Bone 37 (2005) 770 – 775 www.elsevier.com/locate/bone
LRP5 gene polymorphisms and idiopathic osteoporosis in men S.L. Ferrari a , S. Deutsch b , C. Baudoin c , M. Cohen-Solal c , A. Ostertag c , S.E. Antonarakis b , R. Rizzoli a , M.C. de Vernejoul c,⁎ a
Service of Bone Diseases, Department of Rehabilitation and Geriatrics, Geneva University Hospital, Geneva, Switzerland b Department of Genetic Medicine and Development, Geneva University Medical School, Geneva, Switzerland c INSERM U606 and Fédération de rhumatologie, Hôpital Lariboisière, 2 rue Ambroise Paré, 75010 Paris, France Received 4 May 2005; revised 29 June 2005; accepted 30 June 2005 Available online 15 September 2005
Abstract Mutations in the low-density lipoprotein receptor-related protein 5 gene (LRP5) have demonstrated the role of LRP5 in bone mass acquisition. LRP5 variants were recently reported to contribute to the population-based variance in vertebral bone mass and size in males. To investigate whether LRP5 variants are implicated in idiopathic male osteoporosis, we studied 78 men with low BMD (b2.5 T score or b −2 Z score) aged less than 70 years (mean ± SD: 50 ± 16 years) in whom secondary causes of osteoporosis had been excluded and 86 controls (51 ± 10 years). Genotypes and haplotypes were based on LRP5 missense substitutions in exons 9 (c.2047G N A, p.V667M) and 18 (c.4037C N T, p.A1330V), and their association with osteoporosis evaluated after adjustment for multiple clinical and environmental variables using logistic regression. The presence of osteoporosis was significantly associated with LRP5 haplotypes (P = 0.0036) independent of age (P = 0.006), weight (P = 0.004), calcium intake (P = 0.002), alcohol (P = 0.005) and tobacco (P = 0.004) consumption. Accordingly, the odds ratio for osteoporosis was 3.78 (95% CI 1.27–11.26, P b 0.001) in male carriers of haplotype 3 (c.2047A–4037T, n = 20 cases and 12 controls) versus homozygous carriers of haplotype 1 (c.2047G–4037C, n = 42 cases and 61 controls). In conclusion, these data indicate beyond a significant role for environmental factors, an association between LRP5 variants and idiopathic osteoporosis in males, pointing to a role of LRP5 in this disease. © 2005 Elsevier Inc. All rights reserved. Keywords: Osteoporosis; Men; LRP5; Polymorphism
Introduction Idiopathic male osteoporosis has been defined as low bone mass occurring in men aged less than 70 years, when all the potential secondary causes of osteoporosis have been ruled out [1,2]. Osteoporosis in these patients frequently manifests first as a clinical fracture, likely because preventive screening for bone mineral density is rarely performed in men. Markers of bone turnover are usually normal in these cases [2], although bone histomorphometry indicates decreased bone formation [3,4]. Environmental Abbreviation: BMD, Bone mineral density. ⁎ Corresponding author. Fax: +33 1 49 95 84 52. E-mail address:
[email protected] (M.C. de Vernejoul). 8756-3282/$ - see front matter © 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2005.06.017
factors known to influence bone remodeling in women might play a role in idiopathic male osteoporosis. However, we and others previously reported a high prevalence of low bone mass in first degree relatives of patients with idiopathic osteoporosis, pointing towards a genetic predisposition to poor bone mass acquisition [2,5,6]. The genes involved in this complex disorder, however, remain completely unknown. The low-density lipoprotein receptor-related protein 5 (LRP5) gene, which in human maps to chromosome 11q12– 13, is a candidate susceptibility factor for osteoporosis. Loss of function mutations result in a rare autosomal recessive disorder characterized by low bone mass [7], whereas gain of function mutations are associated with high bone mass [8,9] and several osteosclerotic disorders [10]. Transgenic and knock-out mouse models in turn recapitulate the human
S.L. Ferrari et al. / Bone 37 (2005) 770–775
phenotype of high and low bone mass, respectively, [11,12] and have shown the major role of LRP5 on bone formation [11]. More recently, we reported that LRP5 polymorphisms, particularly missense substitutions in exon 9 c.2047G N A (p.V667M) and 18 c.4037C N T (p.A1330V) and their related haplotypes, are associated with bone mass at the lumbar spine in men but not women [13]. Hence, LRP5 polymorphisms contribute significantly to the variance in vertebral bone mineral content and projected area and, to a lesser extent, bone mineral density, in the normal male population. In addition, bone mass acquisition at the lumbar spine in male children with haplotype c.2047A– 4037T was significantly decreased compared to the other haplotypes, suggesting that the deficit in peak bone mass observed in the adults could be generated during growth [13]. These observations prompted us to investigate LRP5 polymorphisms in a cohort of men with idiopathic osteoporosis.
