???????1,473(42.9)???483(41.1)??????990(43.9)???????Comorbidities????????????????????????Hypertension?1,269(37.0)???445(37.9)??????824(36.5)?????.466Diabetesmellitus???888(25.9)??????285(24.3)???????603(26.7)??????.125Dyslipidemia??1,095(31.9)???383(32.6)??????712(31.6)?????.568Frequencyofdrinkingalcohol??????????????????????????????<1d/wk??1,581(46.1)???504(42.9)??????1,077(47.8)???????<.0011to2d/wk?????549(16.0)??????207(17.6)??????342(15.2)?????3to7d/wk?????613(17.9)??????269(22.9)??????344(15.2)?????Missing??687(20.0)??????195(16.6)??????492(21.8)???????Smokingstatus?????????????????????????????Nonsmoker????1,507(43.9)???491(41.8)??????1,016(45.1)?????.199Ex-smoker?????454(13.2)??????180(15.3)??????274(12.2)?????Currentsmoker??????780(22.7)??????310(26.4)???????470(20.8)??????Missing??689(20.1)??????194(16.5)??????495(21.9)???????BMI,kg/m2???23.9±3.4??????24.1±3.4??????23.8±3.3??.027<18·5?????112(3.3)37(3.2)??75(3.3)??<.00118·5-22·9??????1,014(29.3)???355(30.2)??????659(29.2)?????23·0-24·9??????640(18.7)??????218(18.5)??????422(18.7)?????≥25·0?1,024(29.8)???388(33.0)??????636(28.2)???????Missing??640(18.7)??????177(15.1)??????463(20.5)???????HDL,mg/dL??55.3±22.9????55.7±16.0????55.1±26.1.015<40?331(9.6)110(9.4)221(9.8).03340-59?????1,379(40.2)???494(42.0)??????885(39.2)???????≥60???????822(24.0%)??322(27.4%)??500(22.2%)???????Missing??898(26.2%)??248(21.2%)??649(28.8%)???????LDL,mg/dL???110.6±57.1??111.4±80.9??110.1±37.0.668<100??????1,023(29.8%)378(32.2%)??645(28.6%)???????<.001100-159?1,277(37.2%)464(39.5%)??813(36.1%)???????≥160?????229(6.7%)????81(6.9%)??????148(6.5%)???????Missing??901(26.3%)??252(21.4%)??649(28.8%)???????TG,mg/dL?????154.5±125.1162.3±141.3150.0±114.5??????.024<150??????1,626(47.4%)576(19.0%)??1,050(46.6%)??<.001150-199?378(11.0%)???140(11.9%)???238(10.5%)???????≥200?????530(15.5%)??211(18.0%)???319(14.2%)???????Missing??896(26.1)??????248(21.1)??????648(28.7)???????Systemiccorticosteroid(Oralorhigh-doseIV)???????????????????????????Neveruse???????2,090(60.9)???710(60.4)??????1,380(61.2)?????.687Everuse?1,340(39.1)???465(39.6)??????875(38.8)???????Oralcorticosteroid????????????????????????Neveruse???????2,119(61.8)???718(61.1)??????1,401(62.1)?????.584Everuse?1,311(38.2)???457(38.9)??????854(37.9)???????.690Neveruse???????2,119(61.8)???718(61.1)??????1,401(62.1)?????<180cDDDs??1,240(36.1)???434(36.9)??????806(35.7)?????180-364cDDDs????47(1.4)??15(1.3)??32(1.4)??365-729cDDDs????21(0.6)??8(0.7)????13(0.6)??≥730cDDDs3(0.1)????0(0.0)????3(0.1)????High-doseIVcorticosteroid??????????????????????????Neveruse???????3,322(96.9)???1,142(97.2)???2,180(96.7)?????.471Everuse?108(3.1)33(2.8)??75(3.3)??.858Neveruse???????3,322(96.9)???1,142(97.2)???2,180(96.7)?????<180cDDDs??84(2.4)??26(2.2)??58(2.5)??180-364cDDDs????9(0.3)????2(0.2)????7(0.3)????365-729cDDDs????10(0.3)??4(0.3)????6(0.3)????