Skip to main content

Bone mineral density in high-level endurance runners: Part B—genotype-dependent characteristics

Abstract

Purpose

Inter-individual variability in bone mineral density (BMD) exists within and between endurance runners and non-athletes, probably in part due to differing genetic profiles. Certainty is lacking, however, regarding which genetic variants may contribute to BMD in endurance runners and if specific genotypes are sensitive to environmental factors, such as mechanical loading via training.

Method

Ten single-nucleotide polymorphisms (SNPs) were identified from previous genome-wide and/or candidate gene association studies that have a functional effect on bone physiology. The aims of this study were to investigate (1) associations between genotype at those 10 SNPs and bone phenotypes in high-level endurance runners, and (2) interactions between genotype and athlete status on bone phenotypes.

Results

Female runners with P2RX7 rs3751143 AA genotype had 4% higher total-body BMD and 5% higher leg BMD than AC + CC genotypes. Male runners with WNT16 rs3801387 AA genotype had 14% lower lumbar spine BMD than AA genotype non-athletes, whilst AG + GG genotype runners also had 5% higher leg BMD than AG + GG genotype non-athletes.

Conclusion

We report novel associations between P2RX7 rs3751143 genotype and BMD in female runners, whilst differences in BMD between male runners and non-athletes with the same WNT16 rs3801387 genotype existed, highlighting a potential genetic interaction with factors common in endurance runners, such as high levels of mechanical loading. These findings contribute to our knowledge of the genetic associations with BMD and improve our understanding of why some runners have lower BMD than others.

Introduction

Individuals who complete higher levels of weight-bearing physical activity tend to have higher bone mineral density (BMD) (Warburton et al. 2006). Despite this, some athletes, such as endurance runners may be at risk of low BMD and increased risk of stress fracture injury, which negatively impacts both health and performance (Pollock et al. 2010). Excessive training volumes and/or dietary restriction undertaken by this population can result in reduced energy availability, which can negatively impact bone metabolism and potentially reduce BMD (Papageorgiou et al. 2018).

Nonetheless, inter-individual variability in bone phenotypes exist, even within sport-specific cohorts (as demonstrated in Part A), which may be explained in part by genetic factors. Heritability of BMD is reportedly 50–80% (Ralston and Uitterlinden 2010), with 98 loci having been associated with total, femoral neck and lumbar spine dual-energy X-ray absorptiometry (DXA)-derived BMD previously (Trajanoska et al. 2019). Few investigations, have replicated these associations independently or considered gene-environment interactions (Trajanoska et al. 2019). It remains unclear whether certain genes may be sensitive to mechanical loading from physical activity and what the outcome is of such an interaction for BMD and injury risk (Herbert et al. 2019).

Mitchell et al. (2016) investigated genomic components of bone phenotypes and the relationship with physical activity using SNPs that had previously been associated with BMD via GWAS (Estrada et al. 2012). The authors observed nominal interactions at the lumbar spine in children between physical activity and variants such as Wnt family member 16 (WNT16) rs3801387 and axin 1 (AXIN1) rs9921222. In athletes competing in weight-bearing sports specifically, higher total-body BMD was shown in the vitamin D receptor (VDR) FokI rs2228570 GG and GA, but not AA, genotypes compared to non-athlete controls (Nakamura et al. 2002b). Interestingly, within swimmers, a lower total BMD was observed in the GG genotype when compared to non-athlete controls (Nakamura et al. 2002b). Together, these findings suggest that individuals with the GG genotype may be more responsive to mechanical loading, resulting in greater BMD under conditions of high mechanical loading, but lower BMD when the mechanical loading is less, such as in swimming. Thus, phenotypes can vary between individuals of the same genotype, especially when that genotype demonstrates sensitivity to mechanical loading.

It is possible that athletes tend to possess advantageous variants of genes that are responsive to mechanical loading and others that are important for the attainment of peak BMD. This may result in a competitive advantage via more consistent and higher volume training because of the higher BMD and subsequent reduced injury risk. Some endurance runners are at risk of reduced energy availability due to the potential undertaking of high training volumes and/or insufficient energy intake, which may negatively impact BMD. Consequently, possessing an advantageous genetic predisposition for BMD is likely to be of greater importance for the health and performance of populations at risk of low BMD, such as endurance runners, than athletic populations competing in sports that exhibit high peak and multi-directional forces on bone. Those who possess a more advantageous genotype may have a greater response to loading, resulting in higher BMD and a reduced risk of a stress fracture. The potential consequence of a disadvantageous genotype has been demonstrated in dancers, who are also at risk of low BMD. Specifically, genetic variants in the oestrogen receptor and the Wnt/β-catenin pathways were associated with an increased prevalence of low BMD in elite dancers (Amorim et al. 2018). Moreover, the increased prevalence of low BMD in the dancers was not predicted by previously suggested risk factors of body mass, menstrual disturbances and energy availability. Thus, low BMD in the absence of known risk factors further emphasises the potential modulation of the BMD phenotype via genetic characteristics (Amorim et al. 2018).

