Risk Factors Of Vertebral Refracture After Vertebroplasty And Osteoporosis

Osteoporosis is one metabolic skeletal system disease that seriously harms human health.It is characterized by decreased bone mass,destruction of bone microstructure and decreased bone strength.With the aging of population,the incidence of osteoporosis is also increased.As the serious consequence of osteoporosis,osteoporotic fractures are brittle fractures that can be caused by mild violence,and the most of them are complete fractures.Among them,osteoporotic vertebral compression fracture(OVCF)is the most common.Because conservative treatment can cause serious complications such as deep vein thrombosis in lower extremities,lung infection and acne,minimally invasive surgeries are widely accepted by doctors and patients,and mainly include percutaneous vertebroplasty(PVP)and percutaneous kyphoplasty(PKP)which are both called as vertebroplasty.However,many patients undergoing vertebroplasty suffer from vertebral refractures which significantly decrease the quality of life and increase mortality.This project intends to retrospectively collect the medical history and auxiliary examination data of patients undergoing vertebroplasty in Department of Orthopedics,the First Affiliated Hospital of Chongqing Medical University,and then systematically analyze the effects of BMI,number of vertebral fractures,type of surgery,surgical timing,BMD and other factors on the occurrence of refracture after vertebroplasty,which is of great importance for the prevention of refractures.In addition,previous studies have found that decreased BMD and osteoporosis have significantly negative impact on the refracture after vertebroplasty,and therefore the risk factors of osteoporosis should be studied in order to prevent the recurrence of vertebral fractures.Mendelian Randomization(MR)study is mainly used to explore the causal association between exposure factors and outcomes,which is a powerful method to find the risk factors of diseases.Genetic variations are randomly assigned before birth and fixed at conception,thus serving as a valid proxy for exposure factors.Mendelian randomization study can effectively avoid the bias of confounding factors and reverse causal association by using the genetic variations.This Mendelian randomization(MR)study aims to explore the potential risk factors of osteoporosis(including serum adiponectin,hypertension,serum magnesium and depression),which is of great significance for the prevention and treatment of osteoporosis and refracture after vertebroplasty.This paper includes the following five parts:PART I RISK FACTORS OF VERTEBRAL REFRACTURE AFTER VERTEBROPLASTY1 Research objectivesThe risk factors of vertebral refracture after vertebroplasty are systematically studied,including BMI,number of vertebral fractures,type of surgery,surgical timing and BMD,which can provide an important guidance for the prevention of vertebral refracture.2 Research methodsA retrospective study was intended to assess risk factors of vertebral refracture after vertebroplasty due to OVCF: The medical history and auxiliary examination data of patients undergoing vertebroplasty conducted between April 2014 and March 2016 were collected in Department of Orthopedics,the First Affiliated Hospital of Chongqing Medical University.The occurrence and time of vertebral refractures within 4-6years were retrospectively collected.Kaplan-meier survival analysis,univariate and multivariate logistic regression analyses were used to systematically investigate the effects of BMI,number of vertebral fractures,type of surgery,surgical timing(time from injury to surgery)and BMD on vertebral refracture after vertebroplasty.In addition,we also investigated the impact of these risk factors on death after vertebroplasty.3 ResultsA total of 198 patients who underwent vertebroplasty in our hospital were included in the study.The results showed that the risk of refracture was significantly higher in the late surgical timing group(>21d)than in the early surgical timing group(≤21d,P=0.003),and there was no statistical difference in the risk of death between two groups(P=0.148).The low BMD group(T≤-3.5)had a significantly higher incidence of vertebral refracture than high BMD group(T>-3.5,P<0.001),but there was no significant