Content uploaded by José M Olmos
Author content
All content in this area was uploaded by José M Olmos on Jul 31, 2017
Content may be subject to copyright.
UNCORRECTED PROOF
1Original Full Length Article
2Genome-wide association study for radiographic vertebral fractures:
3A potential role for the 16q24 BMD locus versus lessons learned from
4challenging phenotype definition
5LingQ1 Oei
a,b,c
, Karol Estrada
a,b,c
, Emma L. Duncan
d,e
, Claus Christiansen
f
, Ching-Ti Liu
g
,BenteL.Langdahl
h
,
6Barbara Obermayer-Pietsch
i
, José A. Riancho
j,k
, Richard L. Prince
l,m
, Natasja M. van Schoor
n
,
7Eugene McCloskey
o,p
, Yi-Hsiang Hsu
q,r
, Evangelos Evangelou
s,t
, Evangelia Ntzani
s
, David M. Evans
u
,
8Nerea Alonso
v
, Lise B. Husted
h
, Carmen Valero
j,k
, Jose L. Hernandez
j,k
, Joshua R. Lewis
l,m
,
9Stephen K. Kaptoge
w
, Kun Zhu
l,m
, L. Adrienne Cupples
g,x
, Carolina Medina-Gómez
a,b,c
, Liesbeth Vandenput
y
,
10 Ghi Su Kim
z
, Seung Hun Lee
z
, Martha C. Castaño-Betancourt
a,b,c
, Edwin H.G. Oei
aa
,Josefina Martinez
ab
,
11 Anna Daroszewska
v
, Marjolein van der Klift
a
, Dan Mellström
y
, Lizbeth Herrera
a
,MagnusK.Karlsson
ac,ad
,
12 Albert Hofman
b
, Östen Ljunggren
ae
, Huibert A.P. Pols
a,b
,LisetteStolk
a,b,c
, Joyce B.J. van Meurs
a,c
,
13 John P.A. Ioannidis
s,af
, M. Carola Zillikens
a,c
, Paul Lips
n,ag
, David Karasik
q,r
, André G. Uitterlinden
a,b,c
,
14 Unnur Styrkarsdottir
ah
, Matthew A. Brown
d
, Jung-Min Koh
y
, J. Brent Richards
t,ai
,JonathanReeve
aj
,
15 Claes Ohlsson
y
, Stuart H. Ralston
v
, Douglas P. Kiel
q,r
, Fernando Rivadeneira
a,b,c,
⁎
16
a
Department of Internal Medicine, Erasmus Medical Center, Rotterdam, The Netherlands
17
b
Department of Epidemiology, Erasmus Medical Center, Rotterdam, The Netherlands
18
c
Netherlands Genomics Initiative (NGI)-sponsored Netherlands Consortium for Healthy Aging (NCHA), The Netherlands
19
d
Human Genetics Group, University of Queensland Diamantina Institute, Brisbane, Queensland, Australia
20
e
Department of Endocrinology, Royal Brisbane and Women's Hospital, Brisbane, Queensland, Australia
21
f
Center for Clinical and Basic Research (CCBR)-Synarc, Ballerup, Denmark
22
g
Department of Biostatistics, Boston University School of Public Health, Boston, MA, USA
23
h
Department of Endocrinology and Internal Medicine, Aarhus University Hospital, Aarhus C, Denmark
24
i
Department of Internal Medicine, Division of Endocrinology and Metabolism, Medical University Graz, Graz, Austria
25
j
Department of Medicine, University of Cantabria, Santander, Spain
26
k
Department of Internal Medicine, Hospital Universitario Marqués de Valdecilla and Instituto de Formación e Investigación Marqués de Valdecilla (IFIMAV), Santander, Spain
27
l
School of Medicine and Pharmacology, University of Western Australia, Perth, Western Australia, Australia
28
m
Department of Endocrinology and Diabetes, Sir Charles Gairdner Hospital, Perth, Western Australia, Australia
29
n
Department of Epidemiology and Biostatistics, Extramuraal Geneeskundig Onderzoek (EMGO) Institute for Health and Care Research, Vrije Universiteit (VU) University Medical Center,
30 Amsterdam, The Netherlands
31
o
National Institute for Health and Research (NIHR), Musculoskeletal Biomedical Research Unit, University of Sheffield, Sheffield, UK
Bone xxx (2013) xxx–xxx
Abbreviations: BMD, bone mineral density; GWAS, genome-wide association study; SNP, single nucleotide polymorphism; GEFOS, Genetic Factors for Osteoporosis; GENOMOS,
Genetic Markers for Osteoporosis; AOGC, Anglo-Australasian Osteoporosis Genetics Consortium; AOGC-GOS, Anglo-Australasian Osteoporosis Genetics Consortium —Geelong
Osteoporosis Study; AROS, Aarhus Osteoporosis Study; CABRIO-C, Cantabria-Camargo study; CABRIO-CC, Cantabria Case–Control study; CAIFOS, Calcium Intake Fracture Outcome
Study; CaMoS, Canadian Multicentre Osteoporosis Study; DOPS, Danish Osteoporosis Prevention Study; HRT, hormone-replacement therapy; EDOS, Edinburgh Osteoporosis Study;
EPOS, European Prospective Osteoporosis Study; EVOS, European Vertebral Osteoporosis Study;FOS, Framingham Osteoporosis Study; CT, computedtomography; KorAMC, Korean oste-
oporosisstudy at Asan MedicalCenter; LASA, LongitudinalAging Study Amsterdam; MrOS Sweden, OsteoporoticFractures in Men Sweden; PERF, ProspectiveEpidemiological Risk Factor;
FN, femoralneck; LS, lumbar spine; DXA, dual-energy X-ray absorptiometry;BMI, body mass index; OR, odds ratio; MAF, minor allele frequency; PCs, principal components; GEE,gener-
alized estimating equation; VACTERL, Vertebral anomalies, Anal atresia, Cardiovascular anomalies, TracheoEsophagealfistula, Renal and Radial anomalies, Limb defects.
⁎Corresponding author at: Departments of Internal Medicine and Epidemiology, P.O. Box 2040 Ee5-79, 3000CA Rotterdam, The Netherlands. Fax: +31 10 7035430.
E-mail addresses: h.l.d.w.oei@erasmusmc.nl (L. Oei), karol@atgu.mgh.harvard.edu (K. Estrada), emma.duncan@uq.edu.au (E.L. Duncan), cc@Nordicbioscience.com (C. Christiansen),
ctliu@bu.edu (C.-T. Liu), bente.langdahl@aarhus.rm.dk (B.L. Langdahl), barbara.obermayer@medunigraz.at (B. Obermayer-Pietsch), jose.riancho@unican.es (J.A. Riancho),
richard.prince@uwa.edu.au (R.L. Prince), nm.vanschoor@vumc.nl (N.M. van Schoor), e.v.mccloskey@sheffield.ac.uk (E. McCloskey), YiHsiangHsu@hsl.harvard.edu (Y.-H. Hsu),
evangel@cc.uoi.gr (E. Evangelou), entzani@gmail.com (E. Ntzani), Dave.Evans@bristol.ac.uk (D.M. Evans), n.alonso@ed.ac.uk (N. Alonso), LISE.BJERRE.HUSTED@KI.AU.DK (L.B. Husted),
Mariacarmen.Valero@uclm.es (C. Valero), joseluishernandez@ono.com (J.L. Hernandez), joshua.lewis@uwa.edu.au (J.R. Lewis), stephen@srl.cam.ac.uk (S.K. Kaptoge),
kzhu@meddent.uwa.edu.au (K. Zhu), adrienne@bu.edu (L.A. Cupples), m.medinagomez@erasmusmc.nl (C. Medina-Gómez), liesbeth.vandenput@medic.gu.se (L. Vandenput),
gskim3@amc.seoul.kr (G.S. Kim), hun0108@amc.seoul.kr (S.H. Lee), m.castanobetancourt@erasmusmc.nl (M.C. Castaño-Betancourt), e.oei@erasmusmc.nl (E.H.G. Oei),
coquem@hotmail.es (J. Martinez), a.daroszewska@ed.ac.uk (A. Daroszewska), m.vanderklift@erasmusmc.nl (M. van der Klift), dan.mellstrom@vgregion.se (D. Mellström),
l.herreraduran@erasmusmc.nl (L. Herrera), magnus.karlsson@med.lu.se (M.K. Karlsson), a.hofman@erasmusmc.nl (A. Hofman), osten.ljunggren@medsci.uu.se (Ö. Ljunggren),
h.pols@erasmusmc.nl (H.A.P. Pols), l.stolk@erasmusmc.nl (L. Stolk), j.vanmeurs@erasmusmc.nl (J.B.J. van Meurs), jioannid@stanford.edu (J.P.A. Ioannidis), m.c.zillikens@erasmusmc.nl
(M.C. Zillikens), p.lips@vumc.nl (P. Lips), karasik@hrca.harvard.edu (D. Karasik), a.g.uitterlinden@erasmusmc.nl (A.G. Uitterlinden), Unnur.Styrkarsdottir@decode.is (U. Styrkarsdottir),
matt.brown@uq.edu.au (M.A. Brown), jmkoh@amc.seoul.kr (J.-M. Koh), brent.richards@mcgill.ca (J.B. Richards), jonathan@srl.cam.ac.uk (J. Reeve), Claes.Ohlsson@medic.gu.se
(C. Ohlsson), Stuart.Ralston@ed.ac.uk (S.H. Ralston), kiel@hsl.harvard.edu (D.P. Kiel), f.rivadeneira@erasmusmc.nl (F. Rivadeneira).