Only a small number of BMD-associated genetic variants have been explored in athletic populations, or in relation to gene-physical activity interactions, so the genetic influence on BMD in athletic populations is still unknown. Endurance runners, in particular, experience high volumes of mechanical loading at certain sites (e.g. tibia) but less loading at others (e.g. lumbar spine). Despite experiencing high volumes of loading, some endurance runners may have low BMD and thus present a suitable population in which to investigate gene-physical activity interactions vis-à-vis BMD (Herbert et al. 2019). Variability as well as differences in bone phenotypes between high-level endurance runners and non-athletes are reported in Part A of this two-part investigation. The purpose of this second part was to (1) investigate whether the 10 SNPs, individually and collectively, associated with phenotypes including total-body BMD (TBMD), leg BMD (LBMD), lumbar spine BMD (LSBMD), total-body T-score and total-body Z-score in high-level endurance runners; (2) investigate whether being an endurance runner or non-athlete affected any association between genotype and bone phenotypes. The 10 SNPs chosen for investigation all have reported functional effects on bone and have also been identified via genome-wide association study and/or candidate gene association study as being associated with BMD (Table 1). We hypothesised associations between genotype and bone phenotypes in high-level endurance runners, plus the existence of genotype-athlete status interactions on bone phenotypes due to the impact of long-term mechanical loading in runners.

Table 1 Gene functions and previous associations of SNP with bone phenotypes and/or metabolism

Materials and methods

Participants and participant recruitment

Participants comprised 103 high-level endurance runners (45 men, 58 women) and 112 (52 men, 60 women) ethnically matched non-athletes. Briefly, runners were included if they had completed at least one official long-distance event (≥ 3000 m) faster than a pre-determined threshold time (Table 1, Part A) and were considered national/international standard. Please see Part A for a full description of participant characteristics. All experimental procedures were conducted in accordance with the guidelines in the Declaration of Helsinki and approved by the local Ethics Committee of Manchester Metropolitan University.

Protocol

All runners completed a questionnaire that detailed geographic ancestry, performance, training practices, injury and sporting history, whilst female runners also provided menstruation history. Non-athletes completed a questionnaire in relation to ethnicity, general health and physical activity level to establish matched ethnic ancestry and ensure no history of high-level sporting competition. DXA scans on all participants were completed following the manufacturer’s guidelines to obtain TBMD, LBMD, LSBMD, total-body T- and Z-scores.

All participants also provided either a blood, saliva or buccal swab sample, from which DNA was extracted and genotyped for the 10 SNPs, which were selected according to the volume and strength of evidence in the literature of their association with BMD and/or biological function/mechanism in relation to BMD (Table 1).

For blood, a 5 mL sample was collected from a superficial forearm vein into EDTA collection tubes and stored in 1.5 mL microcentrifuge tubes (Eppendorf AG, Hamburg, Germany) at − 20 °C. Saliva samples were collected following a minimum 30-min abstinence from food and drink into Oragene DNA OG-500 collection tubes (DNA Genotek Inc., Ontario, Canada) in accordance with the manufacturer’s guidelines before being stored at room temperature. For buccal cell sample collection, participants brushed one OmniSwab collection tip (Whatman Sterile OmniSwab, GE Healthcare, USA) against the inside of one cheek for 30 s before repeating this with a second swab on the opposite cheek to obtain two samples (following the same abstinence as saliva samples) before being stored at − 20 °C in a 2 mL microcentrifuge tube. DNA was extracted using the Qiagen QIAcube spin protocol (Qiagen, Crawley, UK) and the Qiagen DNA Blood Mini Kit (Qiagen) for whole blood, saliva and buccal samples in accordance with the manufacturer’s guidelines. Approximately 75% of participant DNA was obtained from blood, 23% from saliva and 2% via buccal swabs.