Methods Patients and controls Some male patients with idiopathic osteoporosis included in this study have previously been reported in details [5,6]. Briefly, among men referred to the clinic for management of osteoporosis, we selected 78 cases who fulfilled the following criteria: (1) secondary causes of osteoporosis such as malignant disease, malabsorption, vitamin D deficiency, hypogonadism, hemochromatosis, hyperthyroidism, endogenous or exogenous hypercortisolism were excluded by extensive clinical, radiological and biochemical investigations including plasma levels of free testosterone, thyroid stimulating hormone (TSH) and parathyroid hormone (PTH); (2) their T score was below −2.5 or their Z score was below −2 at the lumbar spine or femoral neck, as previously defined [2,6]; (3) subjects were born in France from European–Caucasian ancestry and were aged less than 70 years; (4) they were not members of the same family. These cases were compared to 86 controls of same ethnicity who were employees at our hospital and aged 20 to 70 years. As for cases, we excluded subjects with a disease or treatment that could affect bone. In addition, their Z and T scores at the lumbar spine and femoral neck had to be higher than −2 and −2.5, respectively. Patients and controls filled identical self-administered semi-quantitative questionnaires reporting dietary calcium intake, alcohol consumption and smoking (6). Bone mineral density (BMD) was measured at the femoral neck and lumbar spine (L2–L4) using a Lunar DPX-L (Lunar Corporation, Madison, Wisconsin). All measurements were performed by a single technician using the same densitometer. Age and gender-adjusted BMD values were based
771
on a French reference population between 20 and 89 years of age from several centers (provided by Lunar France) and expressed as Z and T scores. All subjects were requested to self-report their history of clinical vertebral and nonvertebral fractures. In addition, a spine X-ray was performed in all cases and in controls aged more than 50 years, and vertebral crush fractures were defined as a decrease in vertebral height greater than 25%. All participants gave written informed consent to be included in the study which was approved by the ethical committee of the hospital. Genotyping Genomic DNA was extracted from blood lymphocytes by the phenol chloroform method. SNPs in exon 9, c.2047G N A [accession number rs4988321], and exon 18, c.4037C N T [accession number rs3736228], were determined by the Pyrosequencing method (Pyrosequencing, Uppsala Sweden) as previously described [13]. Specific primers were as follows: Ex9 c.2047G N A, Fwd 5′-GCTTGGCCGCACCCCTTT, Rwd 5′-biotin-GCCTCCTTGACGCCCGTGAG and sequencing 5′-CGAGACCAATAACAACGA; exon 18 c.4037C N T, Fwd 5′-TGCTCCCCGGACCAGTT, Rwd 5′-biotin-GGAGGGCCTCACCGTCACA and sequencing 5′-GGACCGCTCAGACGA. Haplotypes were inferred using the HAPLOTYPER program and default parameters [14]. Accordingly, 3 haplotypes based on exons 9 and 18 SNPs were found in this cohort: H1 (G–C), H2 (G–T) and H3 (A–T). Subjects were subsequently grouped according to their haplotype combinations as follows: homozygous carriers of haplotype 1 (H1/H1), carriers of haplotype 2 without or with haplotype 1 (H2/H2,1) and carriers of haplotype 3 without or with haplotype 1 or 2 (H3/H3,2,1). Statistical methods The results were expressed as mean ± SD. Patients and control values were compared by Student's t test or Chisquare test. We related the probability of developing osteoporosis to the haplotype groups and to six clinical characteristics and environment variables: age, weight, height and calcium intake, included as quantitative variables, alcohol and tobacco consumption, analyzed as binary variables (actual presence or absence). The model included the seven factors and their two by two combined effects (first-order interaction terms). The relationship was estimated using multiple logistic regression performed with GLM function [15]. Odd ratios were derived from the coefficient of this regression. To select the best-fitting and most parsimonious subsets of predictor variables, we used the lack-of-fit technique. Hypotheses testing was assessed by comparing the maximum likelihood observed in two nested models, using the likelihood ratio test [16]. Tests for detecting non-normality and outliers were performed.
