• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of ajhgLink to Publisher's site
Am J Hum Genet. Dec 2004; 75(6): 1032–1045.
Published online Oct 26, 2004. doi:  10.1086/426458
PMCID: PMC1182139

Polymorphisms in the Sclerosteosis/van Buchem Disease Gene (SOST) Region Are Associated with Bone-Mineral Density in Elderly Whites


Osteoporosis has a strong genetic component, but the genes involved are poorly defined. We studied whether the sclerosteosis/van Buchem disease gene (SOST) is an osteoporosis-risk gene by examining its association with bone-mineral density (BMD). Mutations in SOST result in sclerosteosis, and alterations in the SOST gene expression may be causal in the closely related van Buchem disease. We used a set of eight polymorphisms from the SOST gene region to genotype 1,939 elderly men and women from a large population-based prospective-cohort study of Dutch whites. A 3-bp insertion (f=0.38) in the presumed SOST promoter region (SRP3) was associated with decreased BMD in women at the femoral neck (FN) (P=.05) and lumbar spine (LS) (P=.01), with evidence of an allele-dose effect in the oldest age group (P=.006). Similarly, a G variant (f=0.40) in the van Buchem deletion region (SRP9) was associated with increased BMD in men at the FN (P=.007) and LS (P=.02). In both cases, differences between extreme genotypes reached 0.2 SD. We observed no genotype effects on fracture risk, for the 234 osteoporotic fractures validated during 8.2 years of follow-up and for the 146 vertebral prevalent fractures analyzed. We did not find association between any of several frequent haplotypes across the SOST gene region and BMD. We did find evidence of additive effects of SRP3 with the COLIA1 Sp1 polymorphism but not with haplotypes of 3′ polymorphisms in the vitamin-D receptor gene. The SOST-COLIA1 additive effect increased with age and reached 0.5 SD difference in BMD at LS in the oldest age group (P=.02). The molecular mechanism whereby these moderate SOST genotype effects are mediated remains to be elucidated, but it is likely to involve differences in regulation of SOST gene expression.


Osteoporosis is a common disorder characterized by decreased bone-mineral density (BMD), degenerative microarchitectural changes in bone tissue, and an increased fracture risk. Certain aspects of the disorder show strong genetic influences. For example, the heritability of BMD has been estimated to be up to 80%, which suggests the existence of “bone density” genes (Smith et al. 1973; Pocock et al. 1987; Cummings et al. 1995; Flicker et al. 1995). Variance between individuals in the activity of these genes may be manifested in different ways—for example, by differences in peak bone mass or rates of bone loss with increasing age. Although fracture risk is also strongly inherited, it does not correlate absolutely with BMD, which suggests that bone density and bone architecture and/or strength probably are influenced by both shared and distinct genetic factors (Cummings et al. 1995; Flicker et al. 1996). In addition to genetic traits, the expression of osteoporosis is influenced by environmental factors, such as diet, exercise, and exposure to tobacco, alcohol, and sunlight. This multifactorial and polygenetic character makes osteoporosis a so-called complex disease.

A number of approaches have been taken to try to understand the genetic basis of osteoporosis (Uitterlinden et al. 2001; Ralston 2002). Although genome scans in large collections of related individuals have identified a number of potentially interesting chromosomal loci (Devoto et al. 1998, 2001; Duncan et al. 1999; Koller et al. 2000), significant linkage to the same loci is rarely observed in more than one study. Furthermore, such relatively large chromosomal areas still have to be searched for candidate genes, variants of which could explain the original linkage found.

In contrast, the candidate-gene approach builds on biological knowledge to implicate particular genes in bone metabolism. Evidence that would implicate certain genes can include specific expression and function in bone cells, genetically manipulated animal models with a bone phenotype, and human Mendelian disease that expresses a bone phenotype. The contribution of genetic variations in candidate genes involved in bone metabolism (formation, remodeling, microarchitecture, etc.) and bone density or fracture risk has been investigated in population-based association studies. So far, polymorphisms in several candidate genes have been investigated, including those encoding collagen type I α1 (Grant et al. 1996; Uitterlinden et al. 1998), vitamin-D receptor (VDR) (Morrison et al. 1994; Uitterlinden et al. 1996), and estrogen receptor (Kobayashi et al. 1996; Becherini et al. 2000; van Meurs et al. 2003). “Risk alleles” for some of these genes have been observed in some populations, but the polymorphisms are generally associated with modest effects on the phenotypes of interest. Of these, the collagen type I α1 Sp1 polymorphism has been found, in a number of studies, to be consistently associated with differences in bone density and fracture risk, as demonstrated by two meta-analyses (Efstathiadou et al. 2001; Mann et al. 2001).

The gene that encodes sclerostin, or SOST, is mutated in the disorder sclerosteosis (SCL [MIM 269500]), which is characterized by massive, progressive bone overgrowth throughout life (Hansen 1967; Beighton 1988; Balemans et al. 2001; Brunkow et al. 2001). Along with thicker-than-normal trabeculae and cortices, the BMD in affected individuals can be up to twice that of unaffected individuals (Epstein et al. 1979; Stein et al. 1983). In contrast to conditions that result in osteopetrosis, SCL is most likely a defect in osteoblast function, as evidenced by elevated alkaline phosphatase, increased rates of bone formation, and generally normal numbers and activity of osteoclasts. No obvious endocrinological abnormalities have been reported. van Buchem disease (VBD [MIM 239100]) is a similar disorder, with generally less-severe clinical outcomes (van Buchem et al. 1955, 1976; Beighton et al. 1984). Although the exact molecular mechanism for its manifestation is unknown, a 52-kb deletion located 35 kb downstream of the SOST gene was suggested to play some role, perhaps by affecting expression of SOST and/or other tightly linked gene(s) (Balemans et al. 2002b; Staehling-Hampton et al. 2002).

Sclerostin is a cystine knot–containing, secreted glycoprotein most similar in sequence to the DAN (differential screening-selected gene aberrative in neuroblastoma) family of bone-morphogenetic-protein (BMP) antagonists (Brunkow et al. 2001) and more distantly related to the BMP antagonists noggin and chordin. Members of the DAN family—including DAN, Cerberus/CER1, Gremlin/DRM/CKTSF1B1, PRDC, Caronte, and Dante (Hsu et al. 1998; Pearce et al. 1999)—exert their effects by negatively regulating the activity of specific BMPs in a wide variety of processes, such as differentiation, mitogenesis, chemotaxis, and apoptosis. The SOST gene is expressed most highly in cells involved in osteogenesis—specifically differentiated osteoblasts and osteocytes—and, like its related family members, has been shown to act via specific BMPs (Winkler et al. 2003).

