Search tips
Search criteria 


Logo of wtpaEurope PMCEurope PMC Funders GroupSubmit a Manuscript
J Rheumatol. Author manuscript; available in PMC 2007 December 13.
Published in final edited form as:
Published online 2007 February 15.
PMCID: PMC2136207


Mirjam A Lips, Research Student,1 H E Syddall, MSc, Medical Statistician,1 Tom R Gaunt, PhD, Research Fellow,2 Santiago Rodriguez, PhD, Research Fellow,2 Ian NM Day, PhD, Professor of Human Genetics,2 Cyrus Cooper, MD, Professor of Rheumatology,1 E M Dennison, PhD, Reader in Rheumatology,1 and the Hertfordshire Cohort Study Group



We sought evidence of interaction between single nucleotide polymorphisms (SNPs) in the Calcium Sensing Receptor (CASR) gene and early life in determination of bone mineral density (BMD) among individuals from the Hertfordshire Cohort Study.


Four hundred and ninety eight men and 468 women aged 59-71 years were recruited. A lifestyle questionnaire was administered and BMD at lumbar spine and femoral neck was measured. DNA was obtained from whole blood samples using standard extraction techniques. Five SNPs of the CASR gene termed CASRV1 (rs1801725, G>T, S986A), CASRV2 (rs7614486, T>G, untranslated), CASRV3 (rs4300957, untranslated), CASRV4 (rs3804592 G>A, intron) and CASRV5 (rs1393189, T>C, intron) were analysed.


Among women the 11 genotype of the CASRV3 SNP was associated with higher lumbar spine BMD within the lowest birth-weight tertile, while the opposite pattern was observed among individuals in the highest birth-weight tertile (test for interaction on 1df, p=0.005, adjusted for age, BMI, physical activity, dietary calcium intake, cigarette and alcohol consumption, social class, menopausal status and HRT use). Similar relationships were seen at the total femur (p=0.042, fully adjusted) with birth-weight and at the total femur according to weight at one year tertile among women (p<0.001, fully adjusted). One haplotype was associated with lumbar spine BMD in women (p=0.008, fully adjusted); these findings were replicated in a second cohort.


We have found evidence of an interaction between a SNP of the CASR gene and birth weight in determination of bone mass in a UK female population.

Keywords: bone, bone density, cohort studies, genetic studies


Twin and family studies confirm an inherited contribution to peak bone mass, and various candidate genes have been proposed for the genetic regulation of bone mineral, including the genes for the vitamin D receptor (VDR), the estrogen receptor and for type I collagen (Col IA1). However, polymorphisms in these genetic loci explain only a small portion of the observed variance in bone mass in the general population [1-5]. Evidence is also accumulating that the risk of later osteoporosis might be programmed by environmental influences during intrauterine or early postnatal life. Growth in infancy, a marker of such programming predicts adult bone mass independently of adult life style [6-9]. We have recently published evidence of such an interaction between polymorphisms of the growth hormone gene, birth weight and adult bone mass in a cohort of Hertfordshire men and women [10,11]. However, little is known of the interaction between intrauterine environment and genetic markers of calcium metabolism and their association with adult bone mineral density (BMD).

In vitro studies have demonstrated that chondrogenic and osteogenic function can be influenced in response to different extracellular Ca concentrations [12]. Heath et al. described a polymorphism in exon 7 of the Calcium Sensing Receptor gene, coding the intracellular domain of the receptor [13]. This polymorphism, A986S, was found to be a predictor of circulating calcium concentrations in adult women [14;15]. However to date, results relating this gene to BMD have been conflicting. We sought to evaluate the interaction between polymorphisms of the CASR gene and early life in the determination of BMD in a large well-characterized elderly cohort.


In this study, which was designed to examine the relationship between growth in infancy and the subsequent risk of osteoporosis, the selection procedure was as follows: in brief, with the help of the National Health Service Central Registry at Southport, and Hertfordshire Family Health Service Association, we traced men and women who were born during 1931-39 in Hertfordshire, and still lived in East Hertfordshire in 1998. After obtaining written permission from each subject's General Practitioner, we approached each person by letter, asking him or her if they would be willing to be contacted by one of our research nurses. If they agreed, a research nurse performed a home visit, where they administered a structured questionnaire. This included information on socio-economic status, medical history, drug history, cigarette smoking, alcohol consumption, and reproductive variables in women. A total of 768 men and 714 women completed the home visit; of this number, 737 men and 675 women subsequently attended clinic.