All participants were genotyped for the 10 SNPs using the fluorophore-based detection technique of TaqMan real-time polymerase chain reaction (PCR) on either the Fluidigm EP1 (Fluidigm, Cambridge, UK) or StepOnePlus (Applied Biosystems, Paisley, UK). End-point fluorescence measurement of VIC and FAM determined the different genotypes for the 10 SNPs using the software supplied by the respective manufacturers of each PCR machine.

The majority of samples (95%) were genotyped via the Fluidigm EP1 by combining 2 μL GTXpress Master Mix (X2) (Applied Biosystems), 0.2 μL 20X Fast GT Sample Loading Reagent (Fluidigm), 0.2 μL nuclease-free H2O and 1.6 μL of purified DNA into each well of a 192 × 24 microchip. Negative controls were placed into 4 wells on each 192 × 24 microchip, in which nuclease-free H2O replaced the DNA sample. 1.78 μL assay (20X) (Applied Biosystems), 1.78 μL 2X Assay Loading Reagent (Fluidigm) and 0.18 μL ROX reference dye (Invitrogen, Paisley, UK) were combined per assay inlet. An integrated fluid circuit controller RX (Fluidigm) was used to mix samples and assays using a Load Mix (166X) script. PCR was performed using a real-time FC1 Cycler (Fluidigm) GT 192 × 24 Fast v1 protocol. Denaturation began at 95 °C for 120 s followed by 45 cycles of incubation at 95 °C for 2 s and then annealing and extension at 60 °C for 20 s before end-point analysis was completed in the EP1 reader. Genotyping was performed with the Fluidigm SNP genotyping analysis software.

The remaining 5% of samples were genotyped by combining 5 μL Genotyping Master Mix or GTXpress Master Mix (Applied Biosystems), 4.3 μL H2O, 0.5 μL assay (Applied Biosystems), and 0.2 μL of purified DNA (~ 9 ng), for samples derived from blood and saliva into wells on a 96-well plate (MicroAmp EnduraPlate Optical 96-Well Clear Reaction Plate, Applied Biosystems). For DNA taken from buccal swabs, 5 μL Genotyping Master Mix was combined with 3.5 μL H2O, 0.5 μL assay mix, and 1 μL DNA solution (~ 9 ng DNA). Negative controls were also placed into 2 wells on each 96-well plate, in which nuclease-free H2O replaced the DNA sample. Each well on a 96-well plate contained a total reaction volume of ~ 10 μL before being covered with an optical seal (MicroAmp Optical Adhesive Film, Applied Biosystems). PCR was performed using a StepOnePlus Real-Time PCR system (Applied Biosystems). For GTXpress Master Mix, an initial 20 s at 95 °C was followed by 50 cycles of denaturation for 3 s at 95 °C, then annealing and extension at 60 °C for 20 s. For Genotyping Master Mix, denaturation began at 95 °C for 10 min, with 40 cycles of incubation at 92 °C for 15 s and then annealing and extension at 60 °C for 1 min. Genotyping analysis was performed with StepOnePlus software version 2.3.

All samples were analysed in duplicate and were in 100% agreement. Similarly, there was 100% agreement between the StepOnePlus and Fluidigm PCR systems as determined by the analysis of one variant on 94 samples.

Statistical analysis

Using additive (AA vs. Aa vs. aa), dominant (AA + Aa vs. aa) and recessive (AA vs. Aa + aa) genetic models for men and women separately, analysis of variance (ANOVA) was implemented to investigate associations between the 10 SNPs individually, and collectively as a total genotype score (TGS), and bone phenotypes (TBMD, LBMD, LSBMD, T-score, Z-score) in runners. Recessive and dominant models were executed with respect to the allele considered disadvantageous for BMD. For TGS, each SNP was allocated scores according to existing literature (Williams and Folland 2008), where the homozygote associated with higher BMD was given a score of 2, the heterozygotes scored 1 and the other homozygote given 0. The total score was then transformed to lie within 0–100 (e.g. TGS = 100/20 × (2 + 1 + 0 + 1 + 1 + 1 + 0 + 2 + 1 + 2) = 55). Consequently, participants were allocated to either a “low” (≤ 55), “moderate” (60–70) or “high” (≥ 75) TGS group based upon calculated TGS score. Significant associations from any additive model were further explored using Bonferroni post-hoc tests. ANOVA was also used to explore interactions between the 10 SNPs, individually and collectively, and athlete status on bone phenotypes. Following any significant interaction between athlete status and genotype for a bone phenotype, simple main effects with pairwise comparisons were conducted to analyse cohort-dependent differences in these bone phenotypes between runners and non-athletes across the same genotype. To control for Type 1 statistical errors from multiple testing, a false discovery rate (FDR) of 0.2 (Benjamini and Hochberg 1995) was applied within each family of ANOVAs for each SNP and uncorrected P-values are reported except where stated. Consequently, two FDR models for each SNP (each including 15 P-values for every ANOVA implemented) were applied to both the analyses within the runners and the analyses exploring the interactions between genotype and athlete status on bone phenotype. Due to an insufficient number of participants with two copies of the minor allele, COL1A1 rs1800012 female analyses only included 5 P-values for each FDR model whilst LRP5 rs3736228 male analyses only included 10 P-values for one FDR model (interaction analyses). Alpha was set at 0.05 and data are reported as mean (SD).

