Association of lumbar disc degeneration with osteoporotic fractures; the Rotterdam study and...

$: Association of lumbar disc degeneration with osteoporotic fractures; the Rotterdam study and meta-analysis from systematic review$
1

2

3

4Q1

5

6Q3789

10

1112131415161718192021222324252627

52

53

54

55

56

57

58

59

60

Bone xxx (2013) xxx–xxx

BON-10117; No. of pages: 6; 4C:

Contents lists available at SciVerse ScienceDirect

Bone

j ourna l homepage: www.e lsev ie r .com/ locate /bone

Original Full Length Article

Association of lumbar disc degeneration with osteoporotic fractures;the Rotterdam study and meta-analysis from systematic review

OO

FM.C. Castaño-Betancourt a,b,c, L. Oei a,b,c, F. Rivadeneira a,b,c, E.I.T. de Schepper d, A. Hofman c, S. Bierma-Zeinstra d,H.A.P. Pols a, A.G. Uitterlinden a,b,c, J.B.J. Van Meurs a,b,⁎a Department of Internal Medicine, Erasmus Medical Center, 3000 DR Rotterdam, The Netherlandsb The Netherlands Genomics Initiative-sponsored Netherlands Consortium for Healthy Aging (NGI-NCHA), 2300 RC Leiden, The Netherlandsc Department of Epidemiology, Erasmus Medical Center, 3000 DR Rotterdam, The Netherlandsd Department of General Practice, Erasmus Medical Center, GK 1136 Rotterdam, The Netherlands

Abbreviations: LDD, lumbar disc degeneration; OPHnarrowing; VFx, vertebral fractures; BMD, BoneMineral Dteoarthritis; BMI, Body Mass Index.⁎ Corresponding author at: Genetic Laboratory Departm

Ee579b, Erasmus MC, MC PO Box 1738, 3000 DR RotterdaE-mail address: [email protected] (J.B.J. Van M

8756-3282/$ – see front matter © 2013 Elsevier Inc. All rihttp://dx.doi.org/10.1016/j.bone.2013.08.004

Please cite this article as: Castaño-Betancourtand meta-analysis from systematic review, B

Ra b s t r a c t
a r t i c l e i n f o
28
Article history: 29
30

31

32

33

34

35

36

37

38

39

40
CTED PReceived 12 June 2013
Revised 1 August 2013Accepted 5 August 2013Available online xxxx

Edited by: Stuart Ralston

Keywords:Lumbar disc degenerationOsteoporotic fracturesBone Mineral DensityDisc space narrowingVertebral fractures

Objective: To investigate the relation between lumbar disc degeneration (LDD) and all type of osteoporotic (OP)fractures including vertebral.Methods: This study is part of the Rotterdam study, a large prospective population-based cohort study amongmen and women aged 55 years and over. In 2819 participants spine radiographs were scored for LDD(osteophytes and disc space narrowing (DSN)) from L1 till S1, using the Lane atlas. Osteoporotic (OP) fracturedata were collected and verified by specialists during 12.8 years. We considered two types of vertebral fractures(VFx): Clinical VFx (symptomatic fractures recorded by medical practitioners) and Radiographic VFx (using theMcCloskey–Kanis method). Meta-analysis of published studies reporting an association of LDD features and VFxwas performed. Differences in BoneMineral Density (BMD) between participantswith andwithout LDD featureswere analyzed using ANOVA. Risk of OP-fractures was analyzed using Cox regression.Results: In a total of 2385 participants, during 12.8 years follow-up, 558 suffered an OP-fracture. Subjects withLDD had an increased OP fracture risk compared to subjects without LDD (HR: 1.29, CI: 1.04–1.60). LDD-caseshave between 0.3 and 0.72 standard deviations more BMD than non-cases in all analyzed regions including

41

42

43

44

45

46

47

48
RREtotal body BMD and skull BMD (P b 0.001). Only males with LDD had increased risk for OP-fractures compared
tomaleswithout LDD (adjusted-HR: 1.80, 95%CI: 1.20–2.70, P = 0.005). The riskwas alsohigher for VFx inmales(HR: 1.64, CI: 1.03–2.60, P: 0.04). The association LDD–OP-fractures in females was lower and not significant(adjusted-HR: 1.08, 95%CI: 0.82–1.41). Meta-analyses showed that the risk of VFx in subjects with LDD hasbeen studied only in women and there is not enough evidence to confidently analyze the relationship betweenLDD-features (DSN or/and OPH) and VFx due to low power and heterogeneity in phenotype definition in the col-lected studies.Conclusions:Male subjectswith LDD have a higher osteoporotic fracture risk, in spite of systemically higher BMD.

49
© 2013 Elsevier Inc. All rights reserved.
5051
O
61

62

63

64

65

66

67

68

UNCIntroduction

Lumbar disc degeneration (LDD) and osteoporosis are two age-related skeletal diseases which are very prevalent in elderly andknown to be related to pain, increased morbidity and disability in thispopulation [1,2]. In Europe, the mean prevalence of vertebral fractures(VFx) in women between 60 and 64 years is 17% and this increases upto 35% when they are aged 75 years or more [1]. Both, osteoporotic

69

70

71

72

73

74

75

76

, osteophytes; DSN, disc spaceensity; OP, osteoporotic; OA, os-

ent of Internal Medicine, Roomm, The Netherlands.eurs).

ghts reserved.