This evidence, together with the very specific effect on bone mass seen in the autosomal recessive disorder SCL, led us to ask whether the SOST gene plays a role in the genetic control of bone mass and/or predisposition to bone fracture in elderly individuals. We report on the relationship of eight polymorphisms across the SOST gene region to bone density and to the occurrence of osteoporotic fractures in 1,939 elderly individuals from a population-based prospective-cohort study of diseases in the elderly in the Netherlands—the Rotterdam Study (Hofman et al. 1991). Furthermore, we studied the interaction between polymorphisms in the gene that encodes sclerostin and polymorphisms in two genes previously implicated by us and others to play a role in the genetic control of bone density and fractures—that is, VDR and the collagen type I α1 gene (COLIA1).

Subjects and Methods

Study Subjects

Subjects were participants of the Rotterdam Study, a population-based cohort study of subjects aged [gt-or-equal, slanted]55 years who live in the Ommoord district of Rotterdam. The study was designed to investigate the incidence and determinants of chronic disabling diseases in the elderly. The rationale and design have been described elsewhere (Hofman et al. 1991). All 10,275 Rotterdam inhabitants who were aged [gt-or-equal, slanted]55 years were invited for baseline examination, between August 1990 and June 1993. Of those, 7,983 participated. Among the subjects living independently, the overall response rate was 77% for home interview and 71% for examination at a research center, where anthropometric characteristics and BMD were measured and blood samples were taken. The Rotterdam Study was approved by the medical ethics committee of the Erasmus University Medical School, and written informed consent was obtained from each subject.

For the present study, a subgroup of men and women was analyzed (see also Uitterlinden et al. [1998]). Baseline measurements of BMD were available for 5,931 subjects who were living independently. Of those, 1,453 were excluded because they were aged >80 years, used a walking aid, had diabetes mellitus, or were taking drugs known to affect bone metabolism (diuretics, estrogen, thyroid hormone, or cytostatic drugs). Of the 4,478 remaining subjects, we studied a random sample of 1,939 subjects comprising 1,016 women and 923 men.

Clinical Examination and Measurements

Height and weight were measured at baseline examination, with the subject in a standing position and wearing indoor clothing but not shoes. BMD (in g/cm2) was determined by dual-energy X-ray absorptiometry (DEXA) (DPX-L densitometer [Lunar]) at the femoral neck (FN) and lumbar spine (LS) (vertebrae L2, L3, and L4), as described elsewhere (Burger et al. 1994). Dietary intake of calcium (in mg/d) during the preceding year was assessed by a food-frequency questionnaire and was adjusted for energy intake. Age at menopause and current use of cigarettes were assessed by a questionnaire.

Assessment of Incident Nonvertebral Fracture

Follow-up started either January 1, 1991, or at the time of inclusion in the study, if later. For this analysis, follow-up ended either by January 1, 2000, or at the participant’s death, whichever came first. The general practitioners of the participants provided data on morbidity, including nonvertebral fractures, and mortality. For ~80% of the study population, medical events were reported through computerized registers of general-practitioner diagnosis. For the remaining 20%, research physicians collected data from the general practitioners’ medical records of the study participants. All collected data on fractures were verified by review of discharge reports and letters from medical specialists. Fracture events were coded independently by two research physicians, in accordance with the International Classification of Diseases, 10th revision. In case of discrepancy, consensus was attained in a separate session. A medical expert in the field reviewed all coded events for final classification.

Assessment of Vertebral Fracture

Both at baseline and at follow-up visits, between 1997 and 1999, thoracolumbar radiographs of the spine were obtained. The follow-up radiographs were available for the 1,184 individuals who survived an average of 7.4 years after the center visit for baseline assessment and who were still able to come to our research center. All follow-up radiographs were scored for the presence of vertebral fracture by the McCloskey/Kanis method (McCloskey et al. 1993), as described elsewhere (Van der Klift et al. 2002; Schuit et al. 2004). If a vertebral fracture was detected, the baseline radiograph was evaluated as well. If the vertebral fracture was already present at baseline, it was considered a baseline-prevalent fracture. If it was not present at baseline, the fracture was defined to be an incident. For 600 women and 519 men, lateral radiographs of the spine from the 4th thoracic to the 5th lumbar vertebra were measured at follow-up and were examined for the presence of prevalent vertebral fractures by morphometric analysis, as described above.

Identification of SOST-Region Polymorphisms (SRPs)

We used a number of different approaches to search for polymorphisms in the SOST gene region (see Brunkow et al. [2001] and GenBank [accession number AF326736] for complete description of the SOST gene structure): (1) a 21.5-kb genomic fragment—including both coding exons, the ~2.8-kb intron, and ~12 kb of 5′ and ~4.5 kb of 3′ flanking regions—was sequenced to completion from four sources (Afrikaner, Dutch white, and Senegalese individuals and human BAC clone 202l9 from California Institute of Technology BAC B and C libraries); (2) the two coding exons were sequenced to completion in a set of 90 ethnically diverse individuals from the National Institute of General Medical Sciences (NIGMS) Human Diversity Panel (panels HD01–HD09); (3) an additional 87.6-kb region 3′ of the gene was sample sequenced in the four sources of (1) above, by use of PCR primer pairs distributed across the region; and (4) a sequence from the SOST gene region found in National Center for Biotechnology Information (NCBI) databases was examined in silico. The examined SOST-gene flanking regions correspond to the DUSP-SOST and SOST-MEOX1 intergenic intervals (GenBank accession numbers AF326736 and AF397423). All polymorphisms found by these approaches were also compared with those reported in a study by Balemans et al. (2002b).

Determination of Genotypes

Genomic DNA was extracted from samples of peripheral venous blood in accordance with standard procedures. Polymorphism-containing regions in the SOST gene locus were amplified from genomic DNA with PCR. Each PCR was performed in a 10-μl reaction volume that contained 5 ng of genomic DNA, 1.5 mM MgCl2, 0.2 mM deoxy-NTP, 2 pmol of each primer (see table 1), 0.2 units of Taq polymerase (Promega), and 10× PCR buffer (Promega) that contained 20 mM Tris-HCl (pH 8.0), 100 mM KCl, 0.1 mM EDTA, 1 mM DDT, 50% glycerol, 0.5% Nonidet-P40, and 0.5% Tween20. The reactions were performed in a 384-well thermocycler (MJ Research Tetrad), with different cycling protocols for each amplicon (table 1). The genotypes were detected with the single-base extension (SBE) procedure by use of SBE primers of different lengths (table 1). The SBE reactions were performed in accordance with details provided by the manufacturer (ABI Prism SNaPshot Multiplex Kit), with slight modifications. The genotypes thus generated were analyzed with software programs Gene Mapper 1.1 and Genotyper 3.7 (both from Applied Biosystems) and were checked visually. To confirm the accuracy of the genotyping, 150 randomly selected samples were genotyped for a second time with the same method. No discrepancies were found.