≥730cDDDs5(0.1)????1(0.1)????4(0.2)????BMI,bodymassindex;cDDDs,cumulativedefineddailydoses(prednisolone10mg);CI,confidenceinterval;HDL,highdensitylipoprotein;IV,intravenous;LDL,lowdensitylipoprotein;ONFH,osteonecrosisoffemoralhead;OR,oddsratio;TG,triglyceride;THA,totalhiparthroplasty.??Table6AdjustedRiskofTotalHipArthroplastyofOsteonecrosisofFemoralHeadPatients.ConditionalLogisticRegressionAnalysisforONFHDevelopmentCrudeOR??????95%CI???P?????AdjustedOR??95%CI???????PSex????????????????????????????????????????Men1.00ref.?????????1.00ref.??Women???0.890.79to1.01???.0601.030.88to1.20???????.731Age?1.011.00to1.01???<.001?????1.011.01to1.02???????<.001Householdincome????????????????????????????????????????Low1.000.86to1.16???.9891.020.88to1.19???????.786Middle???1.201.05to1.37???.0081.211.06to1.38???????.006High???????1.00ref.?????????1.00ref.??Comorbidities???????????????????????????????????????Hypertension?1.120.99to1.26???.0581.100.91to1.32.340Diabetesmellitus???0.950.83to1.09???.4460.80???????0.68to0.96???.013Dyslipidemia??1.090.97to1.23???.1570.980.81to1.19.842Frequencyofdrinkingalcohol?????????????????????????????????????????????<1d/wk??1.00ref.?????????1.00ref.??1to2d/wk?????1.231.05to1.45???.0111.231.02to1.47.0263to7d/wk?????1.501.30to1.74???<.001?????1.38???????1.16to1.65???<.001Missing??0.900.76to1.06???.2043.120.80to12.23???????.103Smokingstatus?????????????????????????????????????????????Nonsmoker????1.00ref.?????????1.00ref.??Ex-smoker?????1.311.11to1.56???.0020.251.03to1.52.027Currentsmoker??????1.261.09to1.45???.0021.17???????0.98to1.41???.085Missing??0.860.73to1.02???.0850.400.10to1.64???????.203BMI,kg/m2?????????????????????????????????????????<18·5?????0.940.67to1.31???.6980.890.63to1.25???????.51018·5to22·9???1.00ref.?????????1.00ref.??23·0to24·9???0.940.80to1.12???.4940.950.80to1.12.519≥25·0?1.090.95to1.26???.2321.110.96to1.29???????.176Missing??0.770.65to0.93???.0051.040.61to1.77???????.880HDL,mg/dL?????????????????????????????????????????<40?0.930.76to1.15???.5170.910.74to1.12???????.37640to59??1.00ref.?????????1.00ref.??≥60???????1.130.98to1.30???.0941.161.01to1.34???????.043Missing??0.770.66to0.90???<.001?????0.320.03to3.11.325LDL,mg/dL??????????????????????????????????????????<100??????1.00ref.?????????1.00ref.??100to159?????0.970.84to1.11???.6140.990.86to1.13.844≥160?????0.930.73to1.18???.5610.960.75to1.23???????.728Missing??0.740.63to0.87???<.001?????5.091.59to16.28??????.006TG,mg/dL????????????????????????????????????????????<150??????1.00ref.?????????1.00ref.??150to199?????1.070.89to1.29???.4471.050.87to1.27.583≥200?????1.191.02to1.39???.0301.110.94to1.31???????.231Missing??0.780.67to0.91???.0010.490.07to3.60???????.485Medication????????????????????????????????????????????Steroid(oralorhigh–doseIV)?????????????????????????????????????????????Neveruse???????1.00ref.?????????1.00ref.??Everuse?1.090.97to1.22???.1721.070.95to1.21???????.268POsteroid?????????????????????????????????????????????Neveruse???????1.00ref.?????????1.00ref.??Everuse?1.090.97to1.23???.1341.070.95to1.21???????.268Neveruse???????1.00ref.?????????1.00ref.??