Skip to main content
Download PDF

Genomic heterozygosity is associated with a lower risk of osteoarthritis

BMC Genomics volume 25, Article number: 85 (2024) Cite this article

Abstract

Background

Genomic heterozygosity has been shown to confer a health advantage in humans and play a protective role in complex diseases. Given osteoarthritis (OA) is a highly polygenic disease, we set out to determine if an association exists between OA and genomic heterozygosity.

Results

End-stage knee and hip OA patients and healthy controls were recruited from the Newfoundland and Labrador (NL) population. The Arthritis Research UK Osteoarthritis Genetics (arcOGEN) consortium database was utilized as a replication cohort. DNA was extracted from blood samples and genotyped. Individual rates of observed heterozygosity (HetRate) and heterozygosity excess (HetExcess) relative to the expected were mathematically derived, and standardized to a z-score. Logistic regression modeling was used to examine the association between OA and HetRate or HetExcess. A total of 559 knee and hip OA patients (mean age 66.5 years, body mass index (BMI) 33.7 kg/m2, and 55% females) and 118 healthy controls (mean age 56.4 years, BMI 29.5 kg/m2, and 59% female) were included in the NL cohort analysis. We found that OA had an inverse relationship with HetRate and HetExcess with odds ratios of 0.64 (95% CI: 0.45–0.91) and 0.65 (95% CI: 0.45–0.93) per standard deviation (SD), respectively. The arcOGEN data included 2,019 end-stage knee and hip OA patients and 2,029 healthy controls, validating our findings with HetRate and HetExcess odds ratios of 0.60 (95% CI: 0.56–0.64) and 0.44 (95% CI: 0.40–0.47) per SD, respectively.

Conclusions

Our results are the first to clearly show evidence, from two separate cohorts, that reduced genomic heterozygosity confers a risk for the future development of OA.

Peer Review reports

Background

Osteoarthritis (OA) is a chronic heterogeneous disease affecting multiple joint tissues and impacting over 500 million people worldwide [1]. It is characterized by a gradual loss of articular cartilage and subchondral bone changes that typically present clinically in patients as joint pain and functional limitations. Despite its high prevalence, its treatment options remain limited, with early therapies aimed at pain management, and end-stage disease treated with total joint replacement. The development of OA is multifactorial, with environmental, lifestyle, and genetic factors contributing to its progression [2, 3]. Further, OA is a highly polygenic disease and its genetic predisposition is still under heavy investigation with over 100 genetic loci found to date [4].

Genetic diversity within Homo sapiens has played a significant role in our survival and evolution. Literature across multiple species, including humans, supports the dogma that increased genetic diversity improves a species’ fitness [5]. From an evolutionary point of view, this diversity has been developed through breeding between genetically diverse populations, creating genetically diverse offspring. Offspring that inherit genetic variants which are favourable to their survival subsequently pass these variants on to their offspring [6, 7]. This association is termed the heterozygosity-fitness correlation, for which the underlying mechanism is still profoundly debated. A common thread among proposed theories is that the more heterogeneity within an individual’s genome, the greater fitness that individual inherits [5]. A classic example is the Major Histocompatibility Complex (MHC), where more heterozygosity in this genomic region subsequently results in increased antibody diversity and a more robust immune system [8].

With the outburst of genome-wide sequencing, other labs have shown that genome-wide heterozygosity plays a crucial role in human health. Multiple studies have demonstrated that higher heterozygosity rates are interconnected with improved long-term health outcomes and potentially confer healthier aging [9, 10]. Given that prior literature has uncovered a link between heterozygosity and chronic diseases, we undertook this study to investigate if this relationship also applies to OA. In this study, we hypothesized that higher heterozygosity would demonstrate a protective role against hip and knee OA compared to disease-free controls derived from the same population.