Results

Participant characteristics and phenotype data for men and women are as described in Part A of this investigation.

Genetic associations in runners

In the dominant analysis model, female runners with P2RX7 rs3751143 AA genotype had 4% higher TBMD (P = 0.052) and 5% higher LBMD (P = 0.036) than AC + CC genotype female runners but no differences were evident for LSBMD (P = 0.512) as shown in Fig. 1. Additionally, AA genotype female runners had a higher T-score (1.36 vs 0.81; P = 0.047) and Z-score (1.27 vs 0.70; P = 0.017) in comparison to AC + CC genotypes.

Fig. 1
figure1

Mean total bone mineral density (TBMD), leg bone mineral density (LBMD) and lumbar spine bone mineral density (LSBMD) according to P2RX7 rs3751143 genotype, AA (n = 36) vs AC + CC (n = 22), in female high-level endurance runners. Higher TBMD and LBMD but not LSBMD in AA than AC + CC genotypes. Error bars denote standard deviation

In the additive analysis model, a main effect of P2RX7 rs3751143 genotype on TBMD (P = 0.016), LBMD (P = 0.080), T-score (P = 0.018) and Z-score (P = 0.013) but not LSBMD (P = 0.514) existed within the female runners. Following post-hoc analysis, those with AA genotype possessed 5% higher TBMD than AC genotypes (P = 0.036) but no difference was evident for LBMD (P = 0.077) as shown in Fig. 2. AA genotypes also had a higher T-score (1.36 vs 0.64; P = 0.034) and Z-score (1.26 vs 0.57; P = 0.016) than AC genotypes. No genotype-dependent differences for P2RX7 rs3751143 on any bone phenotype were present within runners in the recessive analysis model (corrected P ≥ 0.288).

Fig. 2
figure2

Mean total bone mineral density (TBMD), leg bone mineral density (LBMD) and lumbar spine bone mineral density (LSBMD) according to P2RX7 rs3751143 genotype, AA (n = 36) vs AC (n = 19) vs CC (n = 3), in female high-level endurance runners. Higher TBMD but not LBMD in AA than AC genotypes. No differences between genotypes for LSBMD. Error bars denote standard deviation

No other SNPs, individually or collectively as part of a TGS were associated with any bone phenotypes in men or women after multiple testing correction (corrected P ≥ 0.215; Figs. 3 and 4).

Fig. 3
figure3

Mean total bone mineral density (TBMD), leg bone mineral density (LBMD) and lumbar spine bone mineral density (LSBMD) according to total genotype score (TGS) group, < 55 (n = 7) vs 60–70 (n = 32) vs > 75 (n = 6), in male high-level endurance runners. No differences between groups defined by TGS for any bone phenotype in male runners. Error bars denote standard deviation

Fig. 4
figure4

Mean total bone mineral density (TBMD), leg bone mineral density (LBMD) and lumbar spine bone mineral density (LSBMD) according to total genotype score (TGS) group, < 55 (n = 13) vs 60–70 (n = 34) vs > 75 (n = 11), in female high-level endurance runners. No differences between groups defined by TGS for any bone phenotype in female runners. Numbers within bars are number of runners per TGS group and error bars denote standard deviation

Full results from analyses of genotype and bone phenotypes in runners are provided in Tables 1 and 2 in supplementary material.

Genotype-cohort interactions

A genotype-cohort interaction was evident for WNT16 rs3801387 and all bone phenotypes in men for both the additive (TBMD, P = 0.057; LBMD, P = 0.032; LSBMD, P = 0.042; T-score, P = 0.052 and Z-score, P = 0.045; Table 1 in Supplementary Material) and recessive analysis models (TBMD, P = 0.020; LBMD, P = 0.009; LSBMD, P = 0.021; T-score, P = 0.040; Z-score, P = 0.015). Male runners with WNT16 rs3801387 AA genotype in comparison to their AA genotype non-athlete counterparts, had lower values for TBMD, T-score and Z-score (P ≤ 0.014), as well as LSBMD, where the largest difference was shown (14%; P < 0.001; Fig. 5). In addition, WNT16 rs3801387 AG + GG genotypes had 5% higher LBMD than AG + GG non-athletes (P = 0.049). There were no other genotype-cohort interactions in men, and no genotype-cohort interactions in women.