difference in the risk of postoperative death between two groups(P=0.074).The risk of refracture was significantly higher in the multivertebral fracture group(≥2)than in the single-vertebral fracture group(P=0.005),while there was no significant difference in the risk of death between two groups(P=0.678).The risk of death was significantly higher in low BMI group(≤25kg/m2)than in high BMI group(>25kg/m2,P=0.046),while BMI had no significant effect on the risk of refracture(P=0.29).There was no significant difference in the risk of refracture(P=0.134)or death(P=0.083)between the types of surgery(PVP and PKP).Univariate and multivariate logistic regression analyses showed that higher height and pulmonary diseases were risk factors of refracture,while cerebral infarction was the risk factor of death.4 Summary(1)The risk factors of vertebral refracture after vertebroplasty included low BMD,late surgical timing,multiple vertebral fractures,high height and pulmonary diseases;(2)Risk factors of death after vertebroplasty included low BMI and cerebral infarction;(3)Other factors had no significant effect on the incidence of refracture or death after vertebroplasty: type of surgery,age,weight,cement leakage,smoking,alcohol consumption,hypertension,diabetes or coronary atherosclerotic heart disease.PART II CAUSAL RELATIONSHIP BETWEEN SERUM ADIPONECTIN AND OSTEOPOROSIS1 Research objectivesThis Mendelian randomization study is used to determine whether there is a causal relationship between serum adiponectin and osteoporosis,and find new methods and ideas for the prevention and treatment of osteoporosis from the perspective of serum adiponectin.2 Research methodsThe population samples of serum adiponectin and BMD in this Mendelian randomization study included 67,739 subjects for serum adiponectin and 53,236 subjects for femur neck BMD(FN-BMD),forearm BMD(FA-BMD)and lumbar spine BMD(LS-BMD).The causality between serum adiponectin and BMD(FA-BMD,FN-BMD and LS-BMD)was studied by inverse variance weighting(IVW)analysis,weighted median and MR-Egger regression methods,and various sensitivity analyses were conducted.P<0.05 indicated that the difference was statistically significant.In multiple tests,Bonferroni correction P<0.05/4=0.0125 was considered statistically different.3 ResultsFifteen SNPs were used as instrumental variables of serum adiponectin.According to IVW analysis,high serum adiponectin was the risk factor of low FN-BMD(beta:-0.015,95% CI:-0.023 to-0.006,SE:0.004,P=0.001<Bonferroni correction P),and may reduce FA-BMD(beta:-0.027,95%CI:-0.050 to-0.004,SE: 0.012,P=0.023).Furthermore,the causal association between serum adiponectin and FN-BMD was further confirmed by weighted median method(beta:-0.017,95%CI:-0.029 to-0.005,SE: 0.006,P=0.005<Bonferroni correction P).There was no significant heterogeneity in the association analysis between serum adiponectin and BMD.MR-Egger regression results showed that genetic pleiotropy did not bias the results.No pleiotropic SNPs were found in MR-PRESSO analysis,so no SNPs were removed for causal association assessment.4 Summary(1)High serum adiponectin was the risk factor for FN-BMD reduction;(2)High serum adiponectin may be a risk factor for FA-BMD reduction;(3)High serum adiponectin did not decrease LS-BMD;(4)Multiple sensitivity analyses confirmed the causal association between FN-BMD and serum adiponectin.These results suggested that serum adiponectin reduction may benefit to the prevention and treatment of osteoporosis.PART III CAUSAL RELATIONSHIP BETWEEN HYPERTENSION AND OSTEOPOROSIS1 Research objectivesThis study aims to explore whether there is a causal relationship between blood pressure and osteoporosis,and find new ideas for prevention and treatment of osteoporosis from the perspective of hypertension.2 Research methodsPopulation samples associated with blood pressure and BMD in this Mendelian randomization study included 753,560 subjects for systolic blood pressure(SBP),diastolic blood pressure(DBP)and pulse pressure(PP),and 53,236 subjects for FN-BMD,FA-BMD and LS-BMD.This part explored the causal association between BP(SBP,DBP and PP)and BMD(FA-BMD,FN-BMD and LS-BMD).Mendelian randomised(MR)associations were assessed by IVW analysis,weighted median and MR-Egger regression methods.Various sensitivity