772
S.L. Ferrari et al. / Bone 37 (2005) 770–775
Table 1 Clinical data and BMD in controls and cases
Age (years) Weight (kg) Height (cm) Height ⁎ (cm) BMI (kg/m2) Lumbar spine BMD (g/cm2) Z score Femoral neck BMD (g/cm2) Z score Fractures Vertebral Ribs All Calcium intake (mg/day) Alcohol (number) 0 g/day 0–20 g/day 20 g/day Alcohol (g/day) Current smokers (%) Number of cigarettes per day
Controls (n = 86)
Cases (n = 78)
P value
49.8 (16.1) 75.2 (8.7) 174.7 (6.2) 174.7 (6.2) 24.6 (2.5)
50.7 (9.9) 68.9 (10.6) 171.9 (7.4) 173.4 (6.7) 23.4 (3.8)
0.64 P b 0.0001 0.010 0.30 0.012
1.200 (0.150) 0.151 (1.241)
0.835 (0.114) −2.868 (0.851)
P b 0.0001 P b 0.0001
0.967 (0.127) 0.315 (0.836)
0.764 (0.111) −1.222 (0.878)
P b 0.0001 P b 0.0001
0 1 24 909 (350)
38 13 60 752 (364)
P b 0.0001 P b 0.001 P b 0.0001 0.007 0.73
23 35 21 14.8 (16.3) 25 9 (12)
17 31 22 20.8 (21.5) 36 12 (13)
0.057 0.15 0.064
Data are presented by mean (SD), P value tests the difference between controls and cases assessed by ANOVA. ⁎ Height excluding patients with vertebral fractures.
Statistical analyses were performed using S-PLUS 6.1 statistical package for Windows [17]. Probability values were two-tailed.
Results Characteristics and environmental variables in patients and controls Mean age in the whole cohort was 50 ± 13 years, and cases and controls were matched for age. Patients had significantly lower weight and height than controls, but differences in height disappeared when patients with vertebral crush fractures (n = 38) were excluded from
analysis (Table 1). By definition, BMD was significantly lower among osteoporotic patients, although their Z scores were markedly lower at the lumbar spine than at the femoral neck. Seventy-seven percent of patients versus 25% of controls had experienced at least one fracture. Among those, a vertebral crush fracture was diagnosed in half of the patients but none of the controls. Rib fractures also occurred more frequently in the patients. Table 1 shows also that calcium intake was significantly lower in the patients, whereas their consumption of alcohol and tobacco was marginally higher than controls. Distribution of LRP5 polymorphisms among cases and controls In the whole cohort, genotype frequencies were as follows: exon 9 GG (80%), GA (19%) and AA (1%), exon 18 CC (63%), CT (34%) and TT (3%), both of which are in Hardy–Weinberg equilibrium. The common CC genotype of exon 18 was significantly less frequent among cases (P = 0.036), whereas the GG genotype of exon 9 was marginally less frequent among cases than controls (P = 0.059), when compared to carriers of the rare exon 18T and exon 9A alleles, respectively (Table 2A). Accordingly, homozygous carriers of the H1 (G–C) haplotype were also marginally under-represented among cases compared to carriers of H2 (G–T) and H3 (A–T) (P = 0.066) (Table 2B). We further evaluated the association of LRP5 haplotypes with osteoporosis after adjusting for age, body size and environmental variables using multiple logistic regression. In this case, 14 subjects were excluded (n = 7 cases and 8 controls) because at least one information was missing (mainly for calcium and alcohol intake), leaving 150 subjects in the three pre-defined haplotypic groups (H1/H1, n = 97, H2/H2,1, n = 25 and H3/H3,2,1, n = 28). According to this model, the risk of osteoporosis increased significantly with LRP5 haplotypes H1 to H3 (P = 0.0036) and independently increased with most other environmental variables, except calcium intake that decreased risk and height that was not associated with risk (Table 3). Moreover, significant interactions were found between age, weight and smoking, but no interaction occurred between the haplotypes and any other variable (data not shown).