<180cDDDs??1.090.97to1.23???.1421.080.95to1.21.247180to364cDDDs?1.100.66to1.83???.7161.22???????0.73to2.06???.449365to729cDDDs?1.280.64to2.58???.4821.62???????0.79to3.36???.191≥730cDDDs<0.001???<0.001to999.944<0.001???????<0.001to999.943High–doseIVsteroid????????????????????????????????????????????Neveruse???????1.00ref.?????????1.00ref.??Everuse?0.910.64to1.29???.5920.900.63to1.28???????.552Neveruse???????1.00ref.?????????1.00ref.??<180cDDDs??0.950.64to1.39???.7760.870.59to1.30.504180to364cDDDs?0.600.15to2.41???.4740.71???????0.17to2.87???.626365to729cDDDs?1.200.45to3.20???.7191.35???????0.50to3.64???.554≥730cDDDs0.480.07to3.38???.4580.540.07to4.01.545Steroiduseduration?????????????????????????????????????????????Neveruse???????1.00ref.?????????1.00ref.??<3mobeforediagnosis?1.601.45-1.76<.001???????1.621.47-1.78<.0013-6mobeforediagnosis3.863.22-4.63<.001???????3.913.25-4.70<.0016-12mobeforediagnosis??????5.284.16-6.71<.001???????5.234.10-6.67<.001>12mobeforediagnosis7.575.98-9.59<.001???????7.405.82-9.42<.001?查看原图BMI,bodymassindex;cDDDs,cumulativedefineddailydoses(prednisolone10mg);CI,confidenceinterval;HDL,highdensitylipoprotein;IV,intravenous;LDL,lowdensitylipoprotein;ONFH,osteonecrosisoffemoralhead;OR,oddsratio;PO,peroral;TG,triglyceride;THA,totalhiparthroplasty.??DiscussionWeevaluated3,488ONFHcasesand17,440controlcaseswithmatchingvariablesincludingage,sex,andthetimeofthefollow-upata1:5ratiofromtheNHIS-NSC,includingthefollow-updatafrom2002to2019of1,137,861participantsinanationwidelongitudinalnestedcase-controlstudy.Wedemonstratedthatalowhouseholdincome,diabetes,hypertension,dyslipidemia,heavyalcoholicswhoconsumedalcoholmorethan3daysaweek,andserumTGlevelsof200mg/dLormorewereassociatedwiththedevelopmentofONFH.Inaddition,systemiccorticosteroidusewassignificantlyassociatedwithanincreasedriskofONFHincidencecomparedtothatofnonusers.Specifically,theriskofONFHincreasedrapidlywhenthecDDDswere180ormorecumulativedosesofcorticosteroiduse.WeanalyzedthefactorsaffectingtheprogressionofthediseasebycomparingpatientswhounderwentTHAwiththosewhodidnotundergoTHAduringthefollow-upperiodafterONFHdiagnosis.Men,whohadahigherBMI,heavyalcoholicswhoconsumedalcoholmorethan3daysaweek,currentsmokers,andserumTGlevelsof200mg/dLormorewereassociatedwithTHAafterONFHdiagnosis.However,afteradjustingforcompoundfactors,heavyalcoholicsweretheonlyfactoraffectingtheincidenceofTHAafterONFHdiagnosis,anditmaybeusedtocounselONFHpatients.Otherlipidprofiles,corticosteroidsused,andcumulativedosesofcorticosteroidsdidnotaffecttheincidenceofTHA.我们评估了3488例ONFH病例和17440例对照病例，匹配变量包括年龄、性别和随访时间，比例为1:5，来自NHIS-NSC，包括2002年至2019年全国纵向巢式病例对照研究中1137861名参与者的随访数据。研究表明，家庭收入低、糖尿病、高血压、血脂异常、每周饮酒超过3天的重度酗酒者以及血清TG水平≥200mg/dL与ONFH的发生有关。此外，与不使用皮质类固醇的患者相比，全身性使用皮质类固醇与ONFH发生率增加显著相关。