MC, et al, Association of lumbone (2013), http://dx.doi.org

(OP) fractures and LDD occur also in men however, it has been morestudied in women [3,4].

The relationship between LDD and bone health is unclear. As it hasbeen previously shown, the presence of LDD is associated with higherspine Bone Mineral Density (BMD) [5–7]. In addition, LDD has beenfound associated with higher BMD of the femoral neck, which suggestsa systemic increased BMD in subjects affected by LDD [5,6,8]. In this re-spect LDD behaves very similar to knee or hip osteoarthritis (OA),where also an increased systemic BMD has been found [9,10].

In theory, the higher BMD found in subjects with LDD should corre-sponds to lower fracture risk compared to subjects without LDD. How-ever, the few studies examining the relationship between LDD andvertebral fractures (in women) found conflicting results [7,11–14]. Inpart this might be explained by the different radiological definitionsused for LDD (based on the presence of osteophytes (OPH) and/or discspace narrowing (DSN)) and vertebral fractures (scored by different

ar disc degenerationwith osteoporotic fractures; the Rotterdam study/10.1016/j.bone.2013.08.004

http://dx.doi.org/10.1016/j.bone.2013.08.004

mailto:[email protected]


http://www.sciencedirect.com/science/journal/87563282


T

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

2 M.C. Castaño-Betancourt et al. / Bone xxx (2013) xxx–xxx

UNCO

RREC

methods). Additionally, there are no studies examining the relationshipbetween LDD and all types of OP fractureswhichwould indicatewheth-er the increased BMD found in LDD cases corresponds to a decreasedfracture risk.

Therefore, we investigated the relation between LDD and all type ofosteoporotic fractures including vertebral, in a large prospective cohortthat includes men and women. In addition, we performed a systematicreview of previously published studies.

Materials and methods

The Rotterdam study

This study is part of the Rotterdam study (RS), a large prospectivepopulation-based cohort study among men and women 55 years ofage and older. The study design and rationale are described elsewherein detail [15]. The objective of the study is to investigate the determi-nants, incidence and progression of chronic disabling diseases in the el-derly. The baseline measurements were conducted between 1990 and1993. In total, 7983 participants were examined. The current studywas performed in 2385 study participants for whom data on incidentvertebral fractures, BMD and LDD was available. The medical ethicscommittee of Erasmus University Medical School approved the studyand written informed consent was obtained from each participant.

Data collection for potential risk factors

Home interviews on medical history were performed by trained in-terviewers. Smoking habit was categorized binary as current or formerversus never. The lower limb disability index used was composed ofthe mean score from six different questions regarding activities ofdaily living, using a modified version from the Stanford Health Assess-ment Questionnaire [16].

At baseline measurement, medical information and physical exami-nation including height and weight were obtained. Body Mass Index(BMI) was calculated by dividing weight by height squared (kg/m2).

Radiographic assessment of LDD

Each vertebral level from L1 to S1was reviewed for the presence andseverity of osteophytes (OPH) and vertebral narrowing (disc spacenarrowing (DSN)), using the Lane atlas [3,4]. In this atlas the categoriesare as follows: grade 0 = none; grade 1 = mild; grade 2 = moderate;and grade 3 = severe. DSN was defined as present when there was agrade 1 narrowing at two or more vertebral levels. Because of thesmall proportion of subjects without osteophytes, we used a highercut-off value for this feature; OPH was positive when there wereosteophytes of at least grade 2 at two or more vertebral levels. WhenDSN and OPH were both positive and present at 2 or more levels, theparticipant was assigned as “LDD case”. The definition suggested forLDD was previously found as the best related to clinical symptoms in-cluding lumbar pain [17]. A severity score for each participant was cal-culated adding the individual scores of DSN and OPH (1–3) of allintervertebral levels.

BMD measurements

DXABMD (g/cm2) of the right proximal femur and lumbar spinewasmeasured at baseline using a Lunar DPX-L densitometer (Lunar Radia-tion Corp., Madison, WI, USA). Total body scans were performed at thethird follow-up visit (mean follow-up 6.5 years) using a ProdigyTMfan-beam densitometer (GE Lunar Corporation Madison, WI) and ana-lyzed with EncoreTM software. The software employs an algorithmthat divides body measurements into areas corresponding to totalbody, head, trunk, arms and legs. Other methodological details havebeen described previously [18].