Table 2A Distribution of LRP5 polymorphisms in cases and controls Alleles
Genotype (3 classes)
Exon 9 Cases Controls
(n = 78) (n = 86)
Exon 18 Cases Controls
(n = 78) (n = 86)
Genotype (2 classes)
G
A
P value
GG
GA
AA
P value
GG
GA/AA
136 (87%) 159 (92%)
20 (13%) 13 (I8%)
0.114
58 (74%) 74 (86%)
20 (26%) 11 (13%)
0 (0%) 1 (1%)
0.075
58 (74%) 74 (86%)
20 (26%) 12 (14%)
C
T
P value
CC
TC
TT
P value
CC
TC/TT
119 (76%) 145 (84%)
37 (24%) 27 (16%)
0.067
43 (55%) 61 (71%)
33 (42%) 23 (27%)
2 (3%) 2 (2%)
0.104
43 (55%) 61 (71%)
35 (45%) 25 (29%)
The P value tests the difference between cases and controls assessed by a Chi-square test (qualitative data).
S.L. Ferrari et al. / Bone 37 (2005) 770–775
773
Table 2B Distribution of LRP5 polymorphisms in cases and controls Haplotype combinations a
Haplotypes
Cases Controls
(n = 78) (n = 86)
H1
H2
H3
P value
H1/H1
H2/H2,1
H3/H3,2,1
P value
120 (77%) 146 (85%)
16 (10%) 13 (7,5%)
20 (13%) 13 (7,5%)
0.168
42 (54%) 61 (71%)
16 (20%) 13 (15%)
20 (26%) 12 (14%)
0.066
The P value tests the difference between cases and controls assessed by a Chi-square test (qualitative data). Haplotypes: H1, G–C, H2, G–T, H3, A–T. a Subjects were grouped according to their haplotypes combinations as follows: homozygous carriers of H1, carriers of H2 with H2 or H1, carriers of H3 with any other haplotype.
Estimate of the odd ratios of osteoporosis for LRP5 polymorphisms After adjustment on multiple variables using the multiple logistic regression described above, the odds ratio of osteoporosis for exon 9 GA/AA genotypic group was 2.98 (95% CI: 1.03–8.61, P = 0.022) and for exon 18 CT/TT genotypic group 3.27 (95% CI: 1.43–7.44, P = 0.002). By considering the haplotype combination H1/H1 as the reference, the odds ratio of osteoporosis for H2/H2,1 was 2.87 (95% CI 1.03, 8.01, P = 0.02) and for H3/H3,2,1 3.78 (95% CI 1.27–11.26, P b 0.001) (Fig. 1).