BON-10181; No. of pages: 8; 4C:
8756-3282/$ –see front matter © 2013 Published by Elsevier Inc.
http://dx.doi.org/10.1016/j.bone.2013.10.015
Contents lists available at ScienceDirect
Bone
journal homepage: www.elsevier.com/locate/bone
Please cite this article as: Oei L, et al, Genome-wide association study for radiographic vertebral fractures: A potential role for the 16q24 BMD
locus versus lessons learned from challenging phenotype definition, Bone (2013), http://dx.doi.org/10.1016/j.bone.2013.10.015
UNCORRECTED PROOF
32
p
Academic Unit of Bone Metabolism, Metabolic Bone Centre, University of Sheffield, Sheffield, UK
33
q
Institute for Aging Research, Hebrew SeniorLife, Boston, MA, USA
34
r
Department of Medicine, Harvard Medical School, Boston, MA, USA
35
s
Department of Hygiene and Epidemiology, University of Ioannina School of Medicine, Ioannina, Greece
36
t
Department of Twin Research and Genetic Epidemiology, King's College London, London, UK
37
u
Medical Research Council (MRC) Centre for Causal Analysesin Translational Epidemiology, University of Bristol, Bristol, UK
38
v
Rheumatic Diseases Unit, Institute of Genetics and Molecular Medicine, University of Edinburgh, Edinburgh, UK
39
w
Department of Public Health and Primary Care, University of Cambridge, Cambridge, UK
40
x
Framingham Heart Study, Framingham, MA, USA
41
y
Centre for Bone and Arthritis Research, Institute of Medicine, Sahlgrenska Academy, University of Gothenburg, Gothenburg, Sweden
42
z
Division of Endocrinology and Metabolism, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea
43
aa
Department of Radiology, Erasmus Medical Center, Rotterdam, The Netherlands
44
ab
Q4 Sevicio de Análisis Clínicos, Hospital de Laredo, Laredo, Spain
45
ac
Clinical and Molecular Osteoporosis Research Unit, Department of Clinical Sciences, Lund University, Lund, Sweden
46
ad
Department of Orthopaedics, Malmö University Hospital, Malmö, Sweden
47
ae
Department of Medical Sciences, University of Uppsala, Uppsala, Sweden
48
af
Stanford Prevention Research Center, Stanford University, Stanford, CA, USA
49
ag
Department of Endocrinology, VU University Medical Center, Amsterdam, The Netherlands
50
ah
deCODE Genetics, Reykjavik, Iceland
51
ai
Department of Medicine, Human Genetics and Epidemiology & Biostatistics, Lady Davis Institute, Jewish General Hospital, McGill University, Montreal, Canada
52
aj
Nuffield Department of Orthopaedics, Rheumatology and Musculoskeletal Sciences, University of Oxford, Nuffield Orthopaedic Centre, Oxford, UK
53
54
abstractarticle info
55 Article history:
56 Received 6 February 2013
57 Revised 18 October 2013
58 Accepted 19 October 2013
59 Available online xxxx
60
61 Edited by: Sundeep Khosla
6263
64
65 Keywords:
66 Genome-wide association study
67 Vertebral fracture risk
68 Genetics of osteoporosis
69 GEFOS consortium
70 FOXC2
71 Vertebral fracture risk is a heritable complex trait. The aim of this study was to identifygenetic susceptibility fac-
72 tors for osteoporoticvertebral fractures applying a genome-wide association study (GWAS)approach. The GWAS
73 discovery was based on the Rotterdam Study, a population-based study of elderly Dutch individuals aged
74 N55years; and comprising 329cases and 2666 controls with radiographic scoring (McCloskey–Kanis)and genet-
75 ic data. Replication of one top-associated SNP was pursued by de-novo genotyping of 15 independent studies
76 across Europe, the United States, and Q6Australia and one Asian study. Radiographic vertebral fracture assessment
77 was performed using McCloskey–Kanis or Genant semi-quantitative definitions. SNPs were analyzed in relation
78 to vertebralfracture using logistic regression models corrected for age and sex.Fixed effects inverse variance and
79 Han–Eskin alternative random effects meta-analyses were applied. Genome-wide significance was set at
80 pb5×10
−8
.Q5In the discovery, a SNP (rs11645938) on chromosome 16q24 was associated with the risk for ver-
81 tebral fractures at p=4.6 × 10
−8
. However, the association wasnot significant across 5720cases and 21,791 con-
82 trols from 14 studies. Fixed-effects meta-analysis summary estimate was 1.06 (95% CI: 0.98–1.14; p=0.17),
83 displaying high degreeof heterogeneity (I
2
=57%; Q
het
p= 0.0006). UnderHan–Eskin alternative randomeffects
84 model the summary effect was significant (p=0.0005). The SNP maps to a region previously found associated
85 with lumbar spine bone mineral density (LS-BMD) in two large meta-analyses from the GEFOS consortium. A
86 false positive association in the GWAS discovery cannot be excluded, yet,the low-powered setting of the discov-
87 ery and replication settings (appropriate to identify risk effect size N1.25) may still be consistent with an effect
88 size b1.10, moreof the type expected in complextraits. Larger effortin studies with standardizedphenotype def-
89 initions is needed to confirm or reject the involvement of this locus on the risk for vertebral fractures.
90 © 2013 Published by Elsevier Inc.
9192
93
94
95
Q7 Introduction
96 Vertebralfractures are the most common osteoporotic fractures and
97 represent a significant health issue [1,2]. Epidemiological measures de-
98 rived from population-based studies vary between 1 and 3% per year for
99 incidence and ~10 and 30% for the prevalence in elderly persons, vary-
100 ing by age, gender and geographic region [3–5]. Vertebral fractures are
101 associated with a high morbidity [6–11],mortality[12,13] and a consid-
102 erable financial burden. In the United States the costs of vertebral frac-
103 tures were estimated to be 1.1 billion dollars in the year 2005, and are
104 expected to rise by more than 50% by the year 2025 [14]. A recent report
105 estimated the costs of vertebral fracturesin Europe at 1.5billion euros in
106 2010 [15]. Furthermore, vertebral fractures are likely to become an in-
107 creasingly important health issue with the increasing ageof populations
108 [1,14,15] and their association with increased risk of future osteoporotic
109 fractures at other skeletal sites [7,16,17]. For all of these reasons, a better
110 understanding of the genetic susceptibility to vertebral fracture has the
111 potential to identify underlying biological mechanisms, improve risk
112 prediction and lead to novel disease interventions.
113 Vertebralfracture risk is a heritable complex trait, also influenced by
114 environmental, and gene–environment interactions [18,19]. A positive
115 family history for vertebral fracture constitutes an independent risk fac-
116 tor for future fractures [20], emphasizing the importance of genetics in
117the pathogenesis of the disease. The hypothesis-free genome-wide as-
118sociation study (GWAS) approach has been particularly successful in
119identifying loci associated with many diseases and quantitative com-
120plex traits [21], including osteoporosis [18,22–24].
121The aim of our study was to better understand the genetic architec-
122ture of radiographic vertebral fractures by conducting the first GWAS
123for this trait in a large population-based study of elderly Dutch individ-
124uals and pursuing replication in a large set of studies across Europe, the
125United States, Australia and Asia.