Please cite this article as: Castaño-BetancourtMC, et al, Association of lumband meta-analysis from systematic review, Bone (2013), http://dx.doi.org

ED P

RO

OF

Assessment of osteoporotic fracture

Follow-up started either on January 1, 1991 or at the time of inclu-sion into the study. For this analysis, follow-up ended either at January1, 2007 or, when earlier, at the participant's death or loss to follow-up.For ~80% of the study population, medical events were reportedthrough computerized general practitioner diagnosis registers. For theremaining 20%, research physicians collected data from the generalpractitioners' medical records of the study participants. All collectedfractures were verified by reviewing discharge reports and lettersfrom medical specialists. Fracture events were coded independentlyby two research physicians according to the International Classificationof Diseases, 10th revision (ICD-10). Finally, an expert in osteoporosisreviewed all coded events for final classification. Fractures coded as in-cident vertebral fractureswere considered clinical fractures if theywereidentified on radiographs when subjects with symptoms (principallypain) visited themedical practitioner. All fractures thatwere considerednot osteoporotic (fractures caused by cancer and all hand, foot, skull,and face fractures)were excluded. The period of follow-upwas calculat-ed as the time from enrollment in the study to the first fracture, death,or the end of the planned follow-up period, whichever occurred first.The participants were followed for the occurrence of fracture for ap-proximately 12.8 years (±3.1 SD yr).

Assessment of prevalent and incident radiographic vertebral fracture

Radiographic vertebral fracture: both at baseline, between 1990 and1993, and at the second follow-up visit, between 1997 and 1999, atrained research technician obtained lateral radiographs of thethoracolumbar spine of subjects whowere able to come to the researchcenter. The follow-up radiographs were available for 2819 individuals,who survived an average of 6.3 years after their baseline center visitand who were still able to come to our research center. All follow-upradiographs were evaluated morphometrically in Sheffield by theMcCloskey–Kanis method, as described previously [19]. If a vertebralfracture was detected, the baseline radiograph was evaluated as well.If the fracture was already present at baseline, it was considered a prev-alent fracture. All vertebral fractures were confirmed by visual interpre-tation by an expert in the field to rule out artifacts and other etiologies,such as pathological fractures [20]. Participants with missing data onone or more risk factors were excluded (n = 434).

Literature study

Relevant articles were identified by a systematic search using thedatabase of PubMed with the words [“spine osteoarthritis” or “spineOA” or “disc degeneration”] and [fracture] as keywords in the title orabstract. The following inclusion criteria applied for this review are:a) listed in PubMed, b) publication in the English language, c) study inhumans, d) the article represents original data, e) subjects with andwithout disc degeneration features are compared in the study inrelation to vertebral and/or osteoporotic fractures and f) the full-text ar-ticle was available. Methodological quality assessment is found in theSupplementary material.

Statistical analysis

We compared the baseline characteristics of the study populationand the FN- and LS-BMD between the LDD cases and controls usinganalysis of variance (ANOVA). Categorical variables were analyzedusing chi squared test. Cox's proportional hazards regression was usedto assess association between LDD andOP fractures (or only clinical ver-tebral fractures). The analyses were adjusted by gender (or stratified bygender), age, BMI, lower limb disability and FN-BMD as continuous var-iables. Departure from additive effect of the risk factorswas tested usinginteraction terms in themodel. All these analysesweremade using SPSS



193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

218Q4

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

Table 1t1:1

t1:2 Baseline characteristic for subjects according to LDD features in two or more levels.

t1:3 Variables Controls* n = 2023(82)

LDD** n = 362(18)

P value

t1:4 Female 1161 (57) 207 (57) 0.9t1:5 Age (years) 63.7 (5.7) 65.9 (6.2) b0.001t1:6 BMI (kg/m2) 26.2 (3.7) 26.9 (3.8) 0.04t1:7 FN-BMD (g/cm2) 0.88 (0.13) 0.91 (0.14) b0.001t1:8 Total-BMD (g/cm2) 1.10 (0.12) 1.12 (0.13) 0.001t1:9 Head-BMD (g/cm2) 1.93 (0.28) 2.01 (0.29) b0.001t1:10 Smoking (current/former) 1367 (68) 233 (64) 0.18t1:11 Lower limb disability 0.16 (0.38) 0.21 (0.33) b0.001t1:12 Falling (yes) 241 (12) 52 (14.5) 0.29t1:13 Prevalent radiological VFx* 88 (6.6) 22 (7.4) 0.83

t1:14 Abbreviations: LDD: Lumbar disc degeneration, **defined as mild disc space narrowingt1:15 (DSN) and moderate/severe osteophytes (OPH) in two vertebral levels. *Controls:t1:16 Participants with less than 2 levels affected by DSN and OPH per level. Values presentedt1:17 are mean and standard deviations (SD) for each continuous variable and numbers andt1:18 percentage (%) for categorical variables. BMI (Body Mass Index), smoking, LLD (lowert1:19 limb disability), falling and prevalent radiographic-vertebral fracture (VFx) comparisonst1:20 were adjusted for age and gender. All Bone Mineral Density (BMD) analyses weret1:21 adjusted for age, gender, BMI and LLD.

3M.C. Castaño-Betancourt et al. / Bone xxx (2013) xxx–xxx

V. 15.0. Meta-analyzed results and forest plots included in the literaturestudy were obtained using the Comprehensive Meta-analysis SoftwareVersion 2, Biostat, Englewood NJ (2005). Power calculations weredone with PS version 2.1.31.