Fig. 5
figure5

Mean lumbar spine bone mineral density (LSBMD) in male runners and non-athletes according to WNT16 rs3801387 genotype (AA vs AG vs GG). Significant interaction between runner/non-athlete status and genotype (P = 0.042), with notably lower LSBMD in AA genotype runners than non-athletes (P < 0.001). Numbers within bars are number of runners per genotype group and error bars denote standard deviation

Full genotype-cohort interaction results for the additive model are provided in Tables 1 and 2 in supplementary material.

Discussion

This study is the first to investigate bone phenotypes and BMD-associated genetic variants in high-level endurance runners and non-athletes. We report novel associations between BMD and P2RX7 rs3751143 for high-level endurance runners whilst WNT16 rs3801387 demonstrated genotype-cohort differences in BMD between runners and non-athletes.

In female runners, P2RX7 rs3751143 AA genotypes had higher TBMD, LBMD, T-score and Z-score than AC + CC genotypes. Higher BMD in AA genotypes has been reported previously in non-athletes (Wesselius et al. 2013) whilst lower LSBMD has been shown in CC homozygote osteoporotic women (Husted et al. 2013). The P2RX7 rs3751143 C-allele has been associated with osteoclast apoptosis (Ohlendorff et al. 2007) and reduced bone strength as well as stress fracture incidence in elite athletes (Varley et al. 2016), which indicates that the loss-of-function C-allele may reduce BMD. This is the first study to investigate the association of P2RX7 rs3751143 on these bone phenotypes in an endurance runner cohort, suggesting that possessing a C allele may also negatively impact BMD, similarly to non-athletes. The differences we observed in measured bone phenotypes (except LSBMD) between AA and AC + CC genotypes in runners indicates that possessing the AA genotype is beneficial for BMD, particularly at sites where a greater volume of mechanical loading is occurring. Consequently, P2RX7 rs3751143 may have a greater influence on bone at certain points across the lifespan via interaction with mechanical loading and is thus particularly pertinent when enhancing bone mass in childhood or using exercise to combat BMD loss during ageing. Moreover, some SNPs may have a greater influence on particular bone components. For example, P2RX7 rs1718119 has been associated with cortical thickness whilst sclerostin (SOST) rs1877632 has been associated with trabecular density (Varley et al. 2018). Cortical and trabecular bone adaptation and loss occur at differing rates (Riggs et al. 2008) and thus genotype-dependent differences may influence these phenotypes distinctly.

No significant associations were observed for the TNFRSF11A rs3018362 or TNFRSF11B rs4355801 SNPs. Both SNPs are reported to influence bone metabolism and BMD within non-athlete populations (Richards et al. 2008; Albagha et al. 2010) but appear unrelated to bone phenotypes within high-level endurance runners. Similarly, AXIN1 rs9921222, COMT rs4680, LRP5 rs3736228 and VDR rs2228570 genotypes did not differ for any bone phenotypes in the current study, but these observations are in contrast with previous literature that has shown potential gene-physical activity/mechanical loading interactions with BMD (Nakamura et al. 2002a; Mitchell et al. 2016). These contrasting findings may be due to inconsistencies in the measurement of physical activity used (e.g. questionnaires), as well as specific genotype-dependent differences in response to various types of mechanical loading. For example, Nakamura et al. (2002a, b) investigated handball, volleyball and jumping athletes, all of whom require movements of high-impact loading and forces that are multi-directional in nature, and are thus very different from the lower impact and cyclical movements completed in endurance running. Consequently, the outcomes of potential gene-mechanical loading interactions may differ.