Discussion In this cohort of 78 men with idiopathic osteoporosis, we found that LRP5 genetic variation is associated to the risk of osteoporosis. Thus, haplotype 3 (c.2047A–4037T, p.667M– 1330V) was associated with an odd ratio as high as 3.8 for this disease. As expected, we also found that other interrelated variables, including low body weight, poor calcium intake and smoking, were linked to the occurrence of osteoporosis in these men, but no evidence for interaction with LRP5 alleles was observed. Thus, LRP5 polymorphisms appear to be genetic susceptibility factors for idiopathic osteoporosis in men independent of common environmental risk factors for osteoporosis. Idiopathic male osteoporosis is a rare and unique condition affecting predominantly cancellous bone, as shown again in this study where Z scores were markedly lower at the lumbar spine than the femoral neck. Indeed, pathogenic mechanisms in this disease might be different,
and probably more homogeneous, than those responsible for age-related bone loss. Several observations suggest that genetic factors influencing peak bone mass may play a predominant role in idiopathic male osteoporosis. First, low bone mass is highly prevalent among first degree relatives of osteoporotic males, independently of environmental factors [5,6]. Second, there is a predominantly male-tomale transmission of low bone mass in this disease [2]. Third, biopsies from patients with idiopathic male osteoporosis indicate a defect in bone formation [3,4]. In this regard, it is interesting that LRP5 mediates the effects of Wnt on bone formation and bone mass acquisition [7,11]. Hence, our findings of an association between LRP5 alleles and idiopathic male osteoporosis (mean age 50 years) further support the notion that a genetically driven deficit in bone mass acquisition, rather than accelerated bone loss, is central to the pathogenesis of idiopathic male osteoporosis [2,6]. Our present results are consistent with previous findings from a population-based study, in which LRP5 missense substitutions in exons 9 and 18 were associated with bone mass at the lumbar spine, specifically in males [13]. Noteworthy, the c.2047A–c.4037T haplotype previously shown to be associated with lower Z scores for lumbar spine BMD, bone mineral content (BMC) and projected area in males is the same haplotype (H3) associated with idiopathic male osteoporosis in this study. These findings are consistent with an influence of LRP5 alleles on
Table 3 Multiple logistic regression for osteoporosis in males
Haplotypes of LRP5 Age Weight Height Calcium intake Alcohol Smoking (smokers)
Coefficients
P values a
0.53 0.086 0.031 −0.029 −1.066 4.89 13.12
0.0036 0.006 0.004 0.49 0.002 0.005 0.004
Interactions are not shown. a Likelihood ratio test calculated to adjust the multiple regression according to the lack-of-fit technique.
Fig. 1. Odd ratio for LRP5 polymorphisms and idiopathic male osteoporosis. Legend: Odds and P values were calculated by logistic regression after adjustment for body size and multiple environmental variables, as explained in the Methods. The reference group for exon 9 was GG, for exon 18 CC and for haplotype groups H1/H1.
774
S.L. Ferrari et al. / Bone 37 (2005) 770–775
vertebral bone growth in male children [13]. Furthermore, a recent study performed in subjects of both genders with a large spectrum of BMD levels showed a stronger association between LRP5 polymorphisms and BMD when the analysis was restricted to males only [18]. All together, these data fit with studies showing a genderspecific inheritance of bone mineral density [19]. The precise mechanisms underlying the gender-specific effects of LRP5 genetic variations on bone mass however remain to be elucidated. One limitation of our study is the relatively small sample size due to the rare nature of idiopathic male osteoporosis. This may have precluded significant odd ratios for vertebral fractures to be found in association with LRP5 alleles. In contrast, false positive associations due to population stratification [20,21] are unlikely to have occurred since both osteoporotic and control subjects were born in the same country from European–Caucasian decent. Moreover, all cases and controls were investigated by a single clinician using the same DXA device in order to minimize data collection biases. Although no gene other than LRP5 has been associated with idiopathic male osteoporosis so far, other candidate genes could be involved as well. Among them, IGF-1 polymorphisms [22] deserve to be further investigated as IGF-1 plays a major role in bone mass development [23] and circulating levels of IGF-1 appear to be decreased in males with osteoporosis [24]. Other potential candidate genes for idiopathic male osteoporosis could be searched in the sex steroids pathway [25] since, in those men and their sons, a relative deficit in estradiol levels was observed [26]. Moreover, men with loss-of-function mutations in the estrogen receptor alpha (ESR1) or aromatase (CYP19) gene were reported to have severe osteoporosis [27,28]. However, variations in levels of sex steroids in a cohort with idiopathic osteoporosis and their sons were not related to the (TTTA)s-repeat polymorphism of the CYP19 gene [26]. Much larger studies including older men have shown association with polymorphism of other genes such as ESR1 or SOST [29,30]. However, we selected here a small sample of patients with a well-defined pathology, instead of a cohort. In conclusion, we have identified LRP5 variants as a genetic susceptibility factor for idiopathic male osteoporosis. While these findings need to be confirmed in other populations, they raise the prospect of new diagnostic and therapeutic approaches for this disease.