T238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

ORREC

Results

Study population

Characteristics of the cohort comprising 2385 participants with datafor the two major outcomes: LDD and vertebral fracture are shown inTable 1. At baseline, 362 participants had LDD (moderate OPH andmild-DSN) in two or more intervertebral levels. Subjects with LDDwere older and heavier than controls. Also, LDD subjects had 0.72 and0.32 SD higher LS- and FN-BMD at baseline compared to controls(Fig. 1, P b 0.001 for both FN and LS-BMD differences). Additionally,Fig. 1 shows that total body-BMD and skull-BMD (measured at a latertime point) were also significantly increased in subjects with lumbardisc degeneration. All BMD analyses were adjusted for age, gender,BMI, and lower limb disability. The number of prevalent radiographic-vertebral fractures was not different in the LDD group compared tothe group without LDD after adjustment for age and gender (Table 1,P = 0.83). Themean LDD severity score was higher inmales than in fe-males (mean = 6.6 (SD = 4.3) for males and 5.9 (SD = 4.4) for fe-males (P b 0.001, adjusted for age and BMI)). However, there was nostatistically significant association between LDD-severity score and alltype of OP-fractures (P = 0.13).

UNC

Fig. 1. Bone Mineral Density (BMD; Z-scores) differences between subjects with lumbar disc dneck (FN), and total body and skull. Estimates were adjusted by age, gender, height and BMI. **ticipants at a second follow-up.


ED P

RO

OF

Osteoporotic fracture risk

During 12.8 (SD = 3.12) years of follow-up, 558 participants suf-fered an osteoporotic (OP) fracture. Subjects with LDD had an increasedrisk of OP fractures compared to subjects without LDD (HR: 1.29, CI:1.04–1.60). The risk slightly decreased after adjustment for age, gender,BMI, lower limb disability and FN-BMD (HR: 1.24 (0.99–1.55)). Wefound a significant interaction between gender and LDD on fracturerisk suggesting differences inOP fracture risk between genders (P for in-teraction term: 0.03). Therefore, we stratified the analysis according togender and observed that only males with LDD had an increased OPfracture risk (Table 2, adjusted HR: 1.80 (1.20–2.70), P = 0.005 formales and HR: 1.08, CI: 0.82–1.41, P = 0.59 for females).

Clinical vertebral fractureAfter the follow-up time, 21% of participants having fractures

(n = 116) had a clinically defined vertebral fracture. Participants withLDD had an increased hazard of having a clinical vertebral fracture dur-ing the follow-up (Table 2, adjusted-HR: 1.64, CI: 1.03–2.60, P = 0.04).As it was for overall OP-fractures, the hazard for a clinical vertebral frac-ture was higher only for males with LDD (HR: 2.34, CI: 1.09–5.04 inmales and 1.39, CI: 0.78–2.50 for females).

Radiographic vertebral fractureDuring 6.3 years of follow-up, 106 participants had an incident

radiographic-vertebral fracture. After adjustment for age, gender, BMI,FN-BMD and prevalent radiographic vertebral fracture, subjects withLDDhad 2.14 increased odds of having a radiographic vertebral fracture.However, this was not statistically significant and the broad confidenceinterval revealed low power in the analysis (CI: 0.82–5.58, P = 0.12).Hence, we reviewed the existent literature on the relationship betweenLDD and (vertebral) fractures.

Results of the literature study: separate LDD features and risk forradiographic-vertebral fractures

From 36 studies selected on bases of the search terms in abstract/title, a total of 28 studies were excluded because the abstract clearlyshowed that the study did not analyze the studied relationship LDD/spine OA-fracture. Based on the criterion 4 studies were excluded: nocomparison was made between subjects with and without disc degen-eration in relation to vertebral and/or osteoporotic fractures (criteriae). A total of five studies (four from the literature search + current re-sults from our study) analyzing the relation LDD and radiographic ver-tebral fractures were included in this review. There were no studiesanalyzing the relation of LDDwith other types of fractures in accordancewith the selection criteria previously explained. Further details of theselection procedure of studies can be found in the Supplementary

egeneration (LDD) and without LDD in four different regions: lumbar spine (LS), femoralP value ≤ 0.001. Total body-BMD and skull-BMD were measured in a subset of 1649 par-



T

F

260

261

262Q5

263

264

265Q6

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

Table 2t2:1

t2:2 Risk of vertebral and osteoporotic fracture in participant with lumbar disc degeneration (LDD).

t2:3 n.LDD/n Clinical vertebral fractures (HR & 95% CI) Osteoporotic fractures (HR & 95% CI)

t2:4 Unadjusted P Adjusted risk P Unadjusted P Adjusted risk P

t2:5 All (362/2385) 1.53 (0.98–2.40) 0.06 1.64 (1.03–2.60) 0.04 1.29 (1.04–1.60) 0.02 1.24 (0.99–1.55) 0.06t2:6 Males 1.41 (0.68–2.90) 0.36 2.34 (1.09–5.04) 0.03 1.88 (1.27–2.79) 0.002 1.80 (1.20–2.70) 0.005t2:7 Females 1.22 (0.71–2.10) 0.47 1.39 (0.78–2.50) 0.26 1.11 (0.85–1.43) 0.45 1.08 (0.82–1.41) 0.59

t2:8 Risk for osteoporotic and clinical vertebral fractures in participantswith Lumbar disc degeneration (LDD) defined as categorical variable according to the presence of at leastmild disc spacet2:9 narrowing (DSN ≥ 1) and moderate/severe osteophytosis (OPHQ2 ≥) per intervertebral level in at least two intervertebral levels. Lumbar disc degeneration was evaluated only at lumbart2:10 spine (L1–L5). Risk of vertebral and osteoporotic fracture are hazard ratio (HR) and 95% confidence interval (CI) unadjusted or adjusted for baseline characteristics (gender, age, BMI,t2:11 lower limb disability, femoral neck Bone Mineral Density (FN-BMD)). Number of clinical vertebral fractures = 116; number of osteoporotic fractures = 558.