特别是，当皮质类固醇的累积使用剂量达到180或更高时，ONFH的风险迅速增加。我们通过比较ONFH诊断后随访期间接受THA和未接受THA的患者来分析影响疾病进展的因素。BMI较高的男性、每周饮酒超过3天的重度酗酒者、当前吸烟者、血清TG水平≥200mg/dL的男性在ONFH诊断后与THA相关。然而，在调整复合因素后，重度酗酒者是ONFH诊断后唯一影响THA发生率的因素，可用于指导ONFH患者。其他脂质特征、使用的皮质类固醇和皮质类固醇的累积剂量对THA的发生率没有影响。TherelationshipbetweencorticosteroiduseandONFHdevelopmenthasbeenrevealedinseveralstudies[8,25,26],andtherearealsosomestudiesthathavesuggestedspecificdosesanddurationsofcorticosteroiduse,suchastheassociationwiththedevelopmentofdiseasewhenusingmorethan2gwithin3months[27]orthemeancorticosteroiduseofpatientswhohaveONFHdevelopmentof3,314mg[28].Inthecurrentlargecohortstudyofthegeneralpopulation,weanalyzedvariousfactorsthatcanaffectthedevelopmentofONFH.Asinpreviousstudies,thecaseofcorticosteroiduseatleastonceandadditionalcumulativecorticosteroiddosewereassociatedwiththeincreasedriskofONFHdevelopment,andwhencumulativeprednisoloneusewas1,800mg(180cDDDs)ormore,theriskofONFHdevelopmentincreasedrapidly.Surely,thedevelopmentofONFHrelatedtocorticosteroiduseisthoughttobecausedbyacombinationofnotonlycorticosteroidusebutalsoseveralfactorssuchascomorbidities,alcohol,smoking,andgeneticfactors.Sincecorticosteroidtreatmentisoftenessentialformanydiseases,itisnecessarytobeawareofthecumulativedoseatwhichtheriskofONFHincreasesrapidlywhencontinuouscorticosteroidtreatmentisrequired.?一些研究已经揭示了皮质类固醇的使用与ONFH发展之间的关系[8,25,26]，也有一些研究提出了皮质类固醇使用的特定剂量和持续时间，如在3个月内使用超过2g与疾病发展的关系[27]，或ONFH发展患者的平均皮质类固醇使用量为3,314mg[28]。在目前针对普通人群的大型队列研究中，我们分析了影响ONFH发展的各种因素。与之前的研究一样，至少使用一次皮质类固醇和额外的皮质类固醇累积剂量与ONFH发展的风险增加有关，当泼尼松龙累积使用1800毫克(180cDDDs)或更多时，ONFH发展的风险迅速增加。当然，与皮质类固醇使用相关的ONFH的发展被认为不仅是由皮质类固醇使用引起的，还包括合并症、酒精、吸烟和遗传因素等多种因素。由于皮质类固醇治疗通常对许多疾病至关重要，因此有必要了解在需要持续皮质类固醇治疗时ONFH风险迅速增加的累积剂量。Tothebestofourknowledge,nostudyhascomparedtheincidenceofONFHwithhigh-doseIVandoralcorticosteroidsatthesamecumulativedose.Inthepresentstudy,bothoralcorticosteroidandhigh-doseIVcorticosteroidsincreasedtheriskofdevelopingONFH,andwhencomparingtheoralandhigh-doseIVatthesamecumulativedose,oralcorticosteroidhadahigherrisk;totheAORsoforalcorticosteroidwere1.61(<180cDDDs),4.39(180to365cDDDs),6.42(365to720cDDDs),and5.44(>720cDDDs).TheAORsofhigh-doseIVcorticosteroidwere1.63(<180cDDDs),2.34(180to365cDDDs),3.77(365to720cDDDs),and2.76(>720cDDDs),respectively.Ingeneral,sincetheprednisolonepotencyoforalcorticosteroidsismuchlowerthanthatofhigh-doseIV,thedurationoforalcorticosteroidusewouldhavebeenlongerifthesamecumulativedosewasused.Montetal.reportedthatONFHoccurrencewasassociatedwithmeandailycorticosteroiddose,cumulativedose,andtreatmentduration[8].Ofcourse,ONFHoccurrenceswouldbeaffectedbygeneticsusceptibilityandexposuretovariousriskfactors,butinourstudy,weconfirmedthatmoreONFHoccurredinoralcorticosteroidswitharelativelysmalldailydoseatthesamecumulativedosecomparedtohigh-doseIVcorticosteroid,soitisassumedthatthelongertreatmentperiodofcorticosteroidisalsoassociatedwithdevelopingONFH.据我们所知，没有研究比较相同累积剂量的高剂量静脉注射和口服皮质类固醇对ONFH的发生率。