Results

Characteristics of the NL cohort

This study included 559 OA patients (172 hip, 387 knee) and 118 controls. The OA group comprised of patients that required a total joint replacement, an indicator of end-stage disease. OA patients were found to be older (p < 0.0005) and have a higher average body mass index (BMI) (p < 0.0005) than non-OA controls, while there was no statistically significant difference in sex between OA and controls (Table 1). Additionally, within the OA group, knee OA patients were found to have a higher BMI than hip OA (p < 0.001), but sex and age did not statistically differ between the two joint groups (Table 1). A statistically significant correlation was noted between age, BMI, and heterozygosity (Table 2). Thus, age, sex, and BMI were adjusted for in the subsequent logistic regression models.

Table 1 The characteristics of participants in the NFOAS cohort
Full size table
Table 2 Pearson correlation analysis for age, BMI and heterozygosity rates in NFOAS cohort
Full size table

Heterozygosity rates and OA in the NL cohort

Multivariable logistic regression with adjustments for age, sex, and BMI was performed with OA status as the dependent variable to evaluate the relationship between OA and heterozygosity. Single nucleotide polymorphisms (SNPs) with minor allele frequency (MAF) < 0.01 were excluded from the final analysis as they are less common variants and more likely to be specific to the individual population. First, we examined the relationship between OA and observed heterozygosity (HetRate) and found an inverse relationship with an odds ratio (OR) of 0.64 (p = 0.01; Fig. 1). To determine if this relationship was specific to one joint subgroup, we performed the analyses for hip and knee OA separately. These analyses demonstrated a statistically significant inverse relationship between OA status and HetRate for hip and knee OA with a similar effect size, an OR of 0.64 and 0.59, respectively (Fig. 1).

Fig. 1
figure 1

Association between heterozygosity and OA for NFOAS cohort (SNP MAF > 0.01). Multivariable logistic regression for heterozygosity rate with outcome set to OA status. Odds ratio (OR), 95% confidence interval (95% CI), and p-value with adjustments for age, sex, and BMI

Full size image

As observed heterozygosity rate is influenced by allele frequencies, we next performed multivariable logistic regressions with heterozygote excess (HetExcess), as this variable is independent of allele frequency. Our results showed an inverse relationship between HetExcess and OA status with an OR of 0.65 (p = 0.02, Fig. 1). Separate analyses for hip and knee OA exhibited similar statistical significance, with an OR of 0.64 and 0.61, respectively (Fig. 1).

Next, we explored if this association would remain statistically significant for more common genetic variants in our population. Therefore, we further restricted the SNP MAF to greater than 0.1. We repeated the previous regression analyses with this new dataset and demonstrated that the inverse relationship between OA status and heterozygosity remained statistically significant. This was found for all comparisons and demonstrated a similar effect size to that of SNPs with MAF > 0.01 (Fig. 2).

Fig. 2
figure 2

Association between heterozygosity and OA for NFOAS cohort (SNP MAF > 0.1). Multivariable logistic regression for heterozygosity rate with outcome set to OA status. Odds ratio (OR), 95% confidence interval (95% CI), and p-value with adjustments for age, sex, and BMI

Full size image

Further, we repeated the regression analyses with an additional adjustment for diabetes – one of the comorbidities that we had data on in our cohort. All the observed associations between OA and heterozygosity remained statistically significant (p < 0.05).

Heterozygosity rates and OA in the replication cohort

To replicate and validate our results in an independent population, we utilized the Arthritis Research UK Osteoarthritis Genetics (arcOGEN) genomic database. This is a UK based cohort with 2019 OA patients and 2029 OA-free controls. A simple logistic regression was performed with OA status as the dependent variable to evaluate the relationship between OA and heterozygosity. The arcOGEN analyses validated our results by further showing an inverse relationship between OA and heterozygosity across all analyses. At SNP MAF > 0.01, HetRate and HetExcess ORs were 0.60 (p =  < 0.001) and 0.44 (p =  < 0.001) per standard deviation (SD), respectively (Fig. 3). Subgroup analysis of hip and knee OA separately showed a consistent statistically significant relationship between OA and heterozygosity (Fig. 3). At SNP MAF > 0.1, HetRate and HetExcess ORs were 0.75 (p =  < 0.001) and 0.61 (p =  < 0.001) per SD, respectively (Fig. 4). Subgroup analysis of hip and knee OA separately showed a consistent statistically significant relationship between OA and heterozygosity (Fig. 4).