A genotype-cohort interaction was observed for WNT16 rs3801387 and all bone phenotypes in men for both the additive and recessive analysis models. Male runners with AA genotype possessed lower TBMD, LSBMD, T-score and Z-score than AA genotype non-athletes, whilst AG + GG genotype runners had higher LBMD than their non-athlete counterparts. Interestingly, AA genotype was most advantageous for BMD in non-athletes but most detrimental for runners, with higher values for all bone phenotypes observed in the AG + GG genotype group. Accordingly, the A allele has been previously associated with lower LSBMD, femoral neck BMD and osteoporotic fracture in non-athletes (Estrada et al. 2012). Furthermore, this SNP also interacts with physical activity and LSBMD in children (Mitchell et al. 2016). These findings suggest that possessing WNT16 rs3801387 AA genotype may have a greater impact on BMD in runners than non-athletes, particularly at sites where less loading occurs. Wnt16 is a key regulator of osteoblast-to-osteoclast communication and targeted disruption of Wnt16 in mice results in a 27% loss in bone size and 43–61% loss in bone strength (Zheng et al. 2012). Wnt16 expression is also reported to be influenced by oestrogen receptor signalling in a sex-specific manner with age in mouse tibia (Todd et al. 2015). Oestrogen deficiency has been reported to decrease Wnt16 expression whilst oestrogen replacement increased Wnt16 expression in mouse cortical bone (Alam et al. 2017). Moreover, low oestradiol levels have been reported and associated with reduced BMD in male endurance runners previously (Ackerman et al. 2012). Consequently, it could be hypothesised that interactions between WNT16 variants (and subsequent Wnt16 expression) and low oestradiol levels in male runners could negatively impact BMD.

No further genotype-cohort interactions for any bone phenotypes were observed in either men or women. The findings for TNFRSF11A rs3018362 and P2RX7 rs3751143 SNPs are therefore in agreement with Varley et al. (2018) who showed no genotype-by-time interactions on bone phenotypes following completion of a 12-week training programme in academy football players.

Conclusion

This study is the first to investigate bone phenotypes and BMD-associated genetic variants in high-level endurance runners and non-athletes. We report novel associations between P2RX7 rs3751143 genotype and BMD in female runners, whilst differences in BMD between male runners and non-athletes with the same WNT16 rs3801387 genotype existed, highlighting a potential genetic interaction with factors common in endurance runners, such as mechanical loading. These findings contribute to our knowledge of genetic associations with BMD and improve our understanding of why some runners have lower BMD than others. Independent replication plus identifying other relevant genetic variations could lead towards more personalised exercising programming, partly based upon genetic information, to manage injury risk and thus improve health and performance in endurance runners.

Data availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Abbreviations

ANOVA:

Analysis of variance

BMD:

Bone mineral density

DXA:

Dual-energy X-ray absorptiometry

GWAS:

Genome-wide association study

SNP:

Single nucleotide polymorphism

TGS:

Total genotype score

References

  1. Ackerman KE, Skrinar GS, Medvedova E et al (2012) Estradiol levels predict bone mineral density in male collegiate athletes: a pilot study. Clin Endocrinol (oxf) 76:339–345

    CAS  Article  Google Scholar 

  2. Alam I, Reilly AM, Alkhouli M et al (2017) Bone mass and strength are significantly improved in mice overexpressing human WNT16 in osteocytes. Calcif Tissue Int 100:361–373

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  3. Albagha OME, Visconti MR, Alonso N et al (2010) Genome-wide association study identifies variants at CSF1, OPTN and TNFRSF11A as genetic risk factors for Paget’s disease of bone. Nat Genet 42:520

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  4. Amorim T, Durães C, Machado JC et al (2018) Genetic variation in Wnt/β-catenin and ER signalling pathways in female and male elite dancers and its associations with low bone mineral density: a cross-section and longitudinal study. Osteoporos Int 29:2261–2274

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  5. Baron R, Kneissel M (2013) WNT signaling in bone homeostasis and disease: from human mutations to treatments. Nat Med 19:179

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  6. Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B 57:289–300

    Google Scholar 

  7. Deng F, Tan L, Shen H et al (2013) SNP rs6265 regulates protein phosphorylation and osteoblast differentiation and influences BMD in humans. J Bone Miner Res 28:2498–2507

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  8. Eriksson A-L, Suuriniemi M, Mahonen A et al (2005) The COMT val158met polymorphism is associated with early pubertal development, height and cortical bone mass in girls. Pediatr Res 58:71

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  9. Eriksson AL, Mellström D, Lorentzon M et al (2008) The COMT val158met polymorphism is associated with prevalent fractures in Swedish men. Bone 42:107–112

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  10. Estrada K, Styrkarsdottir U, Evangelou E et al (2012) Genome-wide meta-analysis identifies 56 bone mineral density loci and reveals 14 loci associated with risk of fracture. Nat Genet 44:491

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  11. Garcia-Giralt N, Nogués X, Enjuanes A et al (2002) Two new single-nucleotide polymorphisms in the COL1A1 upstream regulatory region and their relationship to bone mineral density. J Bone Miner Res 17:384–393