Table 1
Rotterdam Study SRPs and Primers for SBE Genotyping

Three anonymous polymorphic restriction-enzyme-recognition sites at the 3′ end of the gene encoding VDR—that is, for BsmI, ApaI and TaqI—were assessed in relation to each other by a direct molecular haplotyping PCR procedure, as described elsewhere (Uitterlinden et al. 1996). This allowed us to determine the phase of the alleles at each of the RFLP loci. As a result, three frequent haplotype alleles were discerned—encoded 1 (baT; frequency in whites 48%), 2 (BAt; 40%), and 3 (bAT; 10%)—that combined to become six genotypes encoded 11, 12, 13, 22, 23, and 33. We excluded the less frequent haplotypes (4 and 5) from the analysis. Women with genotypes containing these two haplotypes represent 1.5% of this population. Detailed information on haplotype alleles and genotype frequencies in the Rotterdam Study can be found elsewhere (Uitterlinden et al. 1996). The Sp1 polymorphism of COLIA1 was detected by PCR with a mismatched primer that introduces a diallelic restriction site, as described elsewhere (Uitterlinden et al. 1998). The test discriminates two alleles, G and T, at position exon 2 −348, which is the first base in the last of four Sp1-binding sites located in the first intron of COLIA1 on 17q21 (Mann et al. 2001).

Statistical Analysis

Linkage disequilibrium (LD) and haplotype analysis

The LD coefficient (D′) between each pair of SNPs was calculated with the help of the software programs PHASE (Stephens et al. 2001) and HaploXT. PHASE was also used to reconstruct haplotypes over the SOST gene region (SRP1–6) for each individual. Hardy Weinberg equilibrium (HWE) was calculated, in accordance with standard procedures, by use of the χ2 analysis.

Association analysis

Subjects were grouped on the basis of genotype. We grouped subjects by allele copy number (0, 1, and 2, corresponding to noncarriers, heterozygous carriers, and homozygous carriers, respectively) for each allele (or haplotype). We allowed for three possible genetic models to explain differences between groups—that is, an allele-dose effect, a dominant effect, or a recessive effect. “Allele dose” was defined as the number of copies of a certain allele in the genotype. In case of a consistent trend reflected as an allele-dose effect, we performed a linear regression analysis to quantify the association. In case of a dominant or recessive effect of the test allele, analysis of (co)variance (ANCOVA) was performed to test for differences between two genotype groups. For dominant alleles, we compared test-allele carriers versus noncarriers, whereas, for recessive effects, subjects homozygous for the test allele were compared with heterozygous carriers combined with noncarriers.

To study interaction of genetic variants (i.e., SOST with VDR and SOST with COLIA1), subjects were analyzed in four groups on the basis of presence or absence of risk alleles for each gene. For reasons of power, we combined heterozygous and homozygous subjects for the risk allele into carriers of at least one copy of the risk allele.

Odds ratios (ORs) with 95% CIs were calculated by logistic-regression analyses to estimate the relative risk of fractures by genotype for the risk allele; the reference group comprised those with no copies of the risk allele. We first calculated crude ORs and then adjusted for potentially confounding factors (e.g., age, BMI, BMD, and age at menopause). To estimate nonvertebral fracture risk by genotype, we used Cox proportional hazard models, thereby taking into account potential differences in follow-up time. To estimate the risk of vertebral fractures, ORs with 95% CIs were calculated using logistic-regression models. We were not able to use Cox proportional hazard models for vertebral fractures, since the exact time of the events were not known. For reasons of power, all vertebral-fracture analyses were done with combined prevalent and incident vertebral fractures.

All statistical analyses were performed using SPSS version 10.1.0. P values are two-sided, and P[less-than-or-eq, slant].05 was considered significant.


SOST Genotype, Allele Frequencies, and HWE

We identified a number of informative polymorphisms in the SOST gene/VBD region by a variety of methods (see the “Subjects and Methods” section). We found only a single coding-region polymorphism that affected the amino acid sequence of sclerostin. This G11988A variant (SRP4) results in a Val→Ile change in the signal sequence. Our analysis of the SOST intron and flanking regions identified an additional 30 SRPs (fig. 1). We determined allele frequencies for 15 of the SRPs in a set of 90 normal DNA samples, which represent 10 individuals of African, 30 of Asian, and 50 of white descent. Nine SRPs were chosen for further study (SRP1–9) on the basis of their distribution across the SOST gene locus and their representation of a wide range (2%–50%) of minor-allele frequencies. SRP1 (C4102G), SRP2 (C10356T), and SRP3 (10565insGGA) are located in the 5′ flanking region; SRP4 (G11988A) is in the sclerostin signal sequence; and SRP5 (C17965G), SRP6 (A18292G), SRP7 (T42722C), SRP8 (A58874G), and SRP9 (A75707G) are located 3′ of the gene, the latter three falling within the region that is deleted in subjects with VBD (fig. 1) (Staehling-Hampton et al. 2002). There were no differences found in allele frequencies between men and women (results not shown).

Figure  1
Structure of SOST gene region and overview of polymorphisms. The SRPs described in this study span ~87.3 kb of genomic sequence. The two SOST exons are indicated by solid boxes, and the 52-kb VBD deletion region is indicated by the heavy horizontal ...

Although present in the NIGMS samples at a frequency of 1.7%, the minor allele of SRP4 (Val/Ile in SOST signal sequence) was not detected in a subset of the Rotterdam elderly white sample and so was excluded from the association study. The genotype distributions for the remaining eight SRPs in the Rotterdam Study were found to be in HWE (P>.05), except for SRP2 (P=.02) (table 1).

LD and Haplotype Analysis

To get more insight into the pattern of LD between alleles at polymorphic loci, pairwise disequilibria measures (D′) were calculated (fig. 2). Strong (although not complete) LD was observed between the SRPs surrounding the SOST gene (SRP1–3 and SRP5–6) and, to a lesser extent, between SRP7, -8 and -9. Using the PHASE program, we reconstructed haplotype alleles for each subject on the basis of the SRPs surrounding the SOST gene (SRP1–6). These are shown in figure 1, together with their frequencies in the 1,939 subjects from the Rotterdam Study. Because of the strong LD, we observed 4 frequent (>5% population frequency) haplotype alleles and 13 less frequent haplotype alleles among the 32 possible haplotypes. The four frequent haplotypes accounted for 91% of the chromosomes in our population. No differences in frequencies were observed between men and women.

Figure  2
Pairwise LD D′ values of eight SRPs across the SOST gene region. The software program PHASE was used to calculate values by use of true genomic distance (A) and by use of the SRPs at equal distance from each other (B).

Association between Genotype and BMD

We next analyzed the association of SOST gene polymorphisms with bone characteristics for the individual polymorphisms and for the haplotype alleles. We found a significant association of BMD with SRP3 in women and with SRP9 in men but not with the other six polymorphisms. Tables Tables22 and and33 show baseline characteristics by SRP3 genotypes for women and SRP9 genotypes for men, respectively. For both SRP3 and SRP9, the three genotype groups did not differ significantly in age, anthropometric characteristics, dietary calcium intake, or, for women, age at menopause.