Acknowledgments This study is part of the NEMO thematic network sponsored by European Community. It was founded through a grant from “Direction de la Recherche Clinique” from Assistance Publique des Hopitaux de Paris to MC de Vernejoul (PHRC AOR01049).
References [1] Bilezikian JP. Osteoporosis in men. J Clin Endocrinol Metab 1999;84:3431–4. [2] Van Pottelbergh I, Goemaere S, Zmierczak H, De Bacquer D, Kaufman JM. Deficient acquisition of bone during maturation underlies idiopathic osteoporosis in men: evidence from a three-generation family study. J Bone Miner Res 2003;18:303–11. [3] Marie PJ, de Vernejoul MC, Connes D, Hott M. Decreased DNA synthesis by cultured osteoblastic cells in eugonadal osteoporotic men with defective bone formation. J Clin Invest 1991;88: 1167–72. [4] Johansson AG, Eriksen EF, Lindh E, Langdahl B, Blum WF, Lindahl A, et al. Reduced serum levels of the growth hormone-dependent insulin-like growth factor binding protein and a negative bone balance at the level of individual remodeling units in idiopathic osteoporosis in men. J Clin Endocrinol Metab 1997;82:2795–8. [5] Cohen-Solal ME, Baudoin C, Omouri M, Kuntz D, de Vernejoul MC. Bone mass in middle aged osteoporotic men and their relatives: familial effect. J Bone Miner Res 1998;13:1909–15. [6] Baudoin C, Cohen-Solal ME, Beaudreuil J, de Vernejoul MC. Genetic and environmental factors affect bone density variances of families of men and women with osteoporosis. J Clin Endocrinol Metab 2002;87:2053–9. [7] Gong Y, Slee RB, Fukai N, Rawadi G, Roman-Roman S, Reginato AM, et al. LDL receptor-related protein 5 (LRP5) affects bone accrual and eye development. Cell 2001;107:513–23. [8] Boyden LM, Mao J, Belsky J, Mitzner L, Farhi A, Mitnick MA, et al. High bone density due to a mutation in LDL-receptor-related protein 5. N Engl J Med 2002;346:1513–21. [9] Little RD, Carulli JP, Del Mastro RG, Dupuis J, Osborne M, Folz C, et al. A mutation in the LDL receptor-related protein 5 gene results in the autosomal dominant high-bone-mass trait. Am J Hum Genet 2002;70:11–9. [10] Van Wesenbeeck L, Cleiren E, Gram J, Beals RK, Benichou O, Scopelliti D, et al. Six novel missense mutations in the LDL Receptorrelated protein 5 (LRP5) gene in different conditions with increased bone density. Am J Hum Genet 2003;72:763–71. [11] Kato M, Patel MS, Levasseur R, Lobov I, Chang BH, Glass II DA, et al. Cbfa1-independent decrease in osteoblast proliferation, osteopenia, and persistent embryonic eye vascularization in mice deficient in Lrp5, a Wnt coreceptor. J Cell Biol 2002;157:303–14. [12] Babij P, Zhao W, Small C, Kharode Y, Yaworsky PJ, Bouxsein ML, et al. High bone mass in mice expressing a mutant LRP5 gene. J Bone Miner Res 2003;18:960–74. [13] Ferrari SL, Deutsch S, Choudhury U, Chevalley T, Bonjour JP, Dermitzakis ET, et al. Polymorphisms in the low-density lipoprotein receptor-related protein 5 (LRP5) gene are associated with variation in vertebral bone mass, vertebral bone size, and stature in whites. Am J Hum Genet 2004;74:866–75. [14] Niu T, Qin ZS, Xu X, Liu JS. Bayesian haplotype inference for multiple linked single-nucleotide polymorphisms. Am J Hum Genet 2002;70:157–69. [15] McCullagh P, Nelder JA. Generalized linear models. 