126Methods
127Datasets assessed
128Sample discovery phase
129The discovery sample was confined to the original Rotterdam Study
130cohort, a large population-based study of Dutch men and women
131aged 55 years and over (mean age at vertebral fracture assessment:
13273.5years). A detailed description of the Rotterdam Study has been re-
133ported previously [25]. In short, the study aimed to assess the incidence
134and determinants of disease anddisability in elderly persons.The study
135has been approved by the Medical Ethics Committee of the Erasmus
136University Medical Center Rotterdam.
2L. Oei et al. / Bone xxx (2013) xxx–xxx
Please cite this article as: Oei L, et al, Genome-wide association study for radiographic vertebral fractures: A potential role for the 16q24 BMD
locus versus lessons learned from challenging phenotype definition, Bone (2013), http://dx.doi.org/10.1016/j.bone.2013.10.015
UNCORRECTED PROOF
137 Sample replication phase
138 The Genetic Factors for Osteoporosis (GEFOS), Genetic Markers for
139 Osteoporosis (GENOMOS) and Anglo-Australasian Osteoporosis Genet-
140 ics Consortium (AOGC) are three consortia studying the genetic deter-
141 minants of osteoporosis-related skeletal phenotypes in populations
142 with available DNA and/or GWAS data [23,26–29]. Within this setting,
143 15 studies with both DNA samples and lateral morphometry-derived
144 vertebral fracture data participated in the replication phase of this pro-
145 ject (Supplementary Table 1). More detailed descriptions are available
146 in the Supplementary material.
147 The AOGC —Geelong Osteoporosis Study (AOGC-GOS) is a
148 cohort drawn from the Geelong general population. Vertebral fracture
149 imaging was performed in case of a clinical indication [30,31].The
150 AOGC —Sheffield (AOGC-SHEFFIELD) study constitutes a large
151 population-based cohort of community-dwelling elderly women aged
152 ≥75 years in Sheffield, UK [32]. AROS (Aarhus Osteoporosis Study) is a
153 case–control study, including 462 osteoporotic patients (vertebral
154 fracture and T-score b−2.5) and 336 controls [33]. AUSTRIOS is a pro-
155 spective cohort study of elderly female patients above 70 recruited in
156 95 nursing homes in four counties in Austria. The AUSTRIOS-B cohort
157 had vertebral fracture data available and was used for this project
158 [34]. The Cantabria-Camargo (CABRIO-C) and Cantabria Case–Control
159 (CABRIO-CC) studies are based in Northern Spain. CABRIO-C is a
160 community-based study designed to evaluate the prevalence of meta-
161 bolic bone diseases in postmenopausal women and men older than
162 50 years attending a primary care center in Santander [35,36].
163 CABRIO-CC is a clinic-based study of control individuals and patients
164 with osteoporosis living in Cantabria, a region in Northern Spain
165 [37,38]. The Calcium Intake Fracture Outcome Study (CAIFOS) is a
166 randomized-controlled trial investigating calcium carbonate supple-
167 mentation in ambulatory women older than 70 years recruited in
168 Perth, Australia [39]. The Canadian Multicentre Osteoporosis Study
169 (CaMoS) is a population-based prospective cohort of unrelated men
170 and women followed for osteoporosis and osteoporotic fractures for
171 the past 14 years [40–42]. The Danish Osteoporosis Prevention Study
172 (DOPS) is a population-based study of perimenopausal women. The
173 women were followed for 10years and approximately 35% were treat-
174 ed with hormone-replacement therapy (HRT) [43]. The Edinburgh Os-
175 teoporosis Study (EDOS) consists of a clinical referral population of
176 patients assessed for evaluation of osteoporosis in Edinburgh, United
177 Kingdom. The European Prospective Osteoporosis Study (EPOS) is the
178 prospective phase of the European Vertebral Osteoporosis Study
179 (EVOS) in which population-based samples had paired duplicate spinal
180 films. Men and women from 36 centers in 19 European countries were
181 recruited [5,44,45]. The Framingham Osteoporosis Study (FOS) is an
182 ancillary study of the Framingham Study, a multigenerational family-
183 based cohort study originally initiated to study the risk factors for car-
184 diovascular disease [46–48]. Vertebral fracture assessment was done
185 on multidetector computed tomography (CT) lateral scout views. The
186 Korean osteoporosis study at Asan Medical Center (KorAMC) study is
187 a hospital registered, cross-sectional study of postmenopausal Korean
188 women in Seoul [49]. The Longitudinal Aging Study Amsterdam
189 (LASA) is an ongoing multidisciplinary cohort study in older persons.
190 A random sample of men and women aged 55years and over, stratified
191 by age, sex, urbanization grade and expected 5-year mortality rate was
192 drawn from the population register of Amsterdam, The Netherlands
193 [50]. The Osteoporotic Fractures in Men Sweden (MrOS Sweden)
194 study is a multicenter, prospective study including elderly men. Study
195 subjects (men aged 69–80 years) were randomly identified using na-
196 tional population registers, contacted and asked to participate.
197 Eligible subjects had to be able to walk without assistance, provide
198 self-reported data, and sign an informed consent [51]. The Prospective
199 Epidemiological Risk Factor (PERF) Study is based on subjects who
200 were screened for or enrolled into randomized controlled clinical trial
201 to identify genetic and other risk factorsof diseases in the elderly in Co-
202 penhagen, Denmark [52].
203Phenotyping
204Osteoporosis-related skeletal phenotypes in the discovery sample
205During the second follow-up visit between 1997 and 1999 all
206Rotterdam Study participants underwent radiographic screening. A
207trained research technician obtained lateral radiographs of the
208thoracolumbar spine following a standard protocol. Radiographs were
209evaluated morphometrically in Sheffield, UK, by the McCloskey–Kanis
210method as described previously [53]. Using this method, central col-
211lapse, anterior and posterior wedge, and crush deformities were identi-
212fied based on a cut point of 3 standard deviation height reductions. All
213vertebral fractures were confirmed by visual interpretation by an expert
214in the field to rule out artifacts and other etiologies, such as pathological
215fractures. Cases were defined as those individuals who had at least one
216vertebral fracture, and controls were defined as those who were free of
217vertebral fractures. Bone mineral density (BMD) of the femoral neck
218(FN) and lumbar spine (LS) was measured by dual-energy X-ray ab-
219sorptiometry (DXA), using a Lunar DPX-L densitometer (Lunar Radia-
220tion Corporation, Madison, WI, USA).
221Other measurements (covariates) in the discovery sample
222An extensive baseline home interview on medical history, risk
223factors for chronic diseases, and medication use was performed on all
224participants by trained interviewers. Smoking habits were coded as
225“current”,“former”and “never”. Self-reported age at natural meno-
226pause between 40 and 60 years, defined as 12 months after periods
227ceased, was collected retrospectively. Information on medication use
228included hormone replacement therapy and systemic corticosteroids.
229Alcohol intake was assessed from a validated semi-quantitative food-
230frequency questionnaire. Height and weight were measured with in-
231door clothing and no shoes. Body mass index (BMI) was calculated as
232weight (in kg) / height (in m
2
).
233Phenotyping replication phase
234Vertebral fracture assessments differed by cohorts which applied
235either the McCloskey–Kanis [53] or the Genant semi-quantitative
236method [54]. Detailed description of the methods and cut-offs applied
237by each study is available in Table 1. Four of the replication studies
238used the McCloskey–Kanis method, which is similar to the discovery
239(Rotterdam Study), of which one study applied the same additional cri-
240terion of absolute height reduction. Phenotyping for covariates wassim-
241ilar to that of the discovery sample.
242Genotyping
243Genome-wide association data
244The Rotterdam Study participants were genotyped using the
245Illumina Infinium HumanHap550 Beadchip in the Genetic Laboratory
246of Erasmus MC Department of Internal Medicine, The Netherlands, fol-
247lowing manufacturers' protocols and quality control standards.
248Single nucleotide polymorphism (SNP) genotyping
249The top associated SNP from the discovery phase (rs11645938) was
250genotyped in 15 studies within three maingenotyping centers: deCODE
251Genetics in Reykjavik, Iceland, Queensland University in Brisbane
252Australia and KBiosciences, Hertfordshire, U.K. (www.kbioscience.co.