At clinic, height was measured to the nearest 0.1cm using a Harpenden pocket stadiometer (Chasmors Ltd, London, UK) and weight to the nearest 0.1kg on a SECA floor scale (Chasmors Ltd, London, UK). Venous whole blood samples were taken at this clinic visit.

Eligible subjects were then invited to re-attend for bone density measurements. Individuals taking drugs known to alter bone metabolism (such as bisphosphonates) were excluded from this part of the study, although women taking hormone replacement therapy (HRT) were allowed to participate. There were no other exclusion criteria. Bone mineral density was measured in 498 men and 468 women, by dual energy X-ray absorptiometry at the lumbar spine and proximal femur (neck, total, intertrochanteric and trochanteric regions, Wards triangle) using a Hologic QDR 4500 instrument (Vertec Scientific, Reading, UK). Measurement precision error, expressed as coefficient of variation, was 1.55% for lumbar spine BMD, 1.45% for total femur and 1.83% for femoral neck BMD for the Hologic QDR 4500; these figures were scans on the same day, getting on and off the table between examinations. Short-term (two month) precision error for the QDR 4500 was less than 1% for both sites (manufacturers figures).

Genomic DNA was extracted from whole blood samples according to standard procedures. Single nucleotide variants in the Calcium Sensing Receptor gene termed CASRV1 (rs1801725, G>T, S986A), CASRV2 (rs7614486, T>G, untranslated), CASRV3 (rs4300957, untranslated), CASRV4 (rs3804592 G>A, intron) and CASRV5 (rs1393189, T>C, intron) were analysed.

The different alleles were termed 1 or 2, resulting in genotypes 11, 12 and 22. Genotypes 12 and 22 were combined in the event of a low frequency of genotype 22. Analyses were conducted separately for men and women using STATA 8. The relationship between each continuously distributed phenotype variable and each SNP was explored using both analysis of variance and linear regression models. Analyses were conducted with and without adjustment for age, BMI, typical activity level, dietary calcium intake, smoking status, alcohol intake, current social class and menopausal status and hormone replacement therapy use for women. Haplotypes were inferred using the PHASE package [16].

The East and North Hertfordshire Ethical Committees granted ethical permission for the study. All participants gave written informed consent.


The characteristics of the study population at baseline are displayed in table 1. The mean age of the men and women studied was 64.3 and 65.6 years respectively. Thirty four percent of the men and 62 percent of the women had never smoked, while 52% of the men (28% of the women) and 15% of the men (10% of the women) were ex-smokers and current smokers respectively. Four percent of men and 18 percent of women were non-drinkers, while 21% of men and 12% of women were moderate drinkers (11-21 and 8-14 units of alcohol per week respectively, 1 unit being a single glass of wine or a single measure of spirits).

Table I
Summary characteristics of study participants

The distribution frequencies of each genotype were as follows: CASRV1: 11-78.1%, 12/22-21.9%; CASRV2: 11-58%, 12-35.7%, 22-6.3%; CASRV3: 11-71.1%, 12-26.0%, 22-2.9%, CASRV4: 11-72.7%, 12-25.0%, 22-2.3%; CASRV5: 11-80.8%, 12/22-19.3%. Among men, the CASRV2 22 genotype was underrepresented, and the CASRV2 12 genotype over-represented (p=0.008); similarly, among men the CASRV3 22 genotype was under-represented and the SASRV3 12 genotype over-represented (p=0.02). None of the other genotypes differed by sex. There was no evidence of departure from Hardy-Weinberg equilibrium. A total of 22.2% had at least one S allele of the A986S polymorphism (coded CASRV1). Taking into account the poor power of the recessive genotype (SS), the population for this gene was divided into two groups according to the presence or absence of the S allele.

There were no statistically significant direct associations between any of the CASR SNPs studied and bone mass at the lumbar spine or femoral neck in the cohort as a whole, whether women taking HRT were excluded or not (Table 2). However, we also investigated the possible interaction between genotype and early environment as predictors of bone mass (Table 3). Among women, the 11 genotype of the CASRV3 SNP was associated with higher lumbar spine BMD within the lowest birth weight tertile, while this genotype was associated with lower lumbar spine BMD in the highest birth weight tertile (test for interaction p=0.005, fully adjusted for age, social class, body mass index, physical activity, calcium intake, cigarette and alcohol consumption, menopausal status & HRT use). Similar relationships were observed at the total femur (test for interaction, fully adjusted p=0.042). In addition, the 1 allele of the CASRV3 SNP was associated with higher total femoral BMD among women in the lowest weight at one-year tertile (test for interaction, fully adjusted p<0.001); in the highest weight at one-year tertile the 1 allele was, again, associated with lower total femoral BMD (figure 1). We expanded these analyses to utilise other cut-points for early life parameters (birth weight <2500g, 2500-3000g, 3000-3500g, 3500-4000g, >4000g). While these revised groupings made little difference to our results, further interpretation was difficult due to the small numbers of women in the lowest birth weight group (11: 11 subjects 12: 10 subjects 22: 0 subjects). No significant genotype-early environment interactions were observed in men.