EC

material. From these five selected studies, two were done in the samepopulation therefore, only the most recent “longitudinal prospective”(Sornay et al.) was included in themeta-analysis. These studies fulfilledthe inclusion criteria requirements and methodological quality assess-ment, including adjustment for age, gender, BMI and BMD in the analy-sis. Only one study (Roux et al.) did not perform BMD adjustment in theanalysis of vertebral fracture risk.

All selected studies were done in postmenopausal women, one ofthem in women with osteoporosis [11]. In all studies, radiographic-LDD features were evaluated from the first till fifth lumbar segment(L1–L5). LDD was defined as the presence of osteophytes (OPH), ordisc space narrowing (DSN) in at least one intervertebral level. A de-tailed description of the studies and definition of LDD and vertebral frac-ture assessment is presented in Supplementary Table 1.

Table 3 shows the results of the studies reviewed. Prevalence of atleast minimal/mild osteophytes in the studied populations differs be-tween 56 and 90% (Table 3). In the study of Roux et al. there was alsoa protective effect of OPH for vertebral fractures, however in thatstudy, the association was not adjusted for BMD; what seems to modifythe relationship of osteophytes–vertebral fractures (Table 2). Post-hocpower calculation demonstrated that to have 80% power to detectOR N =1.2 having an incidence of radiographic vertebral fractures ofaround 5%, a sample size of around 4200 participants would be needed.Consequently, confidence intervals are wide and associations of sepa-rate features of combined LDD definition did not reveal conclusiveevidence (Fig. 2, OR: 0.70, CI: 0.39–1.25, P: 0.23 and OR: 1.05, CI:

UNCO

RR

Fig. 2.Meta-analysis of lumbar disc degeneration features and their relation with vertebral fracDSN ≥ 1 = mild disc space narrowing, OPH ≥ 1 = minimal or mild osteophytes. Studies wer


0.75–1.46, P: 0.79 for the presence of at least mild osteophytes anddisc space narrowing, respectively).

ED P

RO

ODiscussion

This study shows that individuals affected with lumbar disc degen-eration (LDD), in spite of having a systemically higher BMD, are notprotected of osteoporotic fractures. Contrarily, male subjects with LDDhad an increased risk for osteoporotic fractures, including clinical verte-bral fractures. Results of themeta-analysis of the literature showed thatLDD has been studied only in women for its association with radio-graphic vertebral fractures. Additionally, this is the first study analyzingthe association of LDD with all type of fractures in males and females.The results of themeta-analysis were not conclusive regarding the rela-tion of separate LDD features (osteophytes and disc space narrowing)and risk of radiographic vertebral fractures in females.

We found a higher systemic BMD in participants with LDD. Partici-pants with LDD had a higher Bone Mineral Density not only in lumbarspine (where BMD measurements are known to be influenced by thepresence of osteophytes), but also in femoral neck, total body andskull. This is in line with previous findings of higher BMD in LDD pa-tients not only in lumbar spine but also in other body regions [6,7].Mea-surement of skull-BMD has been shown to be less subjective to changeduring aging and influence of environmental and mechanical factors(strain and weight bearing) [21]. Higher skull-BMD in subjects with

tures in women. Values are odds ratios (OR) and 95% confidence intervals. Abbreviations:e included when they had similar adjustments.



T

OF

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

Table 3t3:1

t3:2 Results of the reviewed studies for association of LDD and risk of vertebral fractures.

t3:3 Reference Definition of VFx n. of VFx OR (confidence interval) unadjusted risk OR (confidence interval) (BMD adjusted) P adj.