253uk). Genotyping was carried out by personnel blinded to patient status
254in all centers. The samples genotyped by KBiosciences were part of the
255GENOMOS consortium DNA collection, and comprise most of the partic-
256ipating studies. For KBiosciences, a minimum of 1.5μlofDNAat3.3ng/μl
257(when quantitated by PicoGreen analysis or 7 ng/μlifquantitatedby
258spectrophotometry) was required for one SNP to be assayed using
259their proprietary KASPar PCR technique and Taqman (also used by
260Brisbane University for AOGC samples). Genotype calling was carried
261out using an automated system, the results of which werechecked man-
262ually by study personnel using SNPviewer software (KBiosciences).
3L. Oei et al. / Bone xxx (2013) xxx–xxx
Please cite this article as: Oei L, et al, Genome-wide association study for radiographic vertebral fractures: A potential role for the 16q24 BMD
locus versus lessons learned from challenging phenotype definition, Bone (2013), http://dx.doi.org/10.1016/j.bone.2013.10.015
UNCORRECTED PROOF
263 deCODE used the same KASPar assay from KBiosciences to genotype the
264 PERF study samples. To ensure genotyping validity across study centers,
265 a reference plate wasshipped from KBiosciences tothe AOGC coordinat-
266 ing center. To ensure correct genotyping deCODE Genetics genotyped
267 92 HapMap samples for comparison with the KASPar assay, and
268 both positive and negative samples were present on all genotyping
269 plates. Additionally, duplicate SNP genotyping was performed in the
270 Rotterdam Study (all samples) and CABRIO-C (random selection of
271 187 samples) and no discrepancies were found.
272 Statistical methods
273 Within the discovery cohort, we tested 2,543,887 genotyped or im-
274 puted (HapMap CEU release 22, build 36) [55,56] SNPs for association
275 with risk of osteoporotic vertebral fractures using a logistic regression
276 model (MACH2DAT) [57,58] adjusted for age, gender, and admixture
277 principal components (PCs) derived using EIGENSTRAT to adjust for
278 population substructure [59]. Potential effect modifiers for the relation-
279 ship between genotype and vertebral fracture (i.e. height, weight, BMI,
280 age at menopause, HRT use, corticosteroid use, N3 units alcohol use
281 per day, current and ever smoking) were tested by adding them one
282 at a time to the regression model and evaluating the change in both
283 the effect estimate and significance. The GWAS was performed using a
284 web-based interface (GRIMP) on scalable super-computing grid infra-
285 structures [60]. At a genome-wide significant α-level of 5 × 10
−8
,the
286 design had 0.80 power to detect risk effect sizes (OR) of 1.8 to 2.1 for
287 minor allele frequencies (MAF) of 20% to 10%, respectively.
288 Replication analyses
289 Except for the FOS and AOGC studies, all analyses were carried out
290 centrally by the Rotterdam Coordinating Center. Again a logistic regres-
291 sion model adjusting for age and gender was used. Individuals with ei-
292 ther missing genotype or phenotype data were excluded from analysis.
293 Initially, fixed effects inverse variance meta-analysis was performed
294 (METAL software [61]). The presence of statistically significant hetero-
295 geneity was assessed by Cochran's Q statistic (Q
het
p)andtheextent
296 of the observed heterogeneity was measured by the I
2
metric. Han–
297 Eskin alternative random effects meta-analysis was applied when the
298 I
2
metric exceeded 50% as this model is optimized to detect associations
299under heterogeneity (Metasoft software [62]). SPSS 16.0, PLINK, and R
300software were used for therest of the analyses. In addition, the Framing-
301ham Study analysis used population-based generalized estimating
302equation (GEE) approach correcting for correlations owing to family re-
303lationships and PCs. The replication setting incorporating 5720 cases
304and 21,791 controls from 14 studies was powered to identify a variant
305with a MAF of 0.10 and risk effect size N1.25, associated at pb5×10
−8
.
306Results
307The description of the studies included in the discovery and replica-
308tion phases is shown in Supplementary Table 1. Description of the ver-
309tebral fracture assessment done across studies is presented in Table 1
310while baseline characteristics of the study populations are shown in
311Supplementary Table 2. In the discovery set, 329 of the 2995 Rotterdam
312Study participants had at least one vertebral fracture evident on the
313spinal radiographs. A genotyped SNP (rs11645938) on chromosome
31416q24 (MAF = 10%) was associated at a genome-wide significant level
315(p=4.6×10
−8
) with an increased risk of vertebral fractures (Fig. 1).
316Compared to the risk of non-carriers, the odds of the heterozygous car-
317riers of theminor allele (C) was 1.7times higher (95% confidence inter-
318val [CI] 1.3–2.3) and that of the homozygous carriers was 5.8 times
319higher (95% CI 2.7–12.8) (Supplementary Fig. 1). Fig. 2 shows the re-
320gional association plot of the locus, where a cluster of FOX genes maps
321~200 kb from the associated SNP, containing FOXF1,MTHFSD,FOXC2,
322and FOXL1. Further adjusting for potential confounders did not influ-
323ence either the effect estimate or the significance of the association be-
324tween genotype and vertebral fracture risk. Similarly, the association
325remained significant after adjustment for either LS- or FN-BMD. Sex-
326stratified association analysis for the SNP, showed similar effect esti-
327mates (OR heterozygotemen: 1.8 [95% CI: 1.2–2.8] and OR heterozygote
328women: 1.6 [95%CI: 1.1–2.3]; OR homozygote women: 8.4 [95% CI: 3.0–
32923.0] and OR homozygote men 3.3 [95% CI: 0.9–12.7]).
330The associated SNP rs11645938 was successfully genotyped in 14
331of the replication studies (5722 vertebral fracture cases and 21,793
332controls; MAF ~8–12%) while it was found to be monomorphic in the
333Korean population of the KorAMC study (Table 2). The summary effect
334estimate for vertebral fracture risk obtained from the meta-analysis
335was 1.06 (95% CI: 0.98–1.14; p=0.17) and the effect estimate displayed
Table 1t1:1Q2
t1:2Vertebral fracture assessment.
t1:3Cut-off values used Fractures
confirmed
by expert
a
t1:4Comparison setting McCloskey–Kanis Genant
t1:5Study Morphometry
method used
Number of vertebral
fracture cases
Number of vertebral
fracture controls
Prevalence or
case:control
ratio
b
Relative to
population
reference
Absolute height
reduction
c
3 SD relative
reduction
d
Difference in
vertebral height
ratios of at least:
t1:6RS-I McCloskey–Kanis 329 2666 0.11 Yes Yes Yes NA Yes
t1:7AOGC McCloskey–Kanis 686 3411 0.17 Yes No Yes NA Yes
t1:8AROS Genant 335 130 1:0.39
b
No No NA 20% Yes
t1:9AUSTRIOS-B Genant 803 1261 0.39 No No NA 20% Yes
t1:10 CABRIO-C Genant 195 1185 0.14 No No NA 20% Yes
t1:11 CABRIO-CC Genant 220 354 1:1.61
b
No No NA 20% Yes
t1:12 CAIFOS McCloskey–Kanis 428 600 0.42 Yes No Yes NA Yes
t1:13 CaMoS McCloskey–Kanis 243 1785 0.12 Yes No NA NA Yes
t1:14 DOPS Genant 108 1605 0.06 No No NA 20% Yes
t1:15 EDOS Genant 495 523 0.49 No No NA 20% Yes
t1:16 EPOS McCloskey–Kanis 313 1779 0.15 Yes Yes Yes NA Yes
t1:17 FOS Genant 417 2291 0.15 No No NA 20% Yes
t1:18 KorAMC Genant 101 1193 0.08 No Yes NA 20% Yes
t1:19 LASA Genant 237 268 0.47 No No NA 20% Yes
t1:20 MrOS Sweden
e
Genant 309 2613 0.11 No No NA 20% No
t1:21 PERF Genant 830 2793 0.23 No No NA 20% No
a
Q3 E.g. radiologist/clinician,to rule out artifacts andother etiologies, such as pathological fractures.
b
Prevalence in population-based studies, case:control ratio in case–control studies.t1:23
c
Any of the three vertebral heights (anterior, central, orposterior) shows a minimum decrease of at least 4 mm.t1:24
d
3 SD relativereduction of 2 out of 3 ratios: (ha/hp; hm/hp; hp/hp predicted).t1:25
e
Prevalent X-ray verified vertebral fractures only available for about 1425 subjects.t1:26
4L. Oei et al. / Bone xxx (2013) xxx–xxx
Please cite this article as: Oei L, et al, Genome-wide association study for radiographic vertebral fractures: A potential role for the 16q24 BMD
locus versus lessons learned from challenging phenotype definition, Bone (2013), http://dx.doi.org/10.1016/j.bone.2013.10.015
UNCORRECTED PROOF
336 high degree of heterogeneity with I
2
= 57% and Q
het
p=0.0006(Fig. 3).