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  12. Gori F, Lerner U, Ohlsson C, Baron R (2015) A new WNT on the bone: WNT16, cortical bone thickness, porosity and fractures. Bonekey Rep 4:669

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  13. Guo Y, Dong S-S, Chen X-F et al (2016) Integrating epigenomic elements and GWASs identifies BDNF gene affecting bone mineral density and osteoporotic fracture risk. Sci Rep 6:30558

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  14. Hendrickx G, Boudin E, Fijałkowski I et al (2014) Variation in the Kozak sequence of WNT16 results in an increased translation and is associated with osteoporosis related parameters. Bone 59:57–65

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  15. Herbert AJ, Williams AG, Hennis PJ et al (2019) The interactions of physical activity, exercise and genetics and their associations with bone mineral density: implications for injury risk in elite athletes. Eur J Appl Physiol 119:29–47

    PubMed  Article  PubMed Central  Google Scholar 

  16. Husted LB, Harsløf T, Stenkjær L et al (2013) Functional polymorphisms in the P2X7 receptor gene are associated with osteoporosis. Osteoporos Int 24:949–959

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  17. Jin H, Evangelou E, Ioannidis JPA, Ralston SH (2011) Polymorphisms in the 5′ flank of COL1A1 gene and osteoporosis: meta-analysis of published studies. Osteoporos Int 22:911–921

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  18. Kiel DP, Ferrari SL, Cupples LA et al (2007) Genetic variation at the low-density lipoprotein receptor-related protein 5 (LRP5) locus modulates Wnt signaling and the relationship of physical activity with bone mineral density in men. Bone 40:587–596

    CAS  PubMed  Article  Google Scholar 

  19. Krishnan V, Bryant HU, MacDougald OA (2006) Regulation of bone mass by Wnt signaling. J Clin Invest 116:1202–1209

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  20. Lorentzon M, Eriksson AL, Nilsson S et al (2007) Association between physical activity and BMD in young men is modulated by Catechol-O-Methyltransferase (COMT) genotype: the GOOD study. J Bone Miner Res 22:1165–1172

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  21. Mann V, Ralston SH (2003) Meta-analysis of COL1A1 Sp1 polymorphism in relation to bone mineral density and osteoporotic fracture. Bone 32:711–717

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  22. Mitchell JA, Chesi A, Elci O et al (2016) Physical activity benefits the skeleton of children genetically predisposed to lower bone density in adulthood. J Bone Miner Res 31:1504–1512

    PubMed  Article  PubMed Central  Google Scholar 

  23. Moran JM, Rodriguez-Velasco FJ, Roncero-Martin R et al (2014) The relationship between polymorphisms in the vitamin D receptor gene and bone mineral density in postmenopausal women. ISRN Genet 2014:1

    Article  CAS  Google Scholar 

  24. Nakamura O, Ishii T, Ando Y et al (2002a) Potential role of vitamin D receptor gene polymorphism in determining bone phenotype in young male athletes. J Appl Physiol 93:1973–1979

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  25. Nakamura O, Ishii T, Mankyu H et al (2002b) Contribution of vitamin D receptor genotypes to bone mineral density in young male athletes with different impact loading. Eur J Sport Sci 2:1–8

    Article  Google Scholar 

  26. Ohlendorff SD, Tofteng CL, Jensen J-EB et al (2007) Single nucleotide polymorphisms in the P2X7 gene are associated to fracture risk and to effect of estrogen treatment. Pharmacogenet Genomics 17:555–567

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  27. Papageorgiou M, Dolan E, Elliott-Sale KJ, Sale C (2018) Reduced energy availability: implications for bone health in physically active populations. Eur J Nutr 57:847–859

    PubMed  Article  PubMed Central  Google Scholar 

  28. Paternoster L, Ohlsson C, Sayers A et al (2010) OPG and RANK polymorphisms are both associated with cortical bone mineral density: findings from a metaanalysis of the Avon longitudinal study of parents and children and gothenburg osteoporosis and obesity determinants cohorts. J Clin Endocrinol Metab 95:3940–3948

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  29. Pollock N, Grogan C, Perry M et al (2010) Bone-mineral density and other features of the female athlete triad in elite endurance runners: a longitudinal and cross-sectional observational study. Int J Sport Nutr Exerc Metab 20:418–426