Fig. 3
figure 3

Association between heterozygosity and OA for arcOGEN cohort (SNP MAF > 0.01). Logistic regression for heterozygosity rate with outcome set to OA status. Odds ratio (OR), 95% confidence interval (95% CI), and p-value

Full size image
Fig. 4
figure 4

Association between heterozygosity and OA for arcOGEN cohort (SNP MAF > 0.1). Logistic regression for heterozygosity rate with outcome set to OA status. Odds ratio (OR), 95% confidence interval (95% CI), and p-value

Full size image

Discussion

The genetics of OA continues to be an actively investigated field of research, with an increasing number of genomic variants being discovered. In this study, we took a unique approach by examining the association between genomic heterozygosity and OA. Previous studies have found that increased heterozygosity is associated with healthier aging, a lower risk of death, and lower health parameters such as blood pressure and cholesterol [9,10,11]. Our data showed a statistically significant but small correlation between genomic heterozygosity and age/BMI. The negative correlation with age was most likely due to the fact that the OA patients were older than the controls in our cohort whereas the positive correlation with BMI might be explained by the evolutionary advantage of fat storage and lower energy expenditure. However, to the best of our knowledge, this was the first study to provide evidence that heterozygosity rates are correlated to the risk of end-stage OA. Indeed, we found that both observed heterozygosity and heterozygosity excess were statistically significantly lower in OA patients across all comparisons in our cohort. Additionally, we were able to validate these finding in a large independent UK cohort. These findings advance the field of OA research by further strengthening the known role of genetics in OA development. It also provides evidence that genetic heterozygosity plays an important role in patients being predisposed to chronic diseases such as OA.

Furthermore, this relationship remained significant for SNPs with MAF > 0.1, signifying that the association with heterozygosity also exists for common alleles. These findings indicate that the mechanism behind the protective role seen for increased heterozygosity is not due to purifying selection against deleterious recessive alleles. Xu et al. demonstrated similar results for the association between healthier aging and increased heterozygosity [9]. Their study categorized SNPs by MAF and found that SNPs with more common allele frequencies had a similar significance to those with lower frequencies. They argued this refutes the notion that increased heterozygosity provides its benefit through compensation for deleterious recessive alleles [9]. Instead, it supports the mechanism of heterozygote advantage or overdominance [9]. The heterozygosity-fitness correlation in our study is likely under-pined by the same mechanism.

While an abundance of genetic variants associated with OA have been detected, research is limited by the fact that the majority of these variants are found within non-coding regions of the genome and have a small effect size [4, 12]. This has led to the prevailing theory that the accumulation of multiple risk alleles contributes to the predisposition of OA later in life and that a threshold must exist that tips an individual into OA. This theory also explains the diversity seen in OA patients, since countless combinations of different genotypic variants could contribute to the phenotypes of OA. This polygenetic nature of OA has made early detection and targeted treatments challenging to establish. In this study, we provided a framework to use genomic heterozygosity to detect individuals at risk of OA, which has the potential to enable earlier interventions. Not a novel idea since Lacaze et al. and other groups have demonstrated that genomic risk scores, calculated using known genetic risk variants, can predict an individual’s risk of developing advanced hip and knee OA [13, 14]. A recent large meta-analysis in Cell also demonstrated that polygenic risk scores have predictive power for the odds of developing OA [4]. However, these models are limited because they only consider genetic variants that have already been discovered. Therefore, it cannot account for OA variants that have yet to be revealed and is only as powerful as the most current literatures on risk loci. While polygenic risk scores and genomic heterozygosity both use differences within the genome to predict an individual’s risk of OA, they are very different approaches. This is the first study to show that genomic heterozygosity has an association with OA and therefore the predict power of this approach requires further research.