在本研究中，口服皮质类固醇和高剂量静脉注射皮质类固醇都增加了发生ONFH的风险，并且在相同累积剂量下，口服皮质类固醇与高剂量静脉注射皮质类固醇相比，风险更高;口服皮质类固醇的AORs分别为1.61(<180cDDDs)、4.39(180~365cDDDs)、6.42(365~720cDDDs)和5.44(>720cDDDs)。高剂量静脉注射皮质类固醇的AORs分别为1.63(<180cDDDs)、2.34(180~365cDDDs)、3.77(365~720cDDDs)和2.76(>720cDDDs)。一般来说，由于口服皮质类固醇的强的松龙效力远低于高剂量静脉注射，如果使用相同的累积剂量，口服皮质类固醇的使用时间会更长。Mont等人报道ONFH的发生与皮质类固醇的平均每日剂量、累积剂量和治疗时间有关[8]。当然，ONFH的发生会受到遗传易感性和暴露于各种危险因素的影响，但在我们的研究中，我们证实在相同累积剂量下，相对于高剂量静脉注射皮质类固醇，日剂量相对较小的口服皮质类固醇更易发生ONFH，因此我们假设皮质类固醇治疗时间越长也与ONFH的发生有关。ItiswellknownthatahigherTGlevel,orLDLcholesterol,isanimportantriskfactorforischemicheartdiseaseandstrokebyinhibitingbloodflow,andONFHispresumedtoberelatedtoinhibitionofbloodflowtothefemoralheadinasimilarmechanism[29].Similartopreviousstudies[7,30],higherTGwasariskfactorforONFHinourstudy.Inaddition,itwasconfirmedthatLDLatlessthan100mg/dLhasaprotectiveeffectontheoccurrenceofONFH.众所周知，较高的TG或LDL胆固醇水平通过抑制血流是缺血性心脏病和中风的重要危险因素，而ONFH被认为与股骨头血流的抑制有类似的机制[29]。与以往的研究相似[7,30]，在我们的研究中，高TG是ONFH的危险因素。此外，还证实了低于100mg/dL的LDL对ONFH的发生有保护作用。SinceONFHoccurspredominantlyinyoungerpatientsandpreservationofthenativejointasmuchaspossibleistheprincipleoftreatment,thediseaseprogressionandprognosisareasimportantastheoccurrenceofthedisease.WeoftenseecasesofbilateralONFHdevelopmentwhereonesidehasfemoralcollapseandtheothersideremainsasymptomaticwithoutfemoralheadcollapse.Alternatively,somecaseswereasymptomaticwithoutfemoralheadcollapse,butreceivedTHA.ItisknownthattheprognosisafterONFHdevelopmentisinfluencedbyvariousfactors,anduntilnow,therehavebeenstudiesshowingthatradiologicfactorssuchasthesizeorlocationofnecrosisoracetabularanatomicalfactorshaveaneffect[1,19,31].Montetal.reportedthattheprevalenceoffemoralheadcollapsewas38%(106of282hips)inameta-analysisandvariedfrom17to73%dependingontheriskfactor[16].Inourstudy,34.3%ofpatientsunderwentTHA,anaverageof1.03±2.25yearsafterdiagnosis,similartothepreviousstudy.Althoughradiologicassessmentcouldnotbeperformed,casesthatmayhaveseengreaterphysicalloads,suchasmen,whohadahigherBMI,excessivedrinking,smoking,andahigherserumTGlevelover200mg/dL,wereassociatedwithdiseaseprogression.Inparticular,thefactorsthatincreasedtheriskofdiseaseprogressionweredifferentinmenandwomen.Wefoundthatcorticosteroidsprescribedornotandcumulativedosesofcorticosteroidswereunlikelytoaffectdiseaseprogression.由于ONFH主要发生在年轻患者中，治疗原则是尽可能保留原有关节，因此疾病的进展和预后与疾病的发生同样重要。我们经常看到双侧ONFH发展的病例，其中一侧有股骨头塌陷，另一侧无股骨头塌陷症状。另外，一些病例无股骨头塌陷症状，但接受了THA。众所周知，ONFH发生后的预后受多种因素影响，迄今已有研究表明，坏死的大小或位置等放射学因素或髋臼解剖因素对ONFH的预后有影响[1,19,31]。Mont等人在一项荟萃分析中报道，股骨头塌陷的患病率为38%(282髋中的106髋)，根据危险因素的不同，患病率从17%到73%不等[16]。在我们的研究中，34.3%的患者接受了THA，平均诊断后1.03±2.25年，与之前的研究相似。虽然无法进行放射学评估，但可能出现较大身体负荷的病例，如男性，BMI较高，过度饮酒，吸烟，血清TG水平高于200mg/dL，与疾病进展相关。特别是，增加疾病进展风险的因素在男性和女性中是不同的。我们发现，开具或不开具皮质类固醇以及皮质类固醇的累积剂量不太可能影响疾病进展。