2nd ed. London: Chapman and Hall; 1989. [16] Hastie T, Tibshirani R. Generalized additive models. London: Chapman and Hall; 1990. [17] S-PLUS 6 guide to statistics 2001 Seattle, WA, USA. Insightfull Corporation. [18] Koay MA, Woon PY, Zhang Y, Miles LJ, Duncan EL, Ralston SH, et al. Influence of LRP5 polymorphisms on normal variation in BMD. J Bone Miner Res 2004;19:1619–27. [19] Duncan EL, Cardon LR, Sinsheimer JS, Wass JA, Brown MA. Site and gender specificity of inheritance of bone mineral density. J Bone Miner Res 2003;18(8):1531–8. [20] Peacock M, Turner CH, Econs MJ, Foroud T. Genetics of osteoporosis. Endocr Rev 2002;23:303–26.
S.L. Ferrari et al. / Bone 37 (2005) 770–775 [21] Nguyen TV, Eisman JA. Genetics of fracture: challenges and opportunities. J Bone Miner Res 2000;15:1253–6. [22] Rosen CJ, Kurland ES, Vereault D, Adler RA, Rackoff PJ, Craig WY, et al. Association between serum insulin growth factor-I (IGF-I) and a simple sequence repeat in IGF-I gene: implications for genetic studies of bone mineral density. J Clin Endocrinol Metab 1998;83: 2286–90. [23] Zhang M, Xuan S, Bouxsein ML, von Stechow D, Akeno N, Faugere MC, et al. Osteoblast-specific knockout of the insulin-like growth factor (IGF) receptor gene reveals an essential role of IGF signaling in bone matrix mineralization. J Biol Chem 2002;277:44005–12. [24] Kurland ES, Rosen CJ, Cosman F, McMahon D, Chan F, Shane E, et al. Insulin-like growth factor-I in men with idiopathic osteoporosis. J Clin Endocrinol Metab 1997;82:2799–805. [25] Khosla S, Bilezikian JP. The role of estrogens in men and androgens in women. Endocrinol Metab Clin North Am 2003;32:195–218.
775
[26] Van Pottelbergh I, Goemaere S, Zmierczak H, Kaufman JM. Perturbed sex steroid status in men with idiopathic osteoporosis and their sons. J Clin Endocrinol Metab 2004;89:4949–53. [27] Smith EP, Boyd J, Frank GR, Takahashi H, Cohen RM, Specker B, et al. Estrogen resistance caused by a mutation in the estrogen-receptor gene in a man. New Engl J Med 1994;331:1056–61. [28] Bilezikian JP, Morishima A, Bell J, Grumbach MM. Increased bone mass as a result of estrogen therapy in a man with aromatase deficiency. New Engl J Med 1998;339:560–99. [29] Ioannidis JP, Ralston SH, Bennett ST, Brandi ML, Grinberg D, Karassa FB, et al. GENOMOS Study, differential genetic effects of ESR1 gene polymorphisms on osteoporosis outcomes. JAMA 2004;292:2105–14. [30] Uitterlinden AG, Arp PP, Paeper BW, Charmley P, Proll S, Rivadeneira F, et al. Polymorphisms in the sclerosteosis/van Buchem disease gene (SOST) region are associated with bone-mineral density in elderly whites. Am J Hum Genet 2004;75:1032–45.