337 When considering a Han–Eskin alternative random effects meta-
338 analysis model the summary effect was significant (p= 0.0005).
339 When applying more stringent genotyping criteria (call rate N95%;
340 Hardy–Weinberg equilibrium pN0.05) the association became signifi-
341 cant in both the fixed (p=0.045) and Han–Eskin alternative random ef-
342 fects meta-analysis (p=0.0002). When further restricting analyses only
343 to those studies that used the McCloskey–Kanis assessment a consis-
344 tent, nonetheless not a statistically significant, effect direction was ob-
345 served (replication p= 0.29).
346 Discussion
347 To our knowledge, this is the first GWAS for radiographically deter-
348 mined vertebral fracture. A marker on chromosome 16q24 was
349 genome-wide significantly associated with vertebral fracture in the
350Rotterdam Study discovery set. However, this association was not sig-
351nificant in a replication effort including 15 studies world-wide using
352conventional statistical analysis techniques.
353Work by Stankiewicz et al. implicated deletions/mutations in this
35416q24 locus in the VACTERL association (Vertebralanomalies, Analatre-
355sia, Cardiovascular anomalies, TracheoEsophageal fistula, Renal and Ra-
356dial anomalies, Limb defects), a non-random association of birth defects
357that includes vertebral defects [63].FOXC2, mapping ~200 kb upstream
358from the associated SNP, is highly expressed in human bone tissue, and
359its expression is regulated by bone morphogenetic proteins [64].The
360gene is involved in osteoblast differentiation through activation of ca-
361nonical Wnt/β-catenin signals [65], and in mice Foxc2 functions as a
362transcription factor essential for axial skeletogenesis [66]. The vertebral
363fracture associated SNP maps to a region previously found to be associ-
364ated with LS-BMD in a meta-analysis of 19,125 individuals [23] and fur-
365ther replicated in 83,894 individuals [22]. However, the vertebral
Fig. 1. Manhattan plo t of negative logarithm p-values plotted by chromosome, showing that aSNP on chromosome 16q24 was associated at a genome-wide significant level with oste-
oporotic vertebral fractures (p= 4.6 × 10
−8
) in the Rotterdam Study (encircled).
Fig. 2. Regional association plot showing position on chromosome 16 and association p-values of the analyzed SNPs in the Rotterdam Study with neighboring genes. Included are geno-
typed, HapMap II and 1000 Genomes imputed SNPs. The rectangle is the SNP of interest, and the circles represent neighboring SNPs withtheir respective correlation withthe top marker.
The spikes depict the recombination rates. The position of the rs10048146 SNP that has previously been found as associated with lumbar-spine bonemineral density is indicated with *.
5L. Oei et al. / Bone xxx (2013) xxx–xxx
Please cite this article as: Oei L, et al, Genome-wide association study for radiographic vertebral fractures: A potential role for the 16q24 BMD
locus versus lessons learned from challenging phenotype definition, Bone (2013), http://dx.doi.org/10.1016/j.bone.2013.10.015
UNCORRECTED PROOF
366 fracture SNP was not associated with either LS- or FN-BMD in our study
367 and this signal was independent of the one previously reported for the
368 BMD SNP rs10048146 (r
2
= 0.002).
369 Despite the underlying biological plausibility supporting this associ-
370 ation and even with identifying a genome-wide significant signal in the
371 discovery GWAS, replication in independent studies is still needed
372 [21,67,68]. Subsequently, de-novo direct genotyping of rs11645938 in
373 5720 cases and 21,791 controls, from multiple independent studies
374 around the world, did not provide robust evidence for replication of
375 the association. Therefore, there is a high likelihood of the signal being
376 a false-positive signal. It is expected that discoveries at underpowered
377 settings would have low positive predictive value for true findings and
378 this applies even for signals that pass a stringent genome-wide signifi-
379 cance threshold [69]. However, other considerations might have also
380 contributed to an apparent lack of replication of a potentially true asso-
381 ciation,and these will serve to inform the design of future GWAS of the
382 vertebral fracture phenotype.
383Signals in underpowered settings are likely to display inflated effects
384due to the “winner's curse”phenomenon, where the effect estimate ob-
385served in the first study overestimates the actual risk observed at the
386general population level [70–72]. According to a post-hoc calculation
387for the replication phase, the current design had merely 0.42 power to
388detect an OR of 1.2. The study sample should have included more than
3898000 cases to achieve 0.80 power, and we know that typically GWAS
390of complex traits even requires close to30,000 cases to identify trulyas-
391sociated SNPs with moderate allele frequencies (e.g. MAF = 0.10) in a
392powered setting. Previous efforts have pointed out that SNPs with
393MAF b10% tend to be difficult to replicate due to the lack of statistical
394power [78]. Thus, we cannot yet exclude the possibility that the identi-
395fied association has a very small, yet genuine effect [73]. Larger-scale
396GWAS meta-analyses for osteoporotic vertebral fractures are seriously
397needed.
398Q8GWAS studies rely on the principle of linkage disequilibrium (LD)
399where markers are tested under the assumption they tag an underlying
400causal genetic variant. When the linkage disequilibrium structure in the
401region differs across populations this may result in decreased power
402and lack of replication [74–77]. The rs11645938 marker is not in LD
403with any other marker contained in HapMap and only in moderate LD
404with one marker (r
2
= 0.41 with rs11647070) from the 1000 Genomes
405Project. This observation led us to conclude that existing GWAS without
406the rs11645938 on their arrays would be poorly imputed, which was
407the case in the FOS and AOGC studies, and therefore to overcome this,
408de-novo genotyping of the marker was performed in these studies.
409However, strictly speaking, genotyping in the Australian AOGC and
410CAIFOS studies did not attain conventional criteria for unknown rea-
411sons. Further, the SNP is monomorphic in Asian populations.
412Despite the fact that all studies used radiological assessments, a crit-
413ical issue to bear in mind is the phenotype definition, considering that
414diverse methods and cut-offs exist for the assessment of vertebral frac-
415tures [79]. Phenotype measurement differences are a known possible
416source of heterogeneity, which might be reflected in our study by the
417great variation in vertebral fracture prevalences among the studies. No-
418ticeably, prevalence estimates varied between 6% and 49% in the cohort
419studies. Furthermore, quantitative scoring is based on morphometry
420alone, which may result in inclusion of deformities into the phenotype
421definition that are not truly vertebral fractures [80]. These non-
422fracture deformities are frequently labeled as Genant grade 1 or “mild
423vertebral fractures,”when, in fact, they may be normal variations in ver-
424tebral shape. Therefore, many studies assign an expert to filter out these
425non-fracture deformities. Nevertheless, this triage procedure may not
426have been sufficiently standardized, and this could have introduced
427the statistically significant heterogeneity between studies. Several
428methods exist to explore the existence of associations in heterogeneous
Fig. 3. Forest plot showing meta-analysis results of vertebral fracture risk for rs11645938
in discoveryand replication studies. Effectestimates represented by squaresare displayed
on a logarithmic scale, with horizontal lines corresponding to 95% confidence intervals.
The centerline of the diamond standsfor the overall summary measure, and its horizontal
line indicates the 95% confidence interval.