t3:4 Arden et al. [13] Quantitative McClosckey 41 Not shown LS-OPH: 0.46 (0.21–0.99) b0.05t3:5 Thoracic-OPH: 3.57 (1.55–8.24)t3:6 Sornay Rendu [14] Semi-quantitative Genant 48 Lumbar spine Lumbar spine Not shownt3:7 OPH ≥ 1: 0.8 (0.4–1.6) 1.0 (0.5–2.1)t3:8 DSN ≥ 1: 3.4 (1.5–7.8) 3.5 (1.5–8.3)t3:9 OA grade 2: 1.6 (0.9–2.3) 1.7 (0.9–3.2)t3:10 Sornay Rendu [12] Semi-quantitative Genant 42 LS-DSN: Not shown LS-DSN: 3.27 (1.19–8.98) Not shownt3:11 LS and thoracic spine LS and Thoracic spinet3:12 DSN ≥ 1: 6.88 (1.64–28.9) DSN N =1: 6.59 (1.36–31.94)t3:13 OPH ≥ 1: 1.39 (0.18–10.7) OPH ≧ 1: 0.94 (0.10–8.97)t3:14 OA grade 2: 1.57 (0.81–3.01) OA grade 2: 0.92 (0.42–1.99)t3:15 **Roux C et al. [11] Semi-quantitative Genant 215 DSN ≥ 1: 0.83 (0.55–1.26) DSN ≥ 1: 0.67 (0.43–1.04) 0.071t3:16 OPH ≥ 1: 0.45 (0.21–0.99) OPH ≥ 1: 0.38 (0.17–0.86) 0.02t3:17 LS-DSN ≥ 2: 0.86 (0.56–1.33) LS-DSN ≥ 2: 0.74 (0.47–1.16) 0.191t3:18 DSN ≥ 1: 1.26 (0.72–2.20) DSN ≥ 1: 1.27 (0.71–2.26) 0.42t3:19 This study Quantitative McClosckey 108 OPH ≥ 1: 1.10 (0.43–2.84) OPH ≥ 1: 1.27 (0.48–3.34) 0.63t3:20 OA grade 2: 1.00 (0.57–1.74) OA grade 2: 1.05 (0.59–1.87) 0.86t3:21 LDD: 1.14 (0.61–2.15) LDD: 1.14 (0.57–2.29) 0.70

5M.C. Castaño-Betancourt et al. / Bone xxx (2013) xxx–xxx

UNCO

RREC

LDD suggests that a systemically higher BMD might be present beforeLDD.

In spite of higher BMD in participants with LDD, we found a higherosteoporotic fracture risk. It is possible that the increased BMD in sub-jects with LDD is not enough to compensate for other detrimental ef-fects of disc degeneration on trunk stability and flexibility that mightresult in an increased fracture risk. Loading on the spine is determinedby a person's height, weight, muscle forces, and activity, but can alsobe affected by intervertebral disc degeneration [22–24]. Loss of discheight and its properties produce high tensile strains in the endplateand they have been shown as causal factors for “failure of the vertebra”[25,26]. Additionally, disc degeneration can affect other structures(vertebra itself, muscles and ligaments) producing modification in thedistribution of compressive and tensional forces through the columnthat in normal conditions are evenly distributed. Ligaments of the ante-rior region have changes as a consequence of LDD, causing its remodel-ing and thickening [27,28]. Consequently ligaments lose elasticity andthe trunk's flexibility decreases; this becomes evident during agingwhere the range of spine movement is severely affected. Individualswith LDDhavemore stiffness in trunk and lower legs that could increasethe reaction time during falling and other demanding occupational ac-tivities which are major situations where fractures occur in elderly[29,30].

We found a higher osteoporotic fracture risk in males. Severity ofLDD was higher in males, principally because of higher severity ofdisc space narrowing. Additionally, there is some evidence for an asso-ciation between disc space narrowing and lower back pain especiallyin men proportionally increasing with a higher number of affectedintervertebral disc spaces [17]. However, severity itself did not explainthe increased OP fracture risk in males. Neither was the risk explainedby other factors such as lower limb disability, which was found to behigher in males with LDD and other common risk factors includingage and falling risk. In older males (N = 65 years), clinical vertebralfractures are caused by no known trauma or by low-energy trauma.It is known that most fractures occur in men with normal BMD andclinical vertebral fractures are particularly common in the oldest men[31]. Clinical vertebral fractures have been also related to importantcomorbidities, negative effects on quality of life and increasedmortality[32–34].

The systematic review showed some important aspects. Previousstudies analyzing the association between LDD and fractures havebeen done only in females, the majority had small sample size andthey defined LDD using only separate features. There were also impor-tant methodological differences between these studies in type (cross-sectional versus longitudinal) and fracture assessment. In spite of trying


ED P

RO

to meta-analyze the results of studies included in this review, using ho-mogeneous definitions, power was still insufficient to draw significantconclusion on the relation of separate LDD features and radiological ver-tebral fractures.

The review made evident that there is a need for consensus in thedefinition of radiological LDD. In our opinion, more stringent radiologi-cal definitions are needed; some studies only consider osteophytes todefine LDD. Osteophytes are a common feature in older populationsand its prevalence depends of how stringent the definition is, reaching90% for the presence of minimal/mild osteophytes. The presence ofdisc space narrowing and osteophytes (at least moderate) in the sameintervertebral level should be considered when LDD is defined becauseit has been shown to bemore clinically relevant; this combination of ra-diological definitionwas found to best correlatewith clinical symptoms:lumbar pain and stiffness [17,29].

There are strengths and limitations in this study. This study is uniquein examining the relation of LDD, osteoporotic and vertebral fractures ina large prospective cohort that includes males and females. In addition,the composed definition of several radiographic features used in thisstudy is an advantage because it is more stringent and clinically rele-vant. We also examined separate LDD features in order to compare re-sults with earlier published studies. However, we concluded that evenafter the meta-analysis the number of radiographic vertebral fractureswas insufficient to get conclusive evidence in the relation of radiograph-ic vertebral fractures with separate LDD features.

Conclusions

To conclude, we consider that subjects with LDD in spite of havinghigher systemic BMD are at higher risk of osteoporotic fractures, espe-cially males for whom LDD seems more severe. The exact mechanismsto explain this association merits further investigation, consideringthat both, LDD and clinical vertebral fractures are common, associatedwith comorbidities and decreasing quality of life.