Figure 1
Total femoral BMD by CASRV3 genotype according to weight at 1 year of age in 444 women
Table 2
Lumbar spine, femoral neck and total femoral BMD in relation to genotype
Table 3
Lumbar spine and total femoral BMD in women according to genotype and birth weight and weight at one year

Six haplotypes of the five CASR single nucleotype polymorphisms accounted for 92.3% of all the haplotypes defined in the Hertfordshire Cohort Study dataset; 11111 11121 12211 21111 11112 12111. Significant associations were observed between the 12211 haplotype (the third most frequent in this population) and lumbar spine BMD in women (Table 4, p=0.008, adjusted for age at clinic, BMI, physical activity, dietary calcium score, social class, cigarette and alcohol consumption, years since menopause and HRT use in women). We sought to replicate our findings using an older community from the Hertfordshire Cohort that has been extensively described previously [10]. Among this group of 186 men and 122 women aged 61-73 years at study, our findings were replicated, with the 12211 haplotype once again being associated with higher lumbar spine BMD among women (p=0.05, fully adjusted).

Table 4
Associations of BMD with haplotypes of the CaSR gene among men and women from the Hertfordshire Cohort Study


We have demonstrated evidence of an interaction between a single nucleotide polymorphism of the CASR gene and early environment in the effect on bone mineral density in a Hertfordshire population of healthy elderly women. To our knowledge this is the first report relating CASR genotype to BMD taking into account the possible effects of development in early life. We have also demonstrated that one haplotype was associated with lumbar spine BMD in women, a finding replicated in a second population.

There were a number of limitations to this study. The individuals studied were selected because we have accurate records of their early life; these subjects were all born in Hertfordshire, UK, and still live there. They have however, previously been shown to have anthropometric and lifestyle characteristics similar to those of the general population [17]. The allelic frequency of the recessive alleles of both CASRV1 and CASRV3 polymorphisms was fairly low. However, the effect of the CASRV3 recessive genotype is consistent in BMD measurements at different sites and different ages, suggesting that our results are not a coincidental finding. Our genotype frequencies of the A986S (CASRV1) polymorphism, however, are comparable to those found by other groups.

We included SNPs located in noncoding and intergenic regions rather that exclusively focusing on the coding region. It is hypothesized that variations underlying complex diseases are not limited to the structure of the encoded protein. Gene regulation is the result of the combinatorial action of multiple transcription factors binding at multiple sites in and near a gene and it has recently been shown that gene expression regulatory elements reside in noncoding and intergenic regions [18].

Calcium homeostasis has an important role in the regulation of bone remodelling and alterations of the mechanisms involved in this regulation may contribute to the development of metabolic bone diseases. The main homeostatic regulator of extracellular calcium is the calcitropic hormone PTH. However, the CASR mediates this pathway through its ability to sense small changes in circulating calcium concentration and to activate intracellular pathways if its setpoint is not met. These pathways lead to an increase in PTH which in turn will enhance the activation of vitamin D [19]. Both hormones increase circulating calcium by stimulating calcium reabsorption in renal tubules and intestine and by stimulating bone resorption through promoting the differentiation of osteoclasts from multinucleated precursors [20;21].

The CASR is defective in individuals with familial hypocalciuric hypercalcemia (FHH) and neonatal severe hyperparathyroidism (NSHPT) due to inactivating mutations [22;23]. Heath at al. suggested that the NH2-terminal extracellular and membrane spanning regions of the receptor protein are functional domains for calcium binding and signal transduction, and mutations in these regions lead to familial benign hypocalciuric hypercalcemia (FBHH) [13].