analyses were performed.P<0.05 suggested that the difference was statistically significant.In multiple tests,Bonferroni correction P<0.05/4=0.0125 was considered statistically different.3 ResultsThe instrumental variables of SBP were 21 SNPs,while those of DBP and PP were 37 SNPs,respectively.The causal associations between blood pressure(SBP,DBP and PP)and BMD(FA-BMD,FN-BMD and LS-BMD)were studied.IVW analysis found that high PP was associated with increased FA-BMD(beta: 0.038,95% CI: 0.013 to 0.063,SE: 0.013,P=0.003<Bonferroni correction P),which was also supported by weighted median(beta: 0.034,95%CI: 0.000 to 0.067,SE: 0.017,P=0.046)and MR-Egger regression methods(beta: 0.117,95%CI: 0.026 to 0.208,SE:0.046,P=0.011).However,IVW,weighted median and MR-Egger regression methods showed that PP did not cause the change of FN-BMD or LS-BMD.According to IVW analysis,there was no causal association between SBP and BMD: FA-BMD(beta:-0.015,95%CI:-0.041 to 0.011,SE: 0.013,P=0.258),FN-BMD(beta: 0.010,95% CI:-0.001 to 0.021,SE:0.006,P=0.064)or LS-BMD(beta: 0.001,95% CI:-0.014 to 0.016,SE:0.008,P=0.921).DBP also did not affect FA-BMD(beta: 0.002,95%CI:-0.026 to 0.030,SE: 0.014,P=0.876),FN-BMD(beta:-0.012,95% CI:-0.025 to 0.001,SE: 0.007,P=0.077)or LS-BMD(beta:-0.004,95%CI:-0.019 to 0.012,SE: 0.008,P=0.666).The weighted median method and MR-Egger regression method also confirmed the above results.Only the MR association analysis between PP and LS-BMD showed significant heterogeneity(P=0.045),while other MR association analyses had no significant heterogeneity.The results of MR-Egger regression showed that genetic pleiotropy only caused bias for MR analysis between DBP and FN-BMD.No pleiotropic SNPs were found in MR-PRESSO analysis,so no SNPs were removed for causal association assessment.4 Summary(1)High PP could increase FA-BMD.(2)There was no causal association between PP and FN-BMD,LS-BMD.(3)There was no causal association between SBP and BMD(FA-BMD,FN-BMD,LS-BMD).(4)There was no causal association between DBP and BMD(FA-BMD,FN-BMD,LS-BMD).(5)Multiple sensitivity analyses confirmed the causal associations between blood pressure and BMD.These results suggested that further studies on the causal association between PP and FA-BMD may provide new approaches for the prevention and treatment of osteoporosis.PART IV CAUSAL RELATIONSHIP BETWEEN SERUM MAGNESIUM AND OSTEOPOROSIS1 Research objectivesThis study aims to investigate whether there is a causal relationship between serum magnesium and osteoporosis,and discover new methods for the prevention and treatment of osteoporosis from the perspective of serum magnesium.2 Research methodsPopulation samples associated with serum magnesium and BMD in this Mendelian Randomization study included: 15,366 people for serum magnesium,53,236 people for FN-BMD,FA-BMD and LS-BMD,and426,824 people for HE-BMD.To determine the causal association between serum magnesium concentration and BMD,IVW analysis,weighted median method and MR-Egger regression method were performed using instrumental variables SNPs.P<0.05 indicated that the difference was statistically significant.Multiple sensitivity analyses were performed to verify the results.In multiple tests,Bonferroni correction P<0.05/4=0.0125 was considered statistically different.3 ResultsSix SNPs were used as instrumental variables for serum magnesium concentration.This part studied whether serum magnesium caused the changes of FA-BMD,FN-BMD,LS-BMD or HE-BMD.Serum magnesium did not significantly alter FA-BMD(beta:-1.968,95%CI:-4.397 to 0.461,SE: 1.239,P=0.112),FN-BMD(beta:-1.171,95%CI:-2.567 to 0.225,SE:0.712,P=0.100)or HE-BMD(beta:-0.225,95%CI:-0.365 to 0.914,SE:0.582,P=0.698).However,IVW analysis showed that high serum magnesium was negatively associated with LS-BMD(beta:-1.982,95% CI:-3.328 to-0.635,SE: 0.687,P=0.004<Bonferroni correction P),and this important finding was confirmed by weighted median analysis(beta:-2.295,95%CI:-3.739 to-0.851,SE: 0.737,P=0.002<Bonferroni correction P).In addition,weighted median analysis showed a significant causal association between high serum magnesium and low HE-BMD(beta:-0.929,95%CI:-1.376 to-0.481,SE: 0.228,P<0.001<Bonferroni correction P).There was significant heterogeneity between serum magnesium concentration and FN-BMD,HE-BMD(P<0.05).MR-Egger regression results showed that genetic pleiotropy did not bias the results(P>0.05).Among the 6 SNPs instrumental variables,MR-PRESSO method found one pleiotropic SNP(rs7965584)for FN-BMD and four pleiotropic SNPs(rs4072037,rs7965584,rs3925584 and rs11144134)for HE-BMD.After excluding these pleiotropic SNPs,there was a significantly negative causal association between high serum magnesium and FN-BMD(beta:-1.570,95% CI:-2.770 to-0.371,SE: 0.612,P=0.010<Bonferroni correction P),while other causal associations remained unchanged.4 Summary(1)High serum magnesium was a risk factor for low LS-BMD.