Authors' roles

Study design: MCB, JvM, and FR. Study conduct: MCB. Data collec-tion: EdS, SBZ, FR, LO, and AH. Data analysis: LO andMCB. Data interpre-tation: MCB, LO, JvM and FR. Drafting manuscript: MCB and JvM.Revisingmanuscript content: AGU, JvM, SBZ, AH, EdS and FR. Approvingfinal version ofmanuscript: all authors.MCB and JvM take responsibilityfor the integrity of the data analysis.



393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

411

412

413

414

415416417418419420421422423424425426427428429430431432433434435436437438439440441442

443444445446447448449

512


Disclosure

All authors of this manuscript state that they have no conflicts of in-terest. In addition, all authors had access to all raw data, statistical anal-yses, and material used in this manuscript.

450451452453454455456457458459460461462463464465466467468469470

Acknowledgment

This study is funded by the European Commission Seventh frame-work program TREAT-OA (grant 200800), The Netherlands Society forScientific Research (NWO), Research Institute for Diseases in the Elderly(014-93-015; RIDE2), the Netherlands Genomics Initiative (NGI)/Netherlands Organization for Scientific Research (NWO) project no.050-060-810, and the Netherlands Consortium for Healthy Ageing(NCHA). The Rotterdam study is funded by the Erasmus Medical Centerand Erasmus University, Rotterdam, the Netherlands Organization forthe Health Research and Development (ZonMw), the Research Institutefor Diseases in the Elderly (RIDE), theMinistry of Education, Culture andScience, theMinistry for Health,Welfare and Sports, the European Com-mission (DG XII), and the Municipality of Rotterdam. The authors de-clare no any conflict of interest.

471472473474475476477478

Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.bone.2013.08.004.

T

479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510
N
CO

RREC

References

[1] Cummings SR,Melton LJ. Epidemiologyand outcomes of osteoporotic fractures. LancetMay 18 2002;359(9319):1761–7.

[2] Engel CC, von Korff M, Katon WJ. Back pain in primary care: predictors of highhealth-care costs. Pain May-Jun 1996;65(2–3):197–204.

[3] Lawrence JS. Disc degeneration. Its frequency and relationship to symptoms. AnnRheum Dis Mar 1969;28(2):121–38.

[4] Miller JA, Schmatz C, Schultz AB. Lumbar disc degeneration: correlationwith age, sex,and spine level in 600 autopsy specimens. Spine (Phila Pa 1976) Feb 1988;13(2):173–8.

[5] Jones G, Nguyen T, Sambrook PN, Kelly PJ, Eisman JA. A longitudinal study of the ef-fect of spinal degenerative disease on bone density in the elderly. J Rheumatol May1995;22(5):932–6.

[6] Livshits G, Ermakov S, Popham M, Macgregor AJ, Sambrook PN, Spector TD, et al.Evidence that bone mineral density plays a role in degenerative disc disease: theUK Twin Spine study. Ann Rheum Dis Dec 2010;69(12):2102–6.

[7] Schneider DL, Bettencourt R, Barrett-Connor E. Clinical utility of spine bone densityin elderly women. J Clin Densitom Jul-Sep 2006;9(3):255–60.

[8] Miyakoshi N, Itoi E, Murai H, Wakabayashi I, Ito H, Minato T. Inverse relation be-tween osteoporosis and spondylosis in postmenopausal women as evaluated bybone mineral density and semiquantitative scoring of spinal degeneration. Spine(Phila Pa 1976) Mar 1 2003;28(5):492–5.

[9] Bergink AP, Uitterlinden AG, Van Leeuwen JP, Hofman A, Verhaar JA, Pols HA. Bonemineral density and vertebral fracture history are associated with incident and pro-gressive radiographic knee osteoarthritis in elderly men andwomen: the RotterdamStudy. Bone Oct 2005;37(4):446–56.

[10] Hart DJ, Cronin C, Daniels M, Worthy T, Doyle DV, Spector TD. The relationship ofbone density and fracture to incident and progressive radiographic osteoarthritisof the knee: the Chingford Study. Arthritis Rheum Jan 2002;46(1):92–9.

U

511


ED P

RO

OF

[11] Roux C, Fechtenbaum J, Briot K, Cropet C, Liu-Leage S, Marcelli C. Inverse relationshipbetween vertebral fractures and spine osteoarthritis in postmenopausal womenwith osteoporosis. Ann Rheum Dis Feb 2008;67(2):224–8.

[12] Sornay-Rendu E, Allard C, Munoz F, Duboeuf F, Delmas PD. Disc space narrowing as anew risk factor for vertebral fracture: the OFELY study. Arthritis Rheum Apr2006;54(4):1262–9.

[13] Arden NK, Griffiths GO, Hart DJ, Doyle DV, Spector TD. The association between os-teoarthritis and osteoporotic fracture: the Chingford Study. Br J Rheumatol Dec1996;35(12):1299–304.

[14] Sornay-Rendu E, Munoz F, Duboeuf F, Delmas PD, Study O. Disc space narrowing isassociated with an increased vertebral fracture risk in postmenopausal women:the OFELY Study. J Bone Miner Res Dec 2004;19(12):1994–9.