An association between a CA-repeat polymorphism of the CASR gene and BMD was found by Tsukamoto et al [24]. Cole et al [14;15] reported associations between the A986S polymorphism and calcium levels in a clinical trial of 163 white Canadian women. The AA genotype was associated with lower calcium levels, indicating a loss of function at the receptor level as in FHH. They suggested that this polymorphism would have a potential role as a predictor of disorders that affect bone and mineral metabolism. Lorentzon et al. [25] and Eckstein et al. [26] confirmed this effect of polymorphisms on circulating calcium concentrations and found associations with BMD. However, after correction for physical activity, the polymorphisms no longer had an effect on BMD. Eckstein et al. found the S allele to be over-represented in a group of 80 Israeli with low BMD, though age at menarche was found to be the main predictor of BMD in their group. Several other groups have examined the relationship between extracellular calcium or BMD and the A986S polymorphism without finding a significant association. Tackaks et al. [27] and Cetani et al. [28] studied the polymorphism in a homogenous postmenopausal white Hungarian and Italian, female population respectively and confirmed the functionality of the polymorphism in circulating calcium levels, but could not confirm any influence of the polymorphism on BMD. Young et al. [29] and Bollerslev et al. [30] analysed calcium data for, respectively, 102 New Zealand and 1252 Canadian postmenopausal women. Bollerslev et al. could neither find significance of the A986S polymorphism in association with serum calcium levels nor with BMD. Most recently, Perez-Castrillon et al failed to find an association of the A986S polymorphism with lumbar spine BMD in 48 hypertensive women [31].

In our large homogeneous study group we found the same allele frequencies of the A and S alleles as previously found by other groups. However, we could not find a direct effect of the CASR A986S polymorphism on BMD and can confirm that the results of the latest studies also apply for our UK cohort. There are several possible explanations for this negative result. The CASR may not have a crucial role in the regulation of osteoblast function. However, we also studied 4 other polymorphisms and found an interaction between the CASRV3 (C>T, untranslated) polymorphism and development in early life in determination of BMD in women. We previously reported data suggesting that the intrauterine environment, using birth weight as a marker, and early life, using weight at 1 year of age as a marker, may modulate the relationship between the VDR and GH genes and adult bone mineral density [10;11;32]. Therefore we hypothesized that the effect of polymorphisms of the CASR gene on BMD may be modulated by development in early life. This study provides evidence of an interaction between the early environment and a genotypic locus such that individuals who carry the unfavourable, recessive, genotype, who also grew poorly in fetal life and infancy, are susceptible to lower BMD, whereas the recessive genotype predisposes to higher BMD among individuals that grew better in early life.

Males have greater bone mineral content (BMC) and area, indicating a larger skeletal size, and this is normally reflected in a higher BMD in males compared with females [33]. However, when adequate correction for body size is performed (by calculation of bone mineral apparent density for example), the apparent differences between the sexes are much reduced, or removed. However, there are also known to be sex differences in age-related changes in bone loss rates and bone strength. Hence, in one study, men had approximately 30% larger cross-sectional bone size compared with women at age 67-69 years [34]. At all sites, women had two- to five-fold reduction in bone mass with age compared to men but had comparable increments in bone size. This was reflected in significantly worse bone strength measures with age in women. The sex hormones are likely to be important contributors to the sexual dimorphism we describe. The effects of the sex steroid receptors on skeletal regulation has been tested recently in murine models [35]; androgen receptor deficient mice follow the male pattern of long bone development but imitate females in bone density and trabecular bone, while loss of the α subtype of the estrogen receptor resulted in increased bone length in females but reduced bone length in males. Loss of either receptor resulted in increased osteoblast sensitivity to PTH. Our study suggested an interaction between birth weight and the CASR with regard to BMD among women but not men. Our study group were some years post menopause and hence their BMD would reflect the peak bone mass attained and the rate of subsequent bone loss. It is possible that the interaction we describe may reflect an interaction between birth weight and bone loss. Although the data we present are cross-sectional, the cohort is being followed up, and we will be able to test for an interaction with bone loss rate shortly.

In conclusion, our study shows, for the first time, evidence of an interaction between a polymorphism of the CASR gene and growth in early life in the determination of bone mass in later life in a UK female population. Further work is now indicated to replicate this finding.


We thank the men and women who participated in the study, the General Practitioners who allowed access to their patients and the nurses and radiology staff who administered the bone density measurements. Computing support was provided by Vanessa Cox.

Grants held: This study was supported by grants from the the Arthritis Research Campaign and the Medical Research Council.