Table 2
Characteristics of 1,016 Postmenopausal Women, by SRP3 Genotype
Table 3
Characteristics of 923 Men, by SRP9 Genotype

Table 2 shows FN and LS BMD measurements in postmenopausal women, grouped by the three SRP3 genotypes. At both sites, BMD was highest in the homozygous DEL-DEL group and lowest in the homozygous GGA-GGA group; heterozygotes had intermediate values, which indicates an allele-dose effect. BMD in the DEL-GGA group was 2% less than the DEL-DEL group at both sites, whereas, in the GGA-GGA group, BMD was 4% less than that in the DEL-DEL group (table 2) (P=.01 for FN; P=.004 for LS). The effects of the SRP3 genotype on BMD at the Ward’s triangle and trochanter sites were similar (data not shown). When the BMD values were adjusted for possible confounders, the differences decreased somewhat. In men, no association of SRP3 with BMD was found. When we compared SRP3-allele frequencies between groups composed of women in the highest and lowest quintiles of BMD, we found the GGA-insertion allele to be overrepresented in the lowest BMD quintile (data not shown).

The three SRP9 genotype groups differed significantly in FN BMD and LS BMD in men, but no association was found in women. At both sites, BMD was highest in the GG group and lowest in the AA group (table 3). BMD in the GA group was 1% less than that in the GG group at both sites, whereas, in the AA group, BMD was 3% less than that in the GG group at both sites (P=.02 for FN; P=.03 for LS). The effects of the SRP9 genotype on BMD at the Ward’s triangle and trochanter sites were similar (data not shown). When the BMD values were adjusted for possible confounders, BMD measurements did not change essentially for LS, but differences increased somewhat for FN. When we compared SRP9 allele frequencies between groups composed of men in the highest and lowest quintiles of BMD, we found the A allele to be overrepresented in the lowest BMD quintile (data not shown).

We repeated the analyses using reconstructed haplotypes that contained SRP3 for each subject, estimated by the PHASE program. These analyses showed the associations we observed to be driven by the individual polymorphism (SRP3) rather than by a particular haplotype (data not shown). We also tried the same analyses with haplotypes of SRP7–9, but reconstruction of the haplotypes in each individual could not be estimated reliably by PHASE.

Association between LS-BMD and SRP3 in Females Is Age Related

BMD declines with age, and genetic effects on BMD could theoretically be mediated by differences in the age-related rate of bone loss as well as by differences in peak bone density. We therefore assessed the effect of age on the relationship between the SRP3 and SRP9 genotypes and BMD. The study subjects were stratified by age into tertiles (aged 55–63, 64–71, and 72–80 years). Figure 3 illustrates that, for SRP3, the differences in LS BMD became larger with increasing age in females only. There were no differences in LS BMD between the different genotypes in the youngest age group, whereas an intermediate difference in BMD appeared in the middle age group. In the oldest age group, the difference in BMD between the DEL-GGA group and the DEL-DEL group was 4% (P=.05 [ANCOVA]), and the difference in BMD between the GGA-GGA group and DEL-DEL group was 8% (P=.004 [ANCOVA]). We found significant evidence to suggest an allele-dose effect of the SRP3 GGA-insertion allele with decreased LS BMD (P=.006). When we tested for interaction between genotype and age, the interaction term did not reach significance (P=.13).

Figure  3
LS BMD in women, stratified by tertiles of age and by SRP3 genotype. The three possible SRP3 genotypes are indicated by diamonds (DEL-DEL), squares (GGA-DEL), and triangles (GGA-GGA). P values for linear regression were adjusted for age, height, and weight. ...

No age-related association between BMD and SRP3 genotypes was found at the FN site. Similar analyses for SRP9 genotype in men found no effect of age on the BMD differences between the different genotype groups.

No Association between SRP3 and SRP9 Genotypes and Fractures

In our study population, during an average follow-up period of 8.2 years, we captured 234 incident osteoporotic fractures (164 in women; 70 in men) and 146 prevalent vertebral fractures (81 in women; 65 in men). Table 4 shows the distribution of fractures according to the SRP3 genotype in females or the SRP9 genotype in males. No significant differences were found between the genotype groups. In addition, the risk estimates, as calculated by regression analysis, did not show significant differences from unity, even when adjusted for possible confounding factors such as age, height, and weight (data not shown).

Table 4
Distribution of Fractures and ORs by Genotype in Women and Men

Since the BMD association was observed to be partly dependent on age, we also assessed the effect of age on the fracture-risk estimates. For reasons of power, we stratified the subjects into two age groups (aged 55–67 and 68–80 years). The same analyses as described above were performed, but no significant association was found (data not shown).

Interactions between SOST and VDR or COLIA1 Genotypes

We then studied possible interaction between polymorphisms in the SOST gene and two genes previously found to be associated with differences in BMD; that is, the VDR gene (12q13) and the COLIA1 gene (17q21.3). We first analyzed VDR genotypes on the basis of haplotypes constructed of three polymorphisms at the 3′ end of the VDR gene, as described elsewhere (Uitterlinden et al. 1996), in relation to SRP3 and SRP9 genotypes, but we could not find evidence to suggest interaction or additive effects on BMD or on fracture risk (data not shown).

At the COLIA1 gene, we analyzed the G→T Sp1 polymorphism for which we previously demonstrated an association with BMD and fracture risk in subjects from the Rotterdam Study (Uitterlinden et al. 1998). In the present study, we defined four groups of women, on the basis of their combined genotype, by presence or absence of risk alleles for COLIA1 and SOST. The groups were those having no risk alleles (reference group), those having at least one copy of the COLIA1 risk allele but no SRP3 risk allele (the “COLIA1” group), those having at least one copy of the SRP3 risk allele (the “SRP3” group), and those having at least one copy of either risk allele (the “COLIA1+SRP3” group). We observed a strong additive effect for these two polymorphisms. Whereas SRP3 and COLIA1 Sp1 by themselves were associated with decreased LS BMD, the combined-risk–genotype group showed an even greater effect. Because both SRP3 and COLIA1 Sp1 showed modification by age, we next analyzed the additive genotype effect on LS BMD by tertiles of age for all women (fig. 4). Our results show that the additive effect is absent in the youngest tertile, is becoming apparent in the middle tertile, but shows greatest effect (0.5 SD) and reaches significance (P=.02) in the oldest age tertile. When we repeated the analysis of additive effects on risk of fracture, no genotype-dependent effect was observed, but numbers were small in each of the genotype groups across the age groups.

Figure  4
LS BMD in women, stratified by tertiles of age and by combined genotype for SRP3 and COLIA1 Sp1 polymorphisms. The four possible genotype groups are indicated by diamonds (reference = no risk allele), squares (COLIA1), triangles (SRP3), and circles (COLIA1 ...


We and others have previously reported that loss-of-function mutations in the SOST gene region cause SCL (Balemans et al. 2001; Brunkow et al. 2001) and most likely VBD (Balemans et al. 2002b; Staehling-Hampton et al. 2002). These are monogenetic bone disorders with a severe bone phenotype that involve generalized hyperostosis. A number of recent experiments support the claim that the SOST gene product, sclerostin, is a key regulator of bone mass (Winkler et al. 2003; Van Bezooijen et al. 2004). On the basis of the observed activities of sclerostin in vitro and in vivo, we hypothesized that polymorphisms that occur in the SOST gene region, which have more subtle effects on regulation of expression and function of sclerostin, might influence differences in BMD in the general population. We therefore searched the SOST gene for frequently occurring DNA-sequence variations and evaluated whether SOST polymorphisms are associated with parameters of osteoporosis—that is, differences in BMD and in fracture risk—in a population of elderly whites.