This study's strengths include a well phenotyped NL cohort from a similar ethnic background for both cases and controls. The island of Newfoundland is a young founder population with a high degree of genetic and cultural homogeneity, providing a rich resource for investigating complex genetic diseases such as OA [15, 16]. However, while the limited diversity of this population enables a higher signal-to-noise ratio, the same characteristic limits our ability to generalize results to other populations. Therefore, the replication cohort allows us to further strengthen our findings through replication in an independent European cohort. Still we are limited in extrapolating these findings to other populations with different ethnic backgrounds since NL’s ancestry is primarily European in origin [16]. Also, we only included end-stage hip and knee OA patients in this study, and it is well known that joint-specific genetic patterns exist for OA sites [4]. Therefore, we cannot extrapolate these findings to other joint sites impacted by OA such as hand and spinal OA. As well, these findings cannot be generalized to mild or moderate OA patients. Controls were younger than OA patients, and some of them may develop OA later on, which might potentially bias our findings. However, this misclassification can only bias the results towards null, thus, the actual association might be even stronger than what we observed. Additionally, we did not have data on other comorbidities and factors such as socioeconomic status and genu varum/valgum which might have confounded our findings. However, this is less likely as most of those comorbidities do not have strong evidence to support their associated with OA as per a recent review on this topic [17]. Additionally, this is supported by our further analysis with adjustment for diabetes in the NL cohort that did not alter the results. Most importantly, our NL cohort findings were replicated in an independent cohort collected from a different geographic location, indicating the robustness of the findings.

In summary, these results were the first to show a clear association between genomic heterozygosity and the risk of OA. In particular, our group provides evidence from two separate population cohorts that reduced genomic heterozygosity confers a risk for the future development of OA. This supports our original hypothesis that higher heterozygosity would demonstrate protection against hip and knee OA. Additionally, we propose that the likely mechanism behind our findings is due to heterozygote advantage or overdominance. Future work is required to validate this association in other large OA genomic databases and for OA sites other than the hip and knee. Specifically, further research on non-weight bearing joint such as hand and shoulder, where BMI has less influence, is required. Furthermore, while this study provides evidence for a link between OA and heterozygosity, it does not provide any quantitative measures for the level of heterozygosity that confers a protective effect. Therefore, future predictive modelling for heterozygosity with large OA cohorts that include different stage of disease is required to delineate this model's predictive efficacy further. Once a predictive model is developed then prospective studies can determine if genomic heterozygosity can effectively predict individuals that will develop OA. Once validated, this can lead to early detection and earlier interventions for individuals at risk of OA.

Method

Study participants

The initial analysis utilized two cohorts with participants’ demographic and genome-wide genotype data, referred to as the NL cohort. One contained known end-stage OA patients, and the other healthy OA-free controls, both derived from the population of Newfoundland & Labrador, Canada. Self-reported OA-free controls were selected from the Complex Diseases in the Newfoundland Population; Environment and Genetics (CODING) study and OA cases included were from the Newfoundland Osteoarthritis Study (NFOAS). Inclusion criteria for the CODING study has been previously published and included: 1) a third generation or greater Newfoundlander, 2) age of 20–79 years old at the time of recruitment, 3) no serious metabolic, cardiovascular, or endocrine disorder and, 4) not pregnant at the time of the study [18,19,20]. The characteristics of the NFOAS study have been previously published by our group and included 1086 patients who underwent total knee or hip replacement between November 2011 and September 2017 largely due to OA [21, 22]. Only patients with a primary diagnosis of OA, as defined by the American Rheumatology College’s criteria and judgment of an attending orthopedic surgeon, were included in this study. Both studies were approved by the Health Research Ethics Authority of Newfoundland and Labrador (HREB; NFOAS—# 2011.311 & CODING—#10.33), and written informed consent was obtained from all participants. All methods were performed in accordance with the relevant regulations of this authority.

Genome-wide genotype data from the arcOGEN consortium database was utilized as a replication cohort [23]. We used data on 2029 female controls and 2019 female end-stage OA (hip or knee) patients in our replication because only female controls were available to us. The inclusion and exclusion criteria of the arcOGEN has been previously published [23]. All OA cases were unrelated and were European in origin, while controls were from the TwinsUK cohort [24]. Diagnosis of primary OA (hip or knee) was defined by a clinical level of disease that required total joint replacement. Patients with concurrent diagnoses of hip and knee OA were included in both the hip and knee subgroup analyses. All participants gave informed written consent and the study protocols were reviewed and approved by the Oxfordshire Research Ethics Committee with reference 07/H0606/150.