[15] Hofman A, van Duijn CM, Franco OH, Ikram MA, Janssen HL, Klaver CC, et al. TheRotterdam Study: 2012 objectives and design update. Eur J Epidemiol Aug2011;26(8):657–86.

[16] Lawton MP, Moss M, Fulcomer M, Kleban MH. A research and service oriented mul-tilevel assessment instrument. J Gerontol Jan 1982;37(1):91–9.

[17] de Schepper EI, Damen J, van Meurs JB, Ginai AZ, Popham M, Hofman A. The associa-tion between lumbar disc degeneration and low back pain: the influence of age, gen-der, and individual radiographic features. Spine (Phila Pa 1976) Mar 1 2010;35(5):531–6.

[18] Burger H, van Daele PL, Algra D, van den Ouweland FA, Grobbee DE, Hofman A, et al.The association between age and bone mineral density in men and women aged55 years and over: the Rotterdam Study. Bone Miner Apr 1994;25(1):1–13.

[19] McCloskey EV, Spector TD, Eyres KS, Fern ED, O'Rourke N, Vasikaran S, et al. The as-sessment of vertebral deformity: a method for use in population studies and clinicaltrials. Osteoporos Int May 1993;3(3):138–47.

[20] van der Klift M, de Laet CE, McCloskey EV, Johnell O, Kanis JA, Hofman A, et al. Riskfactors for incident vertebral fractures in men and women: the Rotterdam Study.J Bone Miner Res Jul 2004;19(7):1172–80.

[21] Turner AS, Maillet JM, Mallinckrodt C, Cordain L. Bone mineral density of the skull inpremenopausal women. Calcif Tissue Int Aug 1997;61(2):110–3.

[22] Pollintine P, Dolan P, Tobias JH, AdamsMA. Intervertebral disc degeneration can leadto “stress-shielding” of the anterior vertebral body: a cause of osteoporotic vertebralfracture? Spine (Phila Pa 1976) Apr 1 2004;29(7):774–82.

[23] Polga DJ, Beaubien BP, Kallemeier PM, Schellhas KP, Lew WD, Buttermann GR.Measurement of in vivo intradiscal pressure in healthy thoracic intervertebraldiscs. Spine (Phila Pa 1976) Jun 15 2004;29(12):1320–4.

[24] Adams MA, Dolan P. Spine biomechanics. J Biomech Oct 2005;38(10):1972–83.[25] Fields AJ, Lee GL, Keaveny TM. Mechanisms of initial endplate failure in the human

vertebral body. J Biomech Dec 1 2010;43(16):3126–31.[26] Hulme PA, Boyd SK, Ferguson SJ. Regional variation in vertebral bone morphology

and its contribution to vertebral fracture strength. Bone Dec 2007;41(6):946–57.[27] Iida T, Abumi K, Kotani Y, Kaneda K. Effects of aging and spinal degeneration on me-

chanical properties of lumbar supraspinous and interspinous ligaments. Spine JMar-Apr 2002;2(2):95–100.

[28] Fujiwara A, Tamai K, An HS, Shimizu K, Yoshida H, Saotome K. The interspinous lig-ament of the lumbar spine. Magnetic resonance images and their clinical signifi-cance. Spine (Phila Pa 1976) Feb 1 2000;25(3):358–63.

[29] Scheele J, de Schepper EI, van Meurs JB, Hofman A, Koes BW, Luijsterburg PA.Association between spinal morning stiffness and lumbar disc degeneration: theRotterdam Study. Osteoarthritis and cartilage/OARS, 20(9). Osteoarthritis ResearchSociety; Sep 2012. p. 982–7.

[30] Grisso JA, Kelsey JL, Strom BL, Chiu GY, Maislin G, O'Brien LA. Risk factors for falls as acause of hip fracture in women. The Northeast Hip Fracture Study Group. N Engl JMed May 9 1991;324(19):1326–31.

[31] Freitas SS, Barrett-Connor E, Ensrud KE, Fink HA, Bauer DC, Cawthon PM, et al. Rateand circumstances of clinical vertebral fractures in older men. Osteoporos Int May2008;19(5):615–23.

[32] Cauley JA, Thompson DE, Ensrud KC, Scott JC, Black D. Risk of mortality followingclinical fractures. Osteoporos Int 2000;11(7):556–61.

[33] OglesbyAK,MinshallME, ShenW,Xie S, Silverman SL. The impact of incident vertebraland non-vertebral fragility fractures on health-related quality of life in establishedpostmenopausal osteoporosis: results from the teriparatide randomized, placebo-controlled trial in postmenopausal women. J Rheumatol Jul 2003;30(7):1579–83.

[34] Adachi JD, Ioannidis G, Olszynski WP, Brown JP, Hanley DA, Sebaldt RJ, et al. The im-pact of incident vertebral and non-vertebral fractures on health related quality of lifein postmenopausal women. BMC Musculoskelet Disord Apr 22 2002;3:11.





Association of lumbar disc degeneration with osteoporotic fractures; the Rotterdam study and...

Documents

Transcript of Association of lumbar disc degeneration with osteoporotic fractures; the Rotterdam study and...