Table 2t2:1
t2:2Descriptive information about genotyping of the rs11645938 SNP andassociation statistics per study.
t2:3Study SNP call rate p-Value Hardy–Weinberg
equilibrium
Minor allele frequency Effect estimate (beta) Standard error p-Value
t2:4RS-I 99.9% 0.52 9.65% 0.669 0.122 4.6E−08
t2:5AOGC 93.7% 0.83 9.74% −0.11 0.11 0.33
t2:6AROS 99.3% 0.79 9.76% 0.22 0.31 0.61
t2:7AUSTRIOS-B 97.0% 0.96 11.74% −0.18 0.16 0.26
t2:8CABRIO-C 99.1% 0.84 7.71% −0.51 0.25 0.04
t2:9CABRIO-CC 99.1% 0.61 8.98% −0.16 0.26 0.53
t2:10 CAIFOS 99.2% 0.02 10.01% −0.02 0.15 0.91
t2:11 CaMoS 99.0% 0.12 9.57% 0.09 0.16 0.50
t2:12 DOPS 98.6% 0.17 10.01% −0.13 0.25 0.59
t2:13 EDOS 99.4% 0.99 9.45% 0.07 0.16 0.65
t2:14 EPOS 99.6% 0.80 9.62% 0.15 0.16 0.75
t2:15 FOS 97.6% 0.86 9.97% −0.04 0.14 0.78
t2:16 KorAMC 97.8% NA 0.00% NA NA NA
t2:17 LASA 100.0% 0.82 11.16% −0.06 0.22 0.79
t2:18 MrOS Sweden 98.6% 0.49 9.77% 0.07 0.16 0.69
t2:19 PERF 100.0% 0.79 9.45% 0.04 0.10 0.70
6L. Oei et al. / Bone xxx (2013) xxx–xxx
Please cite this article as: Oei L, et al, Genome-wide association study for radiographic vertebral fractures: A potential role for the 16q24 BMD
locus versus lessons learned from challenging phenotype definition, Bone (2013), http://dx.doi.org/10.1016/j.bone.2013.10.015
UNCORRECTED PROOF
429 data and when we applied a Han–Eskin random effects model, more
430 stringent genotyping criteria or sensitivity analyses for phenotype defi-
431 nition, the results became more consistent. Perhaps selecting Genant
432 grade 2 and 3 types including “moderate”and “severe”vertebral frac-
433 tures [81] could provide a better phenotype definition for future genetic
434 studies. In fact, Liu and colleagues demonstra ted that the heritability of a
435 stricter phenotype (when only more severe deformities counted) was
436 higher than considering all vertebral deformities together [19]. There-
437 fore, phenotype standardization among meta-analysis participants can
438 be a key in replication [71,82]. Unfortunately, data harmonization was
439 not possible because severity grading or qualitative standardized read-
440 ing to enable data harmonization was not available for most of the stud-
441 ies included in our analysis. This consideration, along with the relatively
442 small sample sizes across replication studies,Q9 is a major hurdle to be
443 overcome in future studies focusing on radiographic vertebral fractures.
444 Clinical vertebral fracture is an alternative phenotype definition for fu-
445 ture genetic studies, though achieving sufficient sample sizes will be
446 also challenging; considering that only a small fraction of vertebralfrac-
447 tures come to clinical attention (i.e. are symptomatic). In addition, it
448 would be valuable to gain more insight into incident vertebral fractures.
449 Nevertheless, definition of incident vertebral fractures is accompanied
450 by different and possibly greater precision errors than identification of
451 prevalent vertebral fractures. On the other hand, by comparing images
452 at different follow-ups, the radiological reader has the opportunity to
453 correct possible misclassifications, including misattributions of baseline
454 deformities as fracture cases caused by erroneous vertebral heightread-
455 ings due to for example superimposition of other structures or magnifi-
456 cation errors [83–86].
457 In conclusion, although a GWAS in the population-based Rotterdam
458 Study identified a marker mapping to the 16q24 (FOXC2)BMDlocusas
459 being genome-wide significantly associated with radiographic vertebral
460 fracture in that population, this could not be conclusively replicated by
461 de-novo genotyping across 15 studies worldwide. A false positive asso-
462 ciation in the GWAS discovery cannot yet be excluded. However, these
463 results from a low-powered setting may still be consistent with a
464 small true effect size as is common in complex traits. Larger efforts in
465 subsequent GWAS for radiographic vertebral fracturewith standardized
466 phenotype definitions may confirm or reject the involvement of this
467 locus on the risk for vertebral fractures.
468 Disclosures
469 All authors state that they have no conflicts of interest.
470 Appendix A. Supplementary data
471 Supplementary data to this article can be found online at http://dx.
472 doi.org/10.1016/j.bone.2013.10.015.
473 References
474 [1] Szulc P, Bouxsein ML. Overview of osteoporosis: epidemiology and clinical manage-
475 ment. Vertebral fracture initiative resource document; 2011.
476 [2] Brunton S, Carmichael B, Gold D, Hull B, Kauffman T, Papaioannou A, et al. Vertebral
477 compression fractures in primary care: recommendations from a consensus panel.
478 J Fam Pract 2005;54:781–8.
479 [3] Samelson EJ, Hannan MT, Zhang Y, GenantHK, Felson DT, Kiel DP. Incidence andrisk
480 factors for vertebral fracturein women and men: 25-year follow-up results from the
481 population-based Framingham study. J Bone Miner Res 2006;21:1207–14.
482 [4] Van der Klift M, De Laet CE, McCloskey EV, Hofman A, Pols HA. The incidence of
483 vertebral fractures in men and women: the Rotterdam Study. J Bone Miner Res
484 2002;17:1051–6.
485 [5] Incidence of vertebral fracture in Europe: results from the European Prospective
486 Osteoporosis Study (EPOS). J Bone Miner Res 2002;17:716–24.
487 [6] BurgerH, Van Daele PL,Grashuis K,Hofman A, Grobbee DE,Schutte HE,et al. Vertebral
488 deformities and functional impairment in men and women. J Bone Miner Res
489 1997;12:152–7.
490 [7] Ross PD. Clinical consequences of vertebral fractures. Am J Med 1997;103:30S–42S
491 [discussion 42S-43S].
492[8] Nevitt MC, Ettinger B, Black DM, Stone K, Jamal SA, Ensrud K, et al. The association of
493radiographically detected vertebral fractures with back pain and function: a pro-
494spective study. Ann Intern Med 1998;128:793–800.
495[9] Gold DT. The clinical impact of vertebral fractures: quality of life in women with os-
496teoporosis. Bone 1996;18:185S–9S.
497[10] Oleksik A, Lips P, Dawson A, Minshall ME, Shen W, Cooper C, et al. Health-related
498quality of life in postmenopausal women with low BMD with or without prevalent
499vertebral fractures. J Bone Miner Res 2000;15:1384–92.
500[11] Oleksik AM, Ewing S, Shen W, van Schoor NM, Lips P. Impact of incident vertebral
501fractures on health related quality of life (HRQOL) in postmenopausal women
502with prevalent vertebral fractures. Osteoporos Int 2005;16:861–70.
503[12] Bliuc D, Nguyen ND, Milch VE, Nguyen TV, Eisman JA, Center JR. Mortality risk asso-
504ciated with low-trauma osteoporotic fracture and subsequent fracture in men and
505women. JAMA 2009;301:513–21.
506[13] Ensrud KE, Thompson DE, Cauley JA, Nevitt MC, Kado DM, Hochberg MC, et al.
507Prevalent vertebral deformities predict mortality and hospitalization in older
508women with low bone mass. J Am Geriatr Soc 2000;48:241–9.
509[14] Burge R, Dawson-Hughes B, Solomon DH, Wong JB, King A, Tosteson A. Incidence
510and economic burden of osteoporosis-related fractures in the United States, 2005–
5112025. J Bone Miner Res 2007;22:465–75.
512[15] Ström O, Borgström F, Kanis JA, Compston J, Cooper C, McCloskey EV, et al. Osteoporosis:
513burden, health care provision and opportunities in the EU. Arch Osteoporos 2011.
514http://dx.doi.org/10.1007/s11657-011-0060-1.
515[16] Melton III LJ, Atkinson EJ, Cooper C, O'Fallon WM, Riggs BL. Vertebral fractures pre-
516dict subsequent fractures. Osteoporos Int 1999;10:214–21.
517[17] Black DM, Arden NK, Palermo L, Pearson J, Cummings SR, Study of Osteoporotic
518Fractures Research Group. Prevalent vertebral deformities predict hip fractures
519and new vertebral deformities but not wrist fractures. J Bone Miner Res 1999;14:
520821–8.
521[18] Ralston SH, Uitterlinden AG. Genetics of osteoporosis. Endocr Rev 2010;31:629–62.
522[19] Liu CT, Karasik D, Zhou Y, Hsu YH, GenantHK, Broe KE, et al. Heritability of prevalent
523vertebral fracture and volumetric bone mineral density and geometry at the lumbar
524spine in three generations of the Framingham study. J Bone Miner Res 2011. Q11
525
[20] Kanis JA, Johansson H, Oden A, Johnell O, De Laet C, Eisman JA, et al. A family history
526of fracture and fracture risk: a meta-analysis. Bone 2004;35:1029–37.
527[21] Manolio TA. Genomewide association studies and assessment of the risk of disease.
528N Engl J Med 2010;363:166–76.
529[22] Estrada K, Styrkarsdottir U, Evangelou E, Hsu YH, Duncan EL, Ntzani EE, et al.
530Genome-wide meta-analysis identifies 56 bone mineral density loci and reveals 14
531loci associated with risk of fracture. Nat Genet 2012;44:491–501.
532[23] Rivadeneira F, Styrkarsdottir U, EstradaK, Halldorsson BV, Hsu YH, Richards JB, et al.
533Twenty bone-mineral-density loci identified by large-scale meta-analysis of
534genome-wide association studies. Nat Genet 2009;41:1199–206.
535[24] Richards JB,Zheng HF, Spector TD.Genetics of osteoporosis from genome-wideasso-
536ciation studies: advances and challenges. Nat Rev Genet 2012;13:576–88.
537[25] Hofman A, van Duijn CM, Franco OH, Ikram MA, Janssen HL, Klaver CC, et al. The
538Rotterdam Study: 2012 objectives and design update. Eur J Epidemiol 2011;26:
539657–86.
540[26] Langdahl BL, Uitterlinden AG, Ralston SH, Trikalinos TA, Balcells S, Brandi ML, et al.
541Large-scale analysis of association between polymorphisms in the transforming
542growth factor beta 1 gene (TGFB1) and osteoporosis: the GENOMOS study. Bone
5432008;42:969–81.
544[27] Ralston SH, Uitterlinden AG, Brandi ML, Balcells S, Langdahl BL, Lips P, et al. Large-
545scale evidence for the effect of the COLIA1 Sp1 polymorphism on osteoporosis
546outcomes: the GENOMOS study. PLoS Med 2006;3:e90.
547[28] Duncan EL, Danoy P, Kemp JP, Leo PJ, McCloskey E, Nicholson GC, et al. Genome-
548wide association study using extreme truncate selection identifies novel genes af-
549fecting bone mineral density and fracture risk. PLoS Genet 2011;7:e1001372.
550[29] Sims AM, Shephard N, Carter K, Doan T, Dowling A, Duncan EL, et al. Genetic analy-
551ses in a sampleof individuals withhigh or low BMD shows association withmultiple
552Wnt pathway genes. J Bone Miner Res 2008;23:499–506.
553[30] Henry MJ, Pasco JA, Korn S, Gibson JE, Kotowicz MA, Nicholson GC. Bone mineral
554density reference ranges for Australian men: Geelong Osteoporosis Study.
555Osteoporos Int 2010;21:909–17.
556[31] Henry MJ, Pasco JA, Nicholson GC, Seeman E, Kotowicz MA. Prevalence of osteoporo-
557sis in Australian women: Geelong Osteoporosis Study. J ClinDensitom 2000;3:261–8.
558[32] McCloskey EV, Beneton M, Charlesworth D, Kayan K, deTakats D, Dey A, et al.
559Clodronate reduces the incidence of fractures in community-dwelling elderly
560women unselected for osteoporosis: results of a double-blind, placebo-controlled
561randomized study. J Bone Miner Res 2007;22:135–41.
562[33] Harslof T, Husted LB, Carstens M, Stenkjaer L, Langdahl BL. Genotypes and haplo-
563types of the estrogen receptor genes, but not the retinoblastoma-interacting zinc
564finger protein 1 gene, are associated with osteoporosis. Calcif Tissue Int 2010;87:
56525–35.
566[34] Dobnig H, Piswanger-Solkner JC, Roth M, Obermayer-Pietsch B, Tiran A, Strele A,
567et al. Type 2 diabetes mellitus in nursing home patients: effects on bone turnover,
568bone mass, and fracture risk. J Clin Endocrinol Metab 2006;91:3355–63.
569[35] Martinez J, Olmos JM, Hernandez JL, Pinedo G, Llorca J, Obregon E, et al. Bone turn-
570over markers in Spanish postmenopausal women: the Camargo cohort study. Clin
571Chim Acta 2009;409:70–4.
572[36] Olmos JM, Hernandez JL, Martinez J, Pariente E, Llorca J, Gonzalez-Macias J. Bone
573turnover markers in Spanish adult men: the Camargo cohort study. Clin Chim Acta
5742010;411:1511–5.
575[37] Riancho JA, Valero C, Naranjo A, Morales DJ, Sanudo C, Zarrabeitia MT. Identification
576of an aromatase haplotype that is associated with gene expression and postmeno-
577pausal osteoporosis. J Clin Endocrinol Metab 2007;92:660–5.
7L. Oei et al. / Bone xxx (2013) xxx–xxx
Please cite this article as: Oei L, et al, Genome-wide association study for radiographic vertebral fractures: A potential role for the 16q24 BMD
locus versus lessons learned from challenging phenotype definition, Bone (2013), http://dx.doi.org/10.1016/j.bone.2013.10.015
UNCORRECTED PROOF
578 [38] Zarrabeitia MT, Hernandez JL, Valero C, Zarrabeitia A, Amado JA, Gonzalez-Macias J,
579 et al. Adiposity, estradiol, and genetic variants of steroid-metabolizing enzymes as
580 determinants of bone mineral density. Eur J Endocrinol 2007;156:117–22.
581 [39] Prince RL, Devine A, Dhaliwal SS, Dick IM. Effects of calcium supplementation on
582 clinical fracture and bone structure: results of a 5-year, double-blind, placebo-
583 controlled trial in elderly women. Arch Intern Med 2006;166:869–75.
584 [40] Richards JB,Leslie WD, Joseph L, Siminoski K, Hanley DA, Adachi JD, et al. Changes to
585 osteoporosis prevalence according to method of risk assessment. J Bone Miner Res
586 2007;22:228–34.
587 [41] Richards JB, Papaioannou A, Adachi JD, Joseph L, Whitson HE, Prior JC, et al. Effect of
588 selective serotonin reuptake inhibitors on the risk of fracture. Arch Intern Med
589 2007;167:188–94.
590 [42] TenenhouseA, Joseph L, Kreiger N,Poliquin S,Murray TM, BlondeauL, et al. Estimation
591 of the prevalence of low bone density in Canadian women and men using a
592 population-specific DXA reference standard: the Canadian Multicentre Osteoporosis
593 Study (CaMos). Osteoporos Int 2000;11:897–904.
594 [43] Mosekilde L, Hermann AP, Beck-Nielsen H, Charles P, Nielsen SP, Sorensen OH. The
595 Danish Osteoporosis Prevention Study (DOPS): project design and inclusion of
596 2000 normal perimenopausal women. Maturitas 1999;31:207–19.
597 [44] Ismail AA, O'Neill TW, Cockerill W, Finn JD, Cannata JB, Hoszowski K, etal. Validity of
598 self-report of fractures: results from a prospective study in men and women across
599 Europe. EPOS Study Group. European Prospective Osteoporosis Study Group.
600 Osteoporos Int 2000;11:248–54.
601 [45] O'Neill TW, Felsenberg D, Varlow J, Cooper C, Kanis JA, Silman AJ. The prevalence of
602 vertebral deformityin European men and women: the European Vertebral Osteopo-
603 rosis Study. J Bone Miner Res 1996;11:1010–8.
604 [46] Dawber TR, Meadors GF, Moore Jr FE. Epidemiological approaches to heart disease:
605 the Framingham study. Am J Public Health Nations Health 1951;41:279–81.
606 [47] Kannel WB, Feinleib M, McNamara PM, Garrison RJ, CastelliWP. An investigation of
607 coronary heart disease in families. The Framingham offspring study. Am J Epidemiol
608 1979;110:281–90.
609 [48] Splansky GL, Corey D, Yang Q, Atwood LD, Cupples LA, Benjamin EJ, et al. The Third
610 Generation Cohort of the National Heart, Lung, and Blood Institute's Framingham
611 Heart Study: design, recruitment, and initial examination. Am J Epidemiol 2007;165:
612 1328–35.
613 [49] Koh JM, Oh B, Lee JY, Lee JK, Kimm K, Park BL, et al. Association of FLT3 polymor-
614 phisms with low BMD and risk of osteoporotic fracture in postmenopausal
615 women. J Bone Miner Res 2007;22:1752–8.
616 [50] Deeg DJ, van Tilburg T, Smit JH, de Leeuw ED. Attrition in the Longitudinal Aging
617 Study Amsterdam. The effect of differential inclusion in side studies.J Clin Epidemiol
618 2002;55:319–28.
619 [51] MellstromD, Johnell O, LjunggrenO, Eriksson AL, Lorentzon M, MallminH, et al. Free
620 testosterone is an independent predictor of BMD and prevalent fractures in elderly
621 men: MrOS Sweden. J Bone Miner Res 2006;21:529–35.
