脑卒中患者伴发股骨颈骨折的危险因素分析

Analysis of risk factors for stroke patients with femoral neck fracture

目的分析脑卒中患者股骨颈骨密度值及相关临床资料,探讨脑卒中患者伴发股骨颈骨折的危险因素。方法回顾性分析2017年1月-2019年8月湘雅二医院和湖南省人民医院收治的脑卒中患者163例。男性114例,女性49例。年龄46~89岁,平均66.4岁。按是否伴股骨颈骨折分为单纯脑卒中组(A组,n=126)和脑卒中伴股骨颈骨折组(B组,n=37)。分别记录两组患者年龄、性别、病程、BMI等资料。先单因素方法分析A组和B组在上述指标的差异,再分析股骨颈骨密度值(双能X线法测量)特征,最后采用Logistic回归分析筛选脑卒中患者并发股骨颈骨折的危险因素,建立回归模型并绘制ROC曲线。结果单因素分析显示,年龄、病程、吸烟、下肢运动分期、股四头肌痉挛、卒中后跌倒次数、BMI及认知评分差异可能与脑卒中患者发生股骨颈骨折有关(P<0.05)。股骨颈骨密度值:B组健侧低于A组健侧(P<0.001);A组男性和女性的患侧均低于健侧(P<0.001);A组女性双侧及B组女性健侧均低于男性(P<0.05)。Logistic回归分析显示,年龄(OR=1.155,P<0.001)、性别(OR=2.895,P=0.046)、BMI(OR=1.447,P<0.001)、下肢运动分期(OR=0.082,P=0.010;OR=0.656,P=0.482)及跌倒次数(OR=3.017,P<0.001)是脑卒中患者并发股骨颈骨折的危险因素。模型预测的灵敏度为64.9%,特异度为92.9%,准确率为86.5%。HosmerLemeshow检验显示χ2=6.100,P=0.636。ROC曲线下面积为0.909(P<0.001),提示拟合优度非常好。结论跌倒次数增加是脑卒中患者发生股骨颈骨折的首要危险因素,高龄、女性、高BMI及下肢运动分期低下是次要危险因素。

Objective To analyze the bone mineral density of femoral neck and related clinical data of stroke patients,and to explore the risk factors of femoral neck fracture in stroke patients.Methods A retrospective analysis of 163 stroke patients was carried out.They were divided into simple stroke group(group A,n=126)and stroke with femoral neck fracture group(group B,n=37).Data such as age,gender,duration of illness and body mass index(BMI)were recorded.First,the single factor method was used to analyze the difference of above indicators between groups A and B,and then the bone density data of femoral neck was analyzed.Finally,Logistic regression analysis was used to screen the risk factors of femoral neck fracture in stroke patients,and the regression model was established and the ROC curve was drawn.Results Univariate analysis showed that age,duration of disease,smoking,lower limb motor staging,spasm of quadriceps femoris muscle,post-stroke fall,BMI,and cognitive scores might be associated with femoral neck fractures in stroke patients(P<0.05).The bone density data of femoral neck in the healthy side of group B was lower than that in the same side of group A(P<0.001),the affected side of males and females were lower than in the healthy side of group A,respectively(P<0.001),and the bilateral data of females in group A and the healthy side of female in group B was lower than that of the males(P<0.05).Logistic regression analysis showed age(OR=1.155,P<0.001),gender(OR=2.895,P=0.046),BMI(OR=1.447,P<0.001),and lower limb motor staging(OR=0.082,P=0.010;OR=0.656,P=0.482)and the number of falls(OR=3.017,P<0.001)were risk factors for femoral neck fracture in stroke patients.The model predicted a sensitivity of 64.9%,a specificity of 92.9%,and an accuracy of 86.5%.Theχ2 value of the Hosmer-Lemeshow test was6.100(P=0.636).The area under curve of the ROC was 0.909(P<0.001),indicating that the goodness of fit was very satisfactory.Conclusion Increased number of falls is the primary risk factor for femoral neck fractures in stroke patients.Older age,female,high BMI and lower limb motor staging are secondary risk factors.