Patient and public involvement

Patients were involved in our study as study participants, but neither they nor the public were involved in any of the stages of our research.

Genome-wide genotyping

For the NL cohort, DNA was extracted from whole blood samples collected from all study participants by using standard protocol. Genome-wide genotyping was done by using either the Illumina Human Omni2.5–8 (Illumina, San Diego, CA, USA) microarray platform which genotypes 2.5 million SNPs or the Human Global Diversity Array (Illumina, San Diego, CA, USA) which genotypes 1.9 million SNPs at The Centre for Applied Genomics in Toronto (http://www.tcag.ca/index.html). The genotyped data underwent a strict quality control (QC) procedure including removing SNPs with MAF < 0.01, missing genotype rate > 5%, and out of Hardy–Weinberg Equilibrium, as well as removing individuals with discordant sex information and duplicates. Population structure was examined with the 1000 Genomes Project data as references and outliers were removed. QCed genotype data were then imputed against 1000 Genome Project phase 3 data by Sanger Imputation Server (https://imputation.sanger.ac.uk/?about=1) which generated data on a total of 81,706,022 SNPs across the entire genome. The imputed data underwent QC procedures as described above and were then used for subsequent analyses.

For the arcOGEN cohort, genome-wide genotyping was done by Illumina Human 610-Quad BeadChips (Illimina, San Diego, CA, USA) with a protocol previously described [23]. The same QC procedures as the NL cohort was performed on the directly genotyped data and then imputed by the Sanger Imputation Server with the same parameters as our discovery cohort.

Heterozygosity

The total number of heterozygote genotypes, the total number of homozygote genotypes for alternative alleles, and the expected heterozygosity rates from the imputed genomic data were calculated using Plink Version 1.9 [25]. HetRate was mathematically defined as the total number of heterozygote genotypes divided by the total number of homozygote genotypes for alternative alleles (Fig. 5A). HetExcess was mathematically defined as the observed heterozygosity subtract the expected heterozygosity and then divided by the expected heterozygosity (Fig. 5B). Both HetRate and HetExcess were then standardized to z-scores prior to statistical analysis. Both HetRate and HetExcess were calculated twice, one with all SNPs with MAF > 0.01 and the other with SNPs with MAF > 0.1. There were various approaches used in the literature to represent heterozygosity. We chose these two methods because the observed heterozygosity rate provides an assessment of the current genetic diversity within a population whereas heterozygosity excess can offer insight into the historical and demographic processes that have shaped this diversity.

Fig. 5
figure 5

Equation utilized to calculate observed heterozygosity rate and heterozygosity excess

Full size image

Statistical analysis

For the NL cohort multivariable logistic regression modeling was utilized to investigate the association between heterozygosity and OA. Given age and BMI were found to correlate with heterozygosity, they were included in the regression along with sex to control for potential confounding effects. For the replication cohort simple logistic regression modeling was utilized to investigate the association between heterozygosity and OA. For this cohort all participants were female and data on age and BMI were unavailable. Separate regressions were run for HetRate and HetExcess, at the level of SNP MAF > 0.01 and > 0.1, respectively. Regressions were repeated to examine the association of heterozygosity with hip and knee OA separately. ORs and their 95% confidence intervals were extracted from each logistic analysis to illustrate the correlation between OA and genetic heterozygosity. ORs were calculated as the odds of developing OA for each SD increase in HetRate/HetExcess. An OR of < 1 in our model signifies that as genomic heterozygosity increases the odds of developing OA decrease. The significance level for all regression analyses was defined by an alpha level of 0.05 and all analyses were performed using statistical software SPSS version 28.0.1.0.

Availability of data and materials

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

References

  1. Long H, Liu Q, Yin H, Wang K, Diao N, Zhang Y, Lin J, Guo A. Prevalence Trends of Site-Specific Osteoarthritis From 1990 to 2019: Findings From the Global Burden of Disease Study 2019. Arthritis Rheumatol. 2022;74(7):1172–83.