Of 31 total variants discovered in the area surrounding the SOST gene, we chose eight common polymorphisms and one rare polymorphism for further study. The rare variant (SRP4) was the only example of a polymorphism affecting the amino acid sequence of sclerostin (G11988A/Val10Ile in the signal sequence). Although potentially of interest, its frequency was too low (<3% in whites) to allow association studies of sufficient statistical power, even in a population of the size we had available (n=1,939). Of the polymorphisms we identified, several were present in the NCBI database or had been published elsewhere (Balemans et al. 2002a). The SRPs we identified in the VBD-deletion region are new. The set of nine polymorphisms described in the present study were selected to explore ethnic diversity (SRP1–9) and then to look for association with osteoporosis in whites, as well as to analyze ethnic diversity, LD, and haplotype structure (SRP1–3 and SRP5–9).

For most polymorphisms, the frequencies in the different ethnic groups were similar (in particular, between whites and Asians) (data not shown). We observed strong LD in the immediate surroundings of the SOST gene itself but much less LD with the polymorphisms in the VBD deletion region, which comprises sequences ~35–87 kb downstream of the gene body. The existence of such LD blocks of limited size (10–30 kb) is in line with observations of other areas of the human genome (Kruglyak 1999; Dunning et al. 2000; Abecasis et al. 2001; Daly et al. 2001; Reich et al. 2001). The strong LD is also indicated by the limited number of haplotypes across the SOST gene that we could reconstruct for whites in our Rotterdam Study sample by use of the PHASE program. The number of individuals in the other ethnic groups (in the NIGMS panel) is too low to allow for accurate LD measurement and reconstruction of haplotypes. It will, however, be of interest to see if the haplotypes we identified in whites are different from those in Asians and African Americans, a comparison of which could have important consequences for performing and interpreting association studies by use of these polymorphisms and haplotypes in nonwhite groups.

In our association study of elderly whites, we analyzed eight polymorphisms that were, with the exception of SRP2, all found to be in HWE. In view of the number of HWE tests we performed, we consider this finding to be due to chance.

We observed the GGA-insertion allele (f=0.38) of the SRP3 polymorphism to be associated with decreased BMD in women (but not in men) at the FN and at the LS. The difference between extreme genotypes was 0.2 SD, with evidence of an allele-dose effect. For the SRP9 polymorphism, we found the A allele (f=0.60) to be associated with decreased BMD in men (but not in women) at the FN and at the LS. Also, the difference between extreme genotypes was 0.2 SD, with evidence of an allele-dose effect. These associations seem real, given the size of the population, their consistency (effects at several sites of BMD measurement), and the plausible genetic model of gene-dose effects observed. Although the lack of association for a given polymorphism in both sexes would argue against this, such difference between sexes is not unusual. For example, the COLIA1 Sp1 polymorphism has been found to be associated with BMD only in women and not in men (Uitterlinden et al. 1998; Mann et al. 2001). Although we acknowledge the potential difficulty with multiple comparisons in this study, several of the significance levels we observed are impressive enough to withstand conservative adjustments such as the Bonferroni correction. Together with the arguments presented above, we think the genotype-dependent differences we observe are modest but real. This situation is similar to that seen for effects of common polymorphisms on other complex traits (in addition to BMD and osteoporosis) (Lohmueller et al. 2003). Replication of our findings in other large populations, with a similar age distribution followed by meta-analysis, will help in establishing consistency and true effect size of these polymorphisms (Ioannidis et al. 2001).

When we analyzed the haplotypes across the SOST gene and the VBD deletion region, we observed that the associations were driven by the individual polymorphisms SRP3 and SRP9 rather than by a combination of allelic variants present on a haplotype allele. This could imply that there are no other important regulatory polymorphisms either in the sclerostin-promoter region or in the VBD deletion region, apart from the SRP3 and SRP9 polymorphisms, respectively.

So far, no functionality has been established for the SRP3 or the SRP9 polymorphisms. SRP3 is located ~1.4 kb upstream of the SOST transcriptional start site and thus may be important for interaction of promoter sequences with the transcriptional machinery. The insertion/deletion polymorphism falls within a short region that is highly conserved between humans and mice (80% identity over 315 bp), which suggests functional significance. In the absence of experimental data defining critical promoter sequences, we searched the TRANSFAC database (Wingender et al. 2000, 2001) for consensus binding sites for relevant transcription factors. SRP3 itself does not appear to fall within a known consensus site; however, it is closely flanked by one HOXA9- and two CBFA2-binding sites. It is also interesting to note that the conserved element falls within a region that appears to have a repressive effect on the activity of the proximal SOST promoter, although it is not clear from the published report which SRP3 allele was studied (Sevetson et al. 2004). The inclusion or exclusion of three base pairs in the chromosome could be envisaged to affect the interaction of DNA-protein complexes upstream and downstream of the region. For example, given the loss-of-function phenotype associated with SCL, one could propose that the SRP3 GGA-insertion allele enhances expression of SOST mRNA (through an effect on either positive or negative regulatory factors) and consequently the sclerostin protein, thus leading to decreased BMD. Molecular analysis of the SOST promoter region and complete characterization of important regulatory elements is necessary to establish the influence of these subtle polymorphic variants.

Interestingly, Balemans et al. (2002a) also analyzed the SRP3 polymorphism (termed “SNP5,” which carries a TCC-insertion allele) but did not find any association with BMD in a population of 619 white Scottish perimenopausal women. We believe this can be explained by the age dependency that we observed for the association of SRP3 with BMD in women. In our group of women aged 55–63 years, which is close in age range to those in the Balemans et al. study, we failed to see an association with BMD. Only when we examined women aged [gt-or-equal, slanted]63 years did the association become apparent, and we observed the genotype effect to further increase with age. The SRP3 genotype effect, therefore, seems to be restricted to older age categories, probably due to modification by age-related factors. Such factors could encompass gene-gene interactions and gene-environment interactions, such as changes in diet, changes in hormonal factors, changes in mechanical loading, etc. In this respect, it is of interest to note that sclerostin is specifically and highly expressed in the osteocyte (Winkler et al. 2003; Van Bezooijen et al 2004), a bone cell that has been implicated in the mechanosensitivity of bone tissue. The evident decrease of mobility with age could therefore play an important role in the age-related increase in the genotype effect we observed. However, this obviously needs further study.

The G→A SRP9 polymorphism is located in the 52-kb VBD deletion region, an area that, when deleted, leads to a phenotype very similar to, albeit less severe than, that seen in SCL (Balemans et al. 2002b; Staehling-Hampton et al. 2002). Since there is no evidence of any functional genes within the region, it has been postulated that the VBD-deletion region contains important regulatory sites for the expression of sclerostin (Staehling-Hampton et al. 2002). This notion is supported by recent studies in transgenic mice (G. Loots, personal communication), although the exact mechanism by which SOST gene expression is regulated remains to be elucidated. Within the deletion area, several regions conserved between humans and mice have been identified, but SRP9 is not in any of these (B.W.P., S.P., and M.E.B., unpublished data). Therefore, the functional relevance of the SRP9 polymorphism for expression of sclerostin and how it might influence differences in BMD later in life needs further study.

In view of the importance of sclerostin in regulation of BMD and the previous observations of other gene variants that affect BMD, we studied combined effects of such gene variants. In particular, we analyzed sclerostin polymorphisms in relation to the COLIA1 Sp1 polymorphism and the 3′ BsmI, ApaI, and TaqI VDR polymorphisms. Although we could not see a combined effect of SOST with VDR polymorphisms, we did see an additive effect of the SOST SRP3 and the COLIA1 Sp1 polymorphisms. In addition, we observed that the combined genotype effect increases with age. As discussed above, SRP3 is likely to affect expression of sclerostin, whereas the COLIA1 Sp1 polymorphism has been demonstrated to result in increased expression of the COLIA1 protein, which results in a slight excess of COLIA1 homotrimers (Mann et al. 2001). Such homotrimers have been associated with decreased bone quality, including decreased BMD (Mann et al. 2001). Although the molecular mechanism of the additive effect of SOST and COLIA1 has not been established, variations in sclerostin expression are likely to interact with variations in COLIA1 expression. As a BMP antagonist, sclerostin is a key regulator of bone formation, a process that involves production of collagen and subsequent mineralization. Therefore, the production of collagen (including homotrimers) is likely to be under the influence of sclerostin regulation.

It has been well established that a decreased BMD results in increased fracture risk, and several studies have indicated that a 1-SD decrease in FN BMD is associated with 1.5–2.5-fold increase in fracture risk (Schuit et al. 2004). Although we observed genotype effects on BMD, we did not observe any effect of SRP3 or SRP9 on fracture risk. We hypothesize that this lack of association of the SOST polymorphisms with fracture risk is explained by the modest effect size of these polymorphisms (0.2-SD difference between extreme genotypes) and the fact that the recognizable effect of the SOST polymorphisms is probably restricted to BMD differences. Fracture risk is only partly explained by differences in BMD, which means that other factors, such as bone quality and bone geometry, are at least as important in determination of fracture risk. We analyzed a relatively large population-based study in which 234 incident osteoporotic fractures occurred and 146 prevalent vertebral fractures were diagnosed. This means that the methods set forth in our study have substantial power to detect an increase in fracture risk, given the SRP3 allele frequency of 38%. Yet, in view of the small effect size of SRP3 on BMD, we will need an even larger sample size than the one we had available (n=1,939) to detect such effects on fracture risk.

In conclusion, two variants in the SOST gene region were associated with BMD differences and were found to be modified by sex (SRP3 and SRP9) and by age (SRP3). The effect size we observed on BMD was too small to be detected as genotype-dependent differences in fracture risk in our population study of 1,939 subjects. The mechanism that underlies the BMD differences remain to be established but is likely to involve modulation of expression of sclerostin. Further studies in additional populations of elderly men and postmenopausal women will be helpful in establishing consistency of our findings and in estimating the true effect size of these regulatory sclerostin polymorphisms.


The authors gratefully acknowledge Lynn Jorde, for helpful comments and discussion; David Galas and Jeff Van Ness, for helpful suggestions during the early phase of the study; and Gabriela Loots, for sharing unpublished results from rodent studies. In addition, we thank the participants of the Rotterdam Study and the many field workers at the research center in Ommoord, The Netherlands.

Electronic-Database Information

Accession numbers and URLs for data presented herein are as follows:

GenBank, http://www.ncbi.nih.gov/Genbank/ (for SOST [accession numbers AF326736, AF397423, and AF326739])
HaploXT, http://archimedes.well.ox.ac.uk/pise/ (for LD analysis)
Online Mendelian Inheritance in Man (OMIM), http://www.ncbi.nlm.nih.gov/Omim/ (for SCL and VBD)


Abecasis GR, Noguchi E, Heinzmann A, Traherne JA, Bhattacharyya S, Leaves NI, Anderson GG, Zhang Y, Lench NJ, Carey A, Cardon LR, Moffatt MF, Cookson WOC (2001) Extent and distribution of linkage disequilibrium in three genomic regions. Am J Hum Genet 68:191–197 [PMC free article] [PubMed]
Balemans W, Ebeling M, Patel N, Van Hul E, Olson P, Dioszegi M, Lacza C, Wuyts W, Van Den Ende J, Willems P, Paes-Alves A, Hill S, Bueno M, Ramos F, Tacconi P, Dikkers F, Stratakis C, Lindpaintner K, Vickery B, Foernzler D, Van Hul W (2001) Increased bone density in sclerosteosis is due to the deficiency of a novel secreted protein (SOST). Hum Mol Genet 10:537–543 [PubMed] [Cross Ref]10.1093/hmg/10.5.537
Balemans W, Foernzler D, Parsons C, Ebeling M, Thompson A, Reid DM, Lindpaintner K, Ralston SH, Van Hul W (2002a) Lack of association between the SOST gene and bone mineral density in perimenopausal women: analysis of five polymorphisms. Bone 31:515–519 [PubMed] [Cross Ref]10.1016/S8756-3282(02)00844-X
Balemans W, Patel N, Ebeling M, Van Hul E, Wuyts W, Lacza C, Dioszegi M, Dikkers FG, Hildering P, Willems PJ, Verheij JB, Lindpaintner K, Vickery B, Foernzler D, Van Hul W (2002b) Identification of a 52 kb deletion downstream of the SOST gene in patients with van Buchem disease. J Med Genet 39:91–97 [PMC free article] [PubMed] [Cross Ref]10.1136/jmg.39.2.91
Becherini L, Gennari L, Masi L, Mansani R, Massart F, Morelli A, Falchetti A, Gonnelli S, Fiorelli G, Tanini A, Brandi ML (2000) Evidence of a linkage disequilibrium between polymorphisms in the human estrogen receptor α gene and their relationship to bone mass variation in postmenopausal Italian women. Hum Mol Genet 9:2043–2050 [PubMed] [Cross Ref]10.1093/hmg/9.13.2043
Beighton P (1988) Sclerosteosis. J Med Genet 25:200–203 [PMC free article] [PubMed]
Beighton P, Barnard A, Hamersma H, van der Wouden A (1984) The syndromic status of sclerosteosis and van Buchem disease. Clin Genet 25:175–181 [PubMed]
Brunkow ME, Gardner JC, Van Ness J, Paeper BW, Kovacevich BR, Proll S, Skonier JE, Zhao L, Sabo PJ, Fu Y-H, Alisch RS, Gillett L, Colbert T, Tacconi P, Galas D, Hamersma H, Beighton P, Mulligan J (2001) Bone dysplasia sclerosteosis results from loss of the SOST gene product, a novel cystine knot–containing protein. Am J Hum Genet 68:577–589 [PMC free article] [PubMed]
Burger H, Van Daele PL, Algra D, van den Ouweland FA, Grobbee DE, Hofman A, van Kuijk C, Schutte HE, Birkenhager JC, Pols HA (1994) The association between age and bone mineral density in men and women aged 55 years and over: the Rotterdam Study. Bone Miner 25:1–13 [PubMed]
Cummings SR, Nevitt MC, Browner WS, Stone K, Fox KM, Ensrud KE, Cauley J, Black D, Vogt TM, Study of Osteoporotic Fractures Research Group (1995) Risk factors for hip fracture in white women. N Engl J Med 332:767–773 [PubMed] [Cross Ref]10.1056/NEJM199503233321202
Daly MJ, Rioux JD, Schaffner SF, Hudson TJ, Lander ES (2001) High-resolution haplotype structure in the human genome. Nat Genet 29:229–232 [PubMed] [Cross Ref]10.1038/ng1001-229
Devoto M, Shimoya K, Caminis J, Ott J, Tenenhouse A, Whyte MP, Sereda L, Hall S, Considine E, Williams CJ, Tromp G, Kuivaniemi H, Ala-Kokko L, Prockop DJ, Spotila LD (1998) First-stage autosomal genome screen in extended pedigrees suggests genes predisposing to low bone mineral density on chromosomes 1p, 2p and 4q. Eur J Hum Genet 6:151–157 [PubMed] [Cross Ref]10.1038/sj.ejhg.5200169
Devoto M, Specchia C, Li HH, Caminis J, Tenenhouse A, Rodriguez H, Spotila LD (2001) Variance component linkage analysis indicates a QTL for femoral neck bone mineral density on chromosome 1p36. Hum Mol Genet 10:2447–2452 [PubMed] [Cross Ref]10.1093/hmg/10.21.2447
Duncan EL, Brown MA, Sinsheimer J, Bell J, Carr AJ, Wordsworth BP, Wass JA (1999) Suggestive linkage of the parathyroid receptor type 1 to osteoporosis. J Bone Miner Res 14:1993–1999 [PubMed]
Dunning AM, Durocher F, Healey CS, Teare MD, McBride SE, Carlomagno F, Xu CF, Dawson E, Rhodes S, Ueda S, Lai E, Luben RN, Van Rensburg EJ, Mannermaa A, Kataja V, Rennart G, Dunham I, Purvis I, Easton D, Ponder BAJ (2000) The extent of linkage disequilibrium in four populations with distinct demographic histories. Am J Hum Genet 67:1544–1554 [PMC free article] [PubMed]
Efstathiadou Z, Tsatsoulis A, Ioannidis JP (2001) Association of collagen Ialpha 1 Sp1 polymorphism with the risk of prevalent fractures: a meta-analysis. J Bone Miner Res 16:1586–1592 [PubMed]
Epstein S, Hamersma H, Beighton P (1979) Endocrine function in sclerosteosis. S Afr Med J 55:1105–1110 [PubMed]
Flicker L, Faulkner KG, Hopper JL, Green RM, Kaymacki B, Nowson CA, Young D, Wark JD (1996) Determinants of hip axis length in women aged 10–89 years: a twin study. Bone 18:41–45 [PubMed] [Cross Ref]10.1016/8756-3282(95)00418-1
Flicker L, Hopper JL, Rodgers L, Kaymakci B, Green RM, Wark JD (1995) Bone density determinants in elderly women: a twin study. J Bone Miner Res 10:1607–1613 [PubMed]
Grant SF, Reid DM, Blake G, Herd R, Fogelman I, Ralston SH (1996) Reduced bone density and osteoporosis associated with a polymorphic Sp1 binding site in the collagen type I α 1 gene. Nat Genet 14:203–205 [PubMed] [Cross Ref]10.1038/ng1096-203
Hansen H (1967) Sklerosteose. In: Opitz H, Schmid F (eds) Handbuch der Kinderheilkunde. Vol 6. Springer-Verlag, Berlin, pp 351–355
Hofman A, Grobbee DE, de Jong PT, van den Ouweland FA (1991) Determinants of disease and disability in the elderly: the Rotterdam Elderly Study. Eur J Epidemiol 7:403–422 [PubMed]
Hsu DR, Economides AN, Wang X, Eimon PM, Harland RM (1998) The Xenopus dorsalizing factor Gremlin identifies a novel family of secreted proteins that antagonize BMP activities. Mol Cell 1:673–683 [PubMed] [Cross Ref]10.1016/S1097-2765(00)80067-2
Ioannidis JP, Ntzani EE, Trikalinos TA, Contopoulos-Ioannidis DG (2001) Replication validity of genetic association studies. Nat Genet 29:306–309 [PubMed] [Cross Ref]10.1038/ng749
Kobayashi S, Inoue S, Hosoi T, Ouchi Y, Shiraki M, Orimo H (1996) Association of bone mineral density with polymorphism of the estrogen receptor gene. J Bone Miner Res 11:306–311 [PubMed]
Koller DL, Econs MJ, Morin PA, Christian JC, Hui SL, Parry P, Curran ME, Rodriguez LA, Conneally PM, Joslyn G, Peacock M, Johnston CC, Foroud T (2000) Genome screen for QTLs contributing to normal variation in bone mineral density and osteoporosis. J Clin Endocrinol Metab 85:3116–3120 [PubMed] [Cross Ref]10.1210/jc.85.9.3116
Kruglyak L (1999) Prospects for whole-genome linkage disequilibrium mapping of common disease genes. Nat Genet 22:139–144 [PubMed] [Cross Ref]10.1038/9642