    PubMed  Article  PubMed Central  Google Scholar 

  30. Ralston SH, Uitterlinden AG (2010) Genetics of osteoporosis. Endocr Rev 31:629–662

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  31. Richards JB, Rivadeneira F, Inouye M et al (2008) Bone mineral density, osteoporosis, and osteoporotic fractures: a genome-wide association study. Lancet 371:1505–1512. https://doi.org/10.1016/S0140-6736(08)60599-1

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  32. Riggs BL, Melton LJ, Robb RA et al (2008) A population-based assessment of rates of bone loss at multiple skeletal sites: evidence for substantial trabecular bone loss in young adult women and men. J Bone Miner Res 23:205–214

    PubMed  Article  PubMed Central  Google Scholar 

  33. Robinson JA, Chatterjee-Kishore M, Yaworsky PJ et al (2006) Wnt/β-catenin signaling is a normal physiological response to mechanical loading in bone. J Biol Chem 281:31720–31728

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  34. Simonet WS, Lacey DL, Dunstan CR et al (1997) Osteoprotegerin: a novel secreted protein involved in the regulation of bone density. Cell 89:309–319

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  35. Todd H, Galea GL, Meakin LB et al (2015) Wnt16 is associated with age-related bone loss and estrogen withdrawal in murine bone. PLoS ONE 10:e0140260

    PubMed  PubMed Central  Article  CAS  Google Scholar 

  36. Trajanoska K, Rivadeneira F, Kiel DP, Karasik D (2019) Genetics of bone and muscle interactions in humans. Curr Osteoporos Rep 17:86–95

    PubMed  PubMed Central  Article  Google Scholar 

  37. Varley I, Hughes DC, Greeves JP et al (2015) RANK/RANKL/OPG pathway: genetic associations with stress fracture period prevalence in elite athletes. Bone 71:131–136

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  38. Varley I, Greeves JP, Sale C et al (2016) Functional polymorphisms in the P2X7 receptor gene are associated with stress fracture injury. Purinergic Signal 12:103–113

    CAS  PubMed  PubMed Central  Article  Google Scholar 

  39. Varley I, Hughes DC, Greeves JP et al (2018) SNPs in the vicinity of P2X7R, RANK/RANKL/OPG and Wnt signalling pathways and their association with bone phenotypes in academy footballers. Bone 108:179–185

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  40. Warburton DER, Nicol CW, Bredin SSD (2006) Health benefits of physical activity: the evidence. Can Med Assoc J 174:801–809

    Article  Google Scholar 

  41. Wesselius A, Bours MJL, Henriksen Z et al (2013) Association of P2X7 receptor polymorphisms with bone mineral density and osteoporosis risk in a cohort of Dutch fracture patients. Osteoporos Int 24:1235–1246

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  42. Williams AG, Folland JP (2008) Similarity of polygenic profiles limits the potential for elite human physical performance. J Physiol 586:113–121

    CAS  PubMed  Article  PubMed Central  Google Scholar 

  43. Zheng H-F, Tobias JH, Duncan E et al (2012) WNT16 influences bone mineral density, cortical bone thickness, bone strength, and osteoporotic fracture risk. PLoS Genet 8:e1002745

    CAS  PubMed  PubMed Central  Article  Google Scholar 

Download references

Funding

The research was funded by Manchester Metropolitan University.

Author information

Affiliations

Authors

Contributions

Conceptualisation: AJH, GKS, AGW; Methodology: AJH, GKS, AGW; Formal analysis and investigation: AJH; Writing—original draft preparation: AJH; Writing—review and editing: AJH, GKS, AGW, RME, PJH, CS, SHD, SJL; Supervision: GKS, AGW, SHD, RME, CS, PJH. All authors read and approved the manuscript.

Corresponding author

Correspondence to A. J. Herbert.

Ethics declarations

Conflict of interest

The authors have no relevant financial or non-financial interests to disclose.

Ethical approval

The research study was approved by the appropriate Manchester Metropolitan University ethics committee and certify that the study was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.

Consent to participate

Informed consent was obtained from all individual participants included in the study.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Communicated by Philip D Chilibeck.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (DOCX 25 KB)

Supplementary file2 (DOCX 26 KB)

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Herbert, A.J., Williams, A.G., Lockey, S.J. et al. Bone mineral density in high-level endurance runners: Part B—genotype-dependent characteristics. Eur J Appl Physiol (2021). https://doi.org/10.1007/s00421-021-04789-z

Download citation

Keywords

  • Genetics
  • Single-nucleotide polymorphisms
  • Bone mineral density
  • Endurance
  • Marathon
  • Mechanical loading