Ourstudyhasseveralpotentiallimitationsthatshouldbeaddressedinfurtherstudies.TheONFHwasassessedbyanoperationaldefinitionusingadiagnosiscode,notbyaradiologicevaluationsuchasanx-rayorMRI.However,theincidenceofONFHinthisstudywassimilartothatinapreviousAsiangeneralpopulationstudythatdefineditusinghipjointx-raysand/orMRI[32].BecauseanaccurateselectionofONFHpatientsmaynothavebeenmade,anestedcase-controlanalysisfromapopulation-basedcohortwasperformedtoincreasetheaccuracyoftheanalysis.Additionally,usingtheNHIS-NSCdatabase,prescriptionrecordsforcorticosteroidscouldbeaccessed.However,theactualintakeofcorticosteroidsinsubjectsmightbedifferentfromtheprescriptionrecords.Fortunately,severalstudieshaveshownagoodcorrelationbetweenprescriptionsandrealexposuretodrugs[33,34].Also,wecouldnotobtaininformationabouttraumaticONFH,whichmayhaveinfluencedtheresultsofthecurrentstudy.Inaddition,wecouldnotassessotherinterventionsthatcouldaffectdyslipidemia,microvascularbloodflow,andosteogenesis,suchastheuseoflipid-loweringmedication,aspirin,antiplatelets,andanticoagulantmedication.Thisshouldbeevaluatedinfuturestudies.Furthermore,ananalysisofmultiplecomorbiditiessuchasrheumaticdisease,sicklecelldisease,humanimmunodeficiencyvirus,organtransplantation,andsoonthatcouldaffecttheoccurrenceandprognosisofONFHwasnotperformed.Inparticular,sincetheprognosisofONFHisdifferentforeachdisease,subgroupanalysisaccordingtocomorbidityisabsolutelynecessaryforfuturestudies.Additionally,weanalyzedtheassociationbetweencumulativedoseofcorticosteroidanddiseaseprogressionafterONFHdiagnosisusingcDDD,andweconfirmedthatcumulativedoseofcorticosteroiddidnotaffectdiseaseprogression.However,sincethenumberofpatientswhohave180cDDDormorewassmall,additionalanalysiswithalargernumberofpatientsmaybeneededtoobtainmoreaccurateresults.我们的研究有几个潜在的局限性，应该在进一步的研究中加以解决。ONFH通过使用诊断代码的操作定义进行评估，而不是通过x射线或MRI等放射学评估。然而，本研究中ONFH的发生率与先前的亚洲普通人群研究相似，该研究使用髋关节X光片和/或MRI来定义ONFH[32]。由于可能没有对ONFH患者进行准确的选择，因此进行了基于人群的队列嵌套病例对照分析，以提高分析的准确性。此外，使用NHIS-NSC数据库，可以访问皮质类固醇的处方记录。然而，受试者的实际皮质类固醇摄入量可能与处方记录不同。幸运的是，有几项研究表明，处方与实际接触药物之间存在良好的相关性[33,34]。此外，我们无法获得有关创伤性ONFH的信息，这可能影响了当前研究的结果。此外，我们无法评估其他可能影响血脂异常、微血管血流和成骨的干预措施，如使用降脂药物、阿司匹林、抗血小板和抗凝药物。这应该在未来的研究中进行评估。此外，对风湿病、镰状细胞病、人类免疫缺陷病毒、器官移植等可能影响ONFH发生和预后的多重合并症未进行分析。特别是，由于每种疾病的预后不同，因此根据合并症进行亚组分析对于未来的研究是绝对必要的。此外，我们分析了使用cDDD诊断ONFH后皮质类固醇累积剂量与疾病进展之间的关系，我们证实皮质类固醇累积剂量不影响疾病进展。然而，由于cDDD≥180的患者数量较少，因此可能需要对更多患者进行额外分析，以获得更准确的结果。?ConclusionLowhouseholdincome,diabetes,hypertension,dyslipidemia,heavyalcoholicswhoconsumedalcoholmorethan3daysaweek,serumTGlevelsof200mg/dLormore,andoralorhigh-doseIVcorticosteroiduseareassociatedwithONFHdevelopment.Specifically,theriskofONFHincreasedrapidlywhencumulativeprednisoloneusewas1,800mgormore.However,oralorhigh-doseIVcorticosteroiduseandcumulativedosedidnotaffecttheprognosisofONFH.SincethedevelopmentandprognosisofONFHarecomplexandmultifactorialprocesses,furtherstudyisneeded.低收入家庭、糖尿病、高血压、血脂异常、每周饮酒超过3天的重度酗酒者、血清TG水平≥200mg/dL、口服或高剂量静脉注射皮质类固醇与ONFH的发生有关。具体来说，当累积使用强的松龙超过1800毫克时，ONFH的风险迅速增加。然而，口服或高剂量静脉注射皮质类固醇和累积剂量对ONFH的预后没有影响。由于ONFH的发展和预后是一个复杂的多因素过程，需要进一步研究。北京潞河医院科普号