622 [52] Bagger YZ, Rasmussen HB, Alexandersen P, Werge T, Christiansen C, Tanko LB. Links
623 between cardiovascular disease and osteoporosis in postmenopausal women:
624 serum lipids or atherosclerosis per se? Osteoporos Int 2007;18:505–12.
625 [53] McCloskey EV, Spector TD, Eyres KS, Fern ED, O'Rourke N, Vasikaran S, et al. The as-
626 sessment of vertebral deformity: a method for use in population studies and clinical
627 trials. Osteoporos Int 1993;3:138–47.
628 [54] Genant HK, Wu CY, van Kuijk C, Nevitt MC. Vertebral fracture assessment using a
629 semiquantitative technique. J Bone Miner Res 1993;8:1137–48.
630 [55] The International HapMap Project. Nature 2003;426:789–96.
631 [56] Altshuler DM, Gibbs RA, Peltonen L, Dermitzakis E,Schaffner SF, Yu F, et al. Integrat-
632 ing common and rare genetic variation in diverse human populations. Nature
633 2010;467:52–8.
634 [57] Li Y, Willer C, Sanna S, Abecasis G. Genotype imputation. Annu Rev Genomics Hum
635 Genet 2009;10:387–406.
636 [58] Li Y, Willer CJ, Ding J, Scheet P, AbecasisGR. MaCH: using sequence andgenotype data
637 to estimate haplotypes and unobserved genotypes. Genet Epidemiol 2010;34:816–34.
638 [59] Price AL, Patterson NJ,Plenge RM, Weinblatt ME, Shadick NA, ReichD. Principal com-
639 ponents analysis corrects for stratification in genome-wide association studies. Nat
640 Genet 2006;38:904–9.
641 [60] Estrada K, Abuseiris A, Grosveld FG, Uitterlinden AG, Knoch TA, Rivadeneira F.
642 GRIMP: a web- and grid-based tool for high-speed analysis of large-scale genome-
643 wide association using imputed data. Bioinformatics 2009;25:2750–2.
644 [61] Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide
645 association scans. Bioinformatics 2010;26:2190–1.
646 [62] Han B, Eskin E. Random-effects model aimed at discovering associations in meta-
647 analysis of genome-wide association studies. Am J Hum Genet 2011;88:586–98.
648[63] Stankiewicz P, Sen P, Bhatt SS, Storer M, Xia Z, Bejjani BA, et al. Genomic and genic
649deletions of the FOX gene cluster on 16q24.1 and inactivating mutations of FOXF1
650cause alveolar capillary dysplasia and other malformations. Am J Hum Genet
6512009;84:780–91.
652[64] NifujiA,MiuraN,KatoN,KellermannO,NodaM.Bonemorphogeneticproteinregula-
653tion of forkhead/winged helix transcription factor Foxc2 (Mfh1) in a murine mesoder-
654mal cell line C1 and in skeletal precursor cells. J Bone Miner Res 2001;16:1765–71.
655[65] Kim SH, Cho KW, Choi HS, Park SJ, Rhee Y, JungHS, et al. The forkhead transcription
656factor Foxc2 stimulates osteoblast differentiation. Biochem Biophys Res Commun
6572009;386:532–6.
658[66] Winnier GE, Hargett L, Hogan BL. The winged helix transcription factor MFH1 is re-
659quired for proliferation and patterning of paraxial mesoderm in the mouse embryo.
660Genes Dev 1997;11:926–40.
661[67] Easton DF, Pooley KA, Dunning AM, Pharoah PD, Thompson D, Ballinger DG, et al.
662Genome-wide association study identifies novel breast cancer susceptibility loci.
663Nature 2007;447:1087–93.
664[68] Hoover RN. The evolution of epidemiologic research: from cottage industry to “big”
665science. Epidemiology 2007;18:13–7.
666[69] Panagiotou OA, Ioannidis JP. What should the genome-wide significance threshold
667be? Empirical replication of borderline genetic associations. Int J Epidemiol
6682012;41:273–86.
669[70] Ioannidis JP. Why most discovered true associations are inflated. Epidemiology
6702008;19:640–8.
671[71] Kraft P, Zeggini E, Ioannidis JP. Replication in genome-wide association studies. Stat
672Sci 2009;24:561–73.
673[72] Zhong H, Prentice RL. Correcting “winner's curse”in odds ratios from genomewide
674association findings for major complex human diseases. Genet Epidemiol 2010;34:
67578–91.
676[73] LohmuellerKE, Pearce CL, Pike M, Lander ES, Hirschhorn JN. Meta-analysis of genetic
677association studies supports a contribution of common variants to susceptibility to
678common disease. Nat Genet 2003;33:177–82.
679[74] Montpetit A, Nelis M, Laflamme P, Magi R, Ke X, Remm M, et al. An evaluation of the
680performance of tag SNPs derived from HapMap in a Caucasian population. PLoS
681Genet 2006;2:e27.
682[75] Chen F, Stram DO, Le Marchand L, Monroe KR, Kolonel LN, Henderson BE, et al.
683Caution in generalizing known genetic risk markers for breast cancer across all
684ethnic/racial populations. Eur J Hum Genet 2011;19:243–5.
685[76] He J, WilkensLR, Stram DO, Kolonel LN, HendersonBE, Wu AH, et al. Generalizability
686and epidemiologic characterization of eleven colorectal cancer GWAS hits in multi-
687ple populations. Cancer Epidemiol Biomarkers Prev 2011;20:70–81.
688[77] Sim X, Ong RT, Suo C, Tay WT, Liu J, Ng DP, et al. Transferability of type 2 diabetes
689implicated loci in multi-ethnic cohorts from Southeast Asia. PLoS Genet 2011;7:
690e1001363.
691[78] Stromberg U, Bjork J, Vineis P, Broberg K, Zeggini E. Ranking of genome-wide asso-
692ciation scan signals by different measures. Int J Epidemiol 2009;38:1364–73.
693[79] Oei L, Rivadeneira F, Ly F, Breda SJ, Zillikens MC, Hofman A,et al. Review of radiolog-
694ical scoring methods of osteoporotic vertebral fractures for clinical and research set-
695tings. Eur Radiol 2012. Q12
696
[80] Puisto V, Heliovaara M, Impivaara O, Jalanko T, Kroger H, Knekt P, et al. Severity of
697vertebral fracture and risk of hip fracture: a nested case–control study. Osteoporos
698Int 2011;22:63–8.
699[81] Jiang G, Eastell R, Barrington NA, Ferrar L. Comparison of methods for the visual
700identification of prevalent vertebral fracture in osteoporosis. Osteoporos Int
7012004;15:887–96.
702[82] Kerkhof HJ, Meulenbelt I, Akune T, Arden NK, Aromaa A, Bierma-Zeinstra SM, et al.
703Recommendations for standardization and phenotype definitions in genetic studies
704of osteoarthritis: the TREAT-OA consortium. Osteoarthritis Cartilage 2011;19:254–64.
705[83] Reeve J, Lunt M, Felsenberg D, SilmanAJ, Scheidt-Nave C, Poor G, et al. Determinants
706of the size of incident vertebral deformities in European men and women in the
707sixth to ninth decades of age: the European Prospective Osteoporosis Study
708(EPOS). J Bone Miner Res 2003;18:1664–73.
709[84] Felsenberg D, Silman AJ, Lunt M, Armbrecht G, Ismail AA, Finn JD, et al. Incidence of
710vertebral fracture in Europe: results from the European Prospective Osteoporosis
711Study (EPOS). J Bone Miner Res 2002;17:716–24.
712[85] Lunt M, Ismail AA, Felsenberg D, Cooper C, Kanis JA, Reeve J, et al. Defining incident
713vertebral deformities in population studies: a comparison of morphometric criteria.
714Osteoporos Int 2002;13:809–15.
715[86] Q13Lunt M, Gowin W, Johnell, Armbrecht G, Felsenberg D. A statistical method to min-
716imize magnification errors in serial vertebral radiographs. Osteoporos Int 2001;12:
717909-13.
718
719
8L. Oei et al. / Bone xxx (2013) xxx–xxx
Please cite this article as: Oei L, et al, Genome-wide association study for radiographic vertebral fractures: A potential role for the 16q24 BMD
locus versus lessons learned from challenging phenotype definition, Bone (2013), http://dx.doi.org/10.1016/j.bone.2013.10.015