Reference List

1. Seeman E, Hopper JL, Bach LA, Cooper ME, Parkinson E, McKay J, et al. Reduced bone mass in daughters of women with osteoporosis. N.Engl.J.Med. 1989;320:554–558. [PubMed]
2. Gueguen R, Jouanny P, Guillemin F, Kuntz C, Pourel J, Siest G. Segregation analysis and variance components analysis of bone mineral density in healthy families. J.Bone Miner.Res. 1995;10:2017–2022. [PubMed]
3. Pocock NA, Eisman JA, Hopper JL, Yeates MG, Sambrook PN, Eberl S. Genetic determinants of bone mass in adults. A twin study. J.Clin.Invest. 1987;80:706–710. [PMC free article] [PubMed]
4. Kelly PJ, Eisman JA. Sambrook PN Interaction of genetic and environmental influences on peak bone density. Osteoporos.Int. 1990;1:56–60. [PubMed]
5. Slemenda CW, Christian JC, Williams CJ, Norton JA, Johnston CC., Jr Genetic determinants of bone mass in adult women: a reevaluation of the twin model and the potential importance of gene interaction on heritability estimates. J.Bone Miner.Res. 1991;6:561–567. [PubMed]
6. Cooper C, Cawley M, Bhalla A, Egger P, Ring F, Morton L, et al. Childhood growth, physical activity, and peak bone mass in women. J.Bone Miner.Res. 1995;10:940–947. [PubMed]
7. Cooper C, Fall C, Egger P, Hobbs R, Eastell R, Barker D. Growth in infancy and bone mass in later life. Ann.Rheum.Dis. 1997;56:17–21. [PMC free article] [PubMed]
8. Jones G, Riley M, Dwyer T. Maternal smoking during pregnancy, growth, and bone mass in prepubertal children. J.Bone Miner.Res. 1999;14:146–151. [PubMed]
9. Antoniades L, MacGregor AJ, Andrew T, Spector TD. Association of birth weight with osteoporosis and osteoarthritis in adult twins. Rheumatology. 2003;42:791–796. [PubMed]
10. Dennison EM, Syddall HE, Rodriguez S, Voropanov A, Day IN, Cooper C. Polymorphism in the growth hormone gene, weight in infancy, and adult bone mass. J.Clin.Endocrinol.Metab. 2004;89:4898–4903. [PubMed]
11. Day IN, Chen XH, Gaunt TR, King TH, Voropanov A, Ye S, et al. Late life metabolic syndrome, early growth, and common polymorphism in the growth hormone and placental lactogen gene cluster. J.Clin.Endocrinol.Metab. 2004;89:5569–5576. [PubMed]
12. Chang W, Tu C, Chen TH, Komuves L, Oda Y, Pratt SA, et al. Expression and signal transduction of calcium-sensing receptors in cartilage and bone. Endocrinology. 1999;140:5883–5893. [PubMed]
13. Heath H, III, Odelberg S, Jackson CE, Teh BT, Hayward N, Larsson C, et al. Clustered inactivating mutations and benign polymorphisms of the calcium receptor gene in familial benign hypocalciuric hypercalcemia suggest receptor functional domains. J.Clin.Endocrinol.Metab. 1996;81:1312–1317. [PubMed]
14. Cole DE, Peltekova VD, Rubin LA, Hawker GA, Vieth R, Liew CC, et al. A986S polymorphism of the calcium-sensing receptor and circulating calcium concentrations. Lancet. 1999;353:112–115. [PubMed]
15. Cole DE, Vieth R, Trang HM, Wong BY, Hendy GN, Rubin LA. Association between total serum calcium and the A986S polymorphism of the calcium-sensing receptor gene. Mol.Genet.Metab. 2001;72:168–174. [PubMed]
16. Stephens M, Smith NJ, Donnelly P. A new statistical method for haplotype reconstruction from population data. Am.J.Hum.Genet. 2001;68:978–989. [PubMed]
17. Syddall HE, Aihie Sayer A, Dennison EM, Martin HJ, Barker DJP, Cooper C. Cohort profile: The Hertfordshire Cohort Study. 2005;34:1234–42. [PubMed]
18. Luque RM, Kineman RD, Park S, Peng XD, Gracia-Navarro F, Castano JP, et al. Homologous and heterologous regulation of pituitary receptors for ghrelin and growth hormone-releasing hormone. Endocrinology. 2004;145:3182–3189. [PubMed]