(2)High serum magnesium may be associated with low FN-BMD and HE-BMD.(3)There was no causal association between high serum magnesium and FA-BMD.(4)Multiple sensitivity analyses confirmed the causal association between high serum magnesium and LS-BMD.These results suggest that serum magnesium reduction may benefit to the prevention and treatment of osteoporosis.PART V CAUSAL RELATIONSHIP BETWEEN DEPRESSION AND OSTEOPOROSIS1 Research objectivesThis study is intended to explore whether there is a causal relationship between depression and osteoporosis,and find new approaches for the prevention and treatment of osteoporosis from the perspective of depression.2 Research methodsThis Mendelian randomized study included population samples: 807,553 subjects associated with depression(246,363 depression patients and 561,190 normal people),53,236 subjects associated with FN-BMD,FA-BMD and LS-BMD,as well as 426,824 subjects associated with HE-BMD.In order to determine whether depression was causally associated with FN-BMD,FA-BMD,LS-BMD or HE-BMD,we performed IVW analysis,weighted median and MR-Egger regression methods for the MR association,and multiple sensitivity analyses were conducted.P<0.05 suggested that the difference was statistically significant.In multiple tests,Bonferroni correction P<0.05/4=0.0125 was considered statistically different.3 ResultsSeventy-nine SNPs were used as instrumental variables for depression.This part evaluated the causal association between depression and FA-BMD,FN-BMD,LS-BMD.IVW analysis showed that depression did not cause changes in BMD: FA-BMD(beta: 0.091,95%CI:-0.088 to0.269,SE: 0.091,P=0.320),FN-BMD(beta: 0.066,95% CI:-0.016 to0.148,SE: 0.042,P=0.113),LS-BMD(beta: 0.074,95% CI:-0.029 to0.177,SE: 0.052,P=0.159)or HE-BMD(beta: 0.009,95%CI:-0.043 to0.061,SE: 0.027,P=0.727).These negative associations were also confirmed by weighted median method: FA-BMD(beta: 0.144,95% CI:-0.092 to 0.380,SE: 0.120,P=0.231),FN-BMD(beta: 0.070,95%CI:-0.061 to 0.180,SE: 0.056,P=0.215),LS-BMD(beta: 0.068,95%CI:-0.702 to 0.197,SE: 0.066,P=0.302)or HE-BMD(beta:-0.001,95% CI:-0.034 to 0.033,SE: 0.017,P=0.976).There was no significant heterogeneity in MR analysis between depression and FN-BMD(P=0.075),but there was significant heterogeneity between depression and FA-BMD,LS-BMD,HE-BMD(P<0.05).MR-Egger regression method showed that genetic pleiotropy might affect the MR association between depression and FA-BMD(P=0.014),but did not bias other results(P > 0.05).After the elimination of pleiotropic SNPs identified by MR-PRESSO analysis,there was still no causal association between depression and BMD.4 Summary(1)There was no causal association between depression and FA-BMD.(2)There was no causal association between depression and FN-BMD.(3)There was no causal association between depression and LS-BMD.(4)There was no causal association between depression and HE-BMD.(5)Multiple sensitivity analyses confirmed that there was no causal association between depression and BMD(FN-BMD,FA-BMD,LS-BMD and HE-BMD).These results suggest that depression was not a risk factor of osteoporosis.CONCLUSIONSAccording to the results of the retrospective study,early surgery,effective treatment of osteoporosis and pulmonary diseases can help reduce the risk of vertebral refracture after vertebroplasty.Mendelian randomization studies find that high serum adiponectin and high serum magnesium are risk factors of osteoporosis,and high PP can lead to an increase in FA-BMD.These findings are beneficial to develop new methods and approaches to prevent and treat osteoporosis and vertebral refracture after vertebroplasty from the perspectives of serum adiponectin,serum magnesium and blood pressure.