Lohmueller KE, Pearce CL, Pike M, Lander ES, Hirschhorn JN (2003) Meta-analysis of genetic association studies supports a contribution of common variants to susceptibility to common disease. Nat Genet 33:177–182 [PubMed] [Cross Ref]10.1038/ng1071
Mann V, Hobson EE, Li B, Stewart TL, Grant SF, Robins SP, Aspden RM, Ralston SH (2001) A COL1A1 Sp1 binding site polymorphism predisposes to osteoporotic fracture by affecting bone density and quality. J Clin Invest 107:899–907 [PMC free article] [PubMed]
McCloskey EV, Spector TD, Eyres KS, Fern ED, O’Rourke N, Vasikaran S, Kanis JA (1993) The assessment of vertebral deformity: a method for use in population studies and clinical trials. Osteoporo Int 3:138–147 [PubMed]
Morrison NA, Qi JC, Tokita A, Kelly PJ, Crofts L, Nguyen TV, Sambrook PN, Eisman JA (1994) Prediction of bone density from vitamin D receptor alleles. Nature 367:284–287 [PubMed] [Cross Ref]10.1038/367284a0
Pearce J, Penny G, Rossant J (1999) A mouse cerberus/Dan-related gene family. Dev Biol 209:98–110 [PubMed] [Cross Ref]10.1006/dbio.1999.9240
Pocock NA, Eisman JA, Hopper JL, Yeates MG, Sambrook PN, Eberl S (1987) Genetic determinants of bone mass in adults: a twin study. J Clin Invest 80:706–710 [PMC free article] [PubMed]
Ralston SH (2002) Genetic control of susceptibility to osteoporosis. J Clin Endocrinol Metab 87:2460–2466 [PubMed] [Cross Ref]10.1210/jc.87.6.2460
Reich DE, Cargill M, Bolk S, Ireland J, Sabeti PC, Richter DJ, Lavery T, Kouyoumjian R, Farhadian SF, Ward R, Lander ES (2001) Linkage disequilibrium in the human genome. Nature 411:199–204 [PubMed] [Cross Ref]10.1038/35075590
Schuit SC, Van der Klift M, Weel AE, De Laet CE, Burger H, Seeman E, Hofman A, Uitterlinden AG, van Leeuwen JP, Pols HA (2004) Fracture incidence and association with bone mineral density in elderly men and women: the Rotterdam Study. Bone 34:195–202 [PubMed] [Cross Ref]10.1016/j.bone.2003.10.001
Sevetson B, Taylor S, Pan Y (2004) Cbfa1/RUNX2 directs specific expression of the sclerosteosis gene (SOST). J Biol Chem 279:13849–13858 [PubMed] [Cross Ref]10.1074/jbc.M306249200
Smith DM, Nance WE, Kang KW, Christian JC, Johnston CC Jr (1973) Genetic factors in determining bone mass. J Clin Invest 52:2800–2808 [PMC free article] [PubMed]
Staehling-Hampton K, Proll S, Paeper BW, Zhao L, Charmley P, Brown A, Gardner JC, Galas D, Schatzman RC, Beighton P, Papapoulos S, Hamersma H, Brunkow ME (2002) A 52-kb deletion in the SOST-MEOX1 intergenic region on 17q12-q21 is associated with van Buchem disease in the Dutch population. Am J Med Genet 110:144–152 [PubMed] [Cross Ref]10.1002/ajmg.10401
Stein SA, Witkop C, Hill S, Fallon MD, Viernstein L, Gucer G, McKeever P, Long D, Altman J, Miller NR, Teitelbaum SL, Schlesinger S (1983) Sclerosteosis: neurogenetic and pathophysiologic analysis of an American kinship. Neurology 33:267–277 [PubMed]
Stephens M, Smith NJ, Donnelly P (2001) A new statistical method for haplotype reconstruction from population data. Am J Hum Genet 68:978–989 [PMC free article] [PubMed]
Uitterlinden AG, Burger H, Huang Q, Yue F, McGuigan FE, Grant SF, Hofman A, van Leeuwen JP, Pols HA, Ralston SH (1998) Relation of alleles of the collagen type Iα1 gene to bone density and the risk of osteoporotic fractures in postmenopausal women. N Engl J Med 338:1016–1021 [PubMed] [Cross Ref]10.1056/NEJM199804093381502
Uitterlinden AG, Pols HA, Burger H, Huang Q, Van Daele PL, Van Duijn CM, Hofman A, Birkenhager JC, van Leeuwen JP (1996) A large-scale population-based study of the association of vitamin D receptor gene polymorphisms with bone mineral density. J Bone Miner Res 11:1241–1248 [PubMed]
Uitterlinden AG, van Leeuwen JPTM, Pols HAP (2001) Genetics and genomics of osteoporosis. In: Marcus R, Feldman D, Kelsey J (eds) Osteoporosis. Academic Press, San Diego, pp 639–667
Van Bezooijen RL, Roelen BA, Visser A, Wee-Pals L, De Wilt E, Karperien M, Hamersma H, Papapoulos SE, Ten Dijke P, Lowik CW (2004) Sclerostin is an osteocyte-expressed negative regulator of bone formation, but not a classical BMP antagonist. J Exp Med 199:805–814 [PMC free article] [PubMed] [Cross Ref]10.1084/jem.20031454
van Buchem F, Hadders H, Ubbens R (1955) An uncommon familial systemic disease of the skeleton: hyperostosis corticalis generalisata familiaris. Acta Radiol 44:109–119 [PubMed]
van Buchem F, Prick J, Jaspar H (1976) Hyperostosis corticalis generalisata familiaris (van Buchem disease). Excerpta Medica, Amsterdam
Van der Klift M, De Laet CE, McCloskey EV, Hofman A, Pols HA (2002) The incidence of vertebral fractures in men and women: the Rotterdam Study. J Bone Miner Res 17:1051–1056 [PubMed]
van Meurs JB, Schuit SC, Weel AE, Van der Klift M, Bergink AP, Arp PP, Colin EM, Fang Y, Hofman A, Van Duijn CM, Van Leeuwen JP, Pols HA, Uitterlinden AG (2003) Association of 5′ estrogen receptor alpha gene polymorphisms with bone mineral density, vertebral bone area and fracture risk. Hum Mol Genet 12:1745–1754 [PubMed] [Cross Ref]10.1093/hmg/ddg176
Wingender E, Chen X, Fricke E, Geffers R, Hehl R, Liebich I, Krull M, Matys V, Michael H, Ohnhauser R, Pruss M, Schacherer F, Thiele S, Urbach S (2001) The TRANSFAC system on gene expression regulation. Nucleic Acids Res 29:281–283 [PMC free article] [PubMed] [Cross Ref]10.1093/nar/29.1.281
Wingender E, Chen X, Hehl R, Karas H, Liebich I, Matys V, Meinhardt T, Pruss M, Reuter I, Schacherer F (2000) TRANSFAC: an integrated system for gene expression regulation. Nucleic Acids Res 28:316–319 [PMC free article] [PubMed] [Cross Ref]10.1093/nar/28.1.316
Winkler DG, Sutherland MK, Geoghegan JC, Yu C, Hayes T, Skonier JE, Shpektor D, Jonas M, Kovacevich BR, Staehling-Hampton K, Appleby M, Brunkow ME, Latham JA (2003) Osteocyte control of bone formation via sclerostin, a novel BMP antagonist. EMBO J 22:6267–6276 [PMC free article] [PubMed] [Cross Ref]10.1093/emboj/cdg599

Articles from American Journal of Human Genetics are provided here courtesy of American Society of Human Genetics
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Cited in Books
    Cited in Books
    PubMed Central articles cited in books
  • Gene
    Gene links
  • Gene (nucleotide)
    Gene (nucleotide)
    Records in Gene identified from shared sequence links
  • GEO Profiles
    GEO Profiles
    Related GEO records
  • HomoloGene
    HomoloGene links
  • MedGen
    Related information in MedGen
  • Nucleotide
    Published Nucleotide sequences
  • OMIM
    OMIM record citing PubMed
  • PubMed
    PubMed citations for these articles

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...