19. Brown EM, Gamba G, Riccardi D, Lombardi M, Butters R, Kifor O. Cloning and characterization of an extracellular Ca(2+)-sensing receptor from bovine parathyroid. Nature. 1993;366:575–580. [PubMed]
20. Purroy J, Spurr NK. Molecular genetics of calcium sensing in bone cells. Hum.Mol.Genet. 2002;11:2377–2384. [PubMed]
21. Spurr NK. Genetics of calcium-sensing--regulation of calcium levels in the body. Curr.Opin.Pharmacol. 2003;3:291–294. [PubMed]
22. Pollak MR, Brown EM, Chou YH, Hebert SC, Marx SJ, Steinmann B. Mutations in the human Ca(2+)-sensing receptor gene cause familial hypocalciuric hypercalcemia and neonatal severe hyperparathyroidism. Cell. 1993;75:1297–1303. [PubMed]
23. Pollak MR, Brown EM, Estep HL, McLaine PN, Kifor O, Park J. Autosomal dominant hypocalcaemia caused by a Ca(2+)-sensing receptor gene mutation. Nat.Genet. 1994;8:303–307. [PubMed]
24. Tsukamoto K, Orimo H, Hosoi T, Miyao M, Ota N, Nakajima T. Association of bone mineral density with polymorphism of the human calcium-sensing receptor locus. Calcif.Tissue Int. 2000;66:181–183. [PubMed]
25. Lorentzon M, Lorentzon R, Lerner UH, Nordstrom P. Calcium sensing receptor gene polymorphism, circulating calcium concentrations and bone mineral density in healthy adolescent girls. Eur.J.Endocrinol. 2001;144:257–261. [PubMed]
26. Eckstein M, Vered I, Ish-Shalom S, Shlomo AB, Shtriker A, Koren-Morag N, et al. Vitamin D and calcium-sensing receptor genotypes in men and premenopausal women with low bone mineral density. Isr.Med.Assoc.J. 2002;4:340–344. [PubMed]
27. Takacs I, Speer G, Bajnok E, Tabak A, Nagy Z, Horvath C. Lack of association between calcium-sensing receptor gene “A986S” polymorphism and bone mineral density in Hungarian postmenopausal women. Bone. 2002;30:849–852. [PubMed]
28. Cetani F, Pardi E, Borsari S, Vignali E, Dipollina G, Braga V, et al. Calcium-sensing receptor gene polymorphism is not associated with bone mineral density in Italian postmenopausal women. Eur.J.Endocrinol. 2003;148:603–607. [PubMed]
29. Young R, Wu F, Van de WN, Ames R, Gamble G, Reid IR. Calcium sensing receptor gene A986S polymorphism and responsiveness to calcium supplementation in postmenopausal women. J.Clin.Endocrinol.Metab. 2003;88:697–700. [PubMed]
30. Bollerslev J, Wilson SG, Dick IM, Devine A, Dhaliwal SS, Prince RL. Calcium-sensing receptor gene polymorphism A986S does not predict serum calcium level, bone mineral density, calcaneal ultrasound indices, or fracture rate in a large cohort of elderly women. Calcif.Tissue Int. 2004;74:12–17. [PubMed]
31. Perez-Castrillon JL, Sanz A, Silva J, Justo I, Velasco E, Duenas A. Calcium-sensing receptor gene A986S polymorphism and bone mass in hypertensive women. Arch Med Res. 2006;37:607–11. [PubMed]
32. Dennison EM, Arden NK, Keen RW, Syddall H, Day IN, Spector TD, et al. Birthweight, vitamin D receptor genotype and the programming of osteoporosis. Paediatr.Perinat.Epidemiol. 2001;15:211–219. [PubMed]
33. Tuck SP, Pearce MS, Rawlings DJ, Birrell FN, Parker L, Francis RM. Differences in bone mineral density and geometry in men and women: the Newcastle Thousand Families Study at 50 years old. Br J Radiol. 2005;78:493–498. [PubMed]
34. Sigurdsson G, Aspelund T, Chang M, Jonsdottir B, Sigurdsson S, Eiriksdottir G, Gudmondsson A, Harris TB, Gudnason V, lang TF. Increasing sex difference in bone strength in old age: The Age, Gene/Environment Susceptibility-Reykjavik study (AGES-REYKJAVIK) Bone. 2006;39:644–651. [PubMed]
35. Tozum TF, Oppenlander ME, Koh-Paige AJ, Robins DM, McCauley LK. Effects of sex steroid receptor specificity in the regulation of skeletal metabolism. Calcif Tissue Int. 2004;75:60–70. [PubMed]