    Article  PubMed  PubMed Central  Google Scholar 

  2. Vina ER, Kwoh CK. Epidemiology of osteoarthritis: literature update. Curr Opin Rheumatol. 2018;30(2):160–7.

    Article  PubMed  PubMed Central  Google Scholar 

  3. Fernandez-Moreno M, Rego I, Carreira-Garcia V, Blanco FJ. Genetics in osteoarthritis. Curr Genomics. 2008;9(8):542–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Boer CG, Hatzikotoulas K, Southam L, Stefansdottir L, Zhang Y, Coutinho de Almeida R, Wu TT, Zheng J, Hartley A, Teder-Laving M, et al. Deciphering osteoarthritis genetics across 826,690 individuals from 9 populations. Cell. 2021;184(18):4784-4818e4717.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Charlesworth D, Willis JH. The genetics of inbreeding depression. Nat Rev Genet. 2009;10(11):783–96.

    Article  CAS  PubMed  Google Scholar 

  6. Birchler JA, Yao H, Chudalayandi S, Vaiman D, Veitia RA. Heterosis. Plant Cell. 2010;22(7):2105–12.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  7. Timberlake WE. Heterosis. In: Maloy S, Hughes K, editors. Brenner’s Encyclopedia of Genetics. 2nd ed. London: Academic; 2013. p. 451–3.

    Chapter  Google Scholar 

  8. Penn DJ, Damjanovich K, Potts WK. MHC heterozygosity confers a selective advantage against multiple-strain infections. Proc Natl Acad Sci U S A. 2002;99(17):11260–4.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Xu K, Kosoy R, Shameer K, Kumar S, Liu L, Readhead B, Belbin GM, Lee HC, Chen R, Dudley JT. Genome-wide analysis indicates association between heterozygote advantage and healthy aging in humans. BMC Genet. 2019;20(1):52.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Campbell H, Carothers AD, Rudan I, Hayward C, Biloglav Z, Barac L, Pericic M, Janicijevic B, Smolej-Narancic N, Polasek O, et al. Effects of genome-wide heterozygosity on a range of biomedically relevant human quantitative traits. Hum Mol Genet. 2007;16(2):233–41.

    Article  CAS  PubMed  Google Scholar 

  11. Bihlmeyer NA, Brody JA, Smith AV, Lunetta KL, Nalls M, Smith JA, Tanaka T, Davies G, Yu L, Mirza SS, et al. Genetic diversity is a predictor of mortality in humans. BMC Genet. 2014;15:159.

    Article  PubMed  PubMed Central  Google Scholar 

  12. Beier F. The impact of omics research on our understanding of osteoarthritis and future treatments. Curr Opin Rheumatol. 2023;35(1):55–60.

    Article  PubMed  Google Scholar 

  13. Lacaze P, Wang Y, Polekhina G, Bakshi A, Riaz M, Owen A, Franks A, Abidi J, Tiller J, McNeil J, et al. Genomic Risk Score for Advanced Osteoarthritis in Older Adults. Arthritis Rheumatol. 2022;74(9):1480–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  14. Sedaghati-Khayat B, Boer CG, Runhaar J, Bierma-Zeinstra SMA, Broer L, Ikram MA, Zeggini E, Uitterlinden AG, van Rooij JGJ, van Meurs JBJ. Risk Assessment for Hip and Knee Osteoarthritis Using Polygenic Risk Scores. Arthritis Rheumatol. 2022;74(9):1488–96.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Rahman P, Jones A, Curtis J, Bartlett S, Peddle L, Fernandez BA, Freimer NB: The Newfoundland population: a unique resource for genetic investigation of complex diseases. Hum Mol Genet 2003, 12 Spec No 2:R167–172.

  16. Zhai G, Zhou J, Woods MO, Green JS, Parfrey P, Rahman P, Green RC. Genetic structure of the Newfoundland and Labrador population: founder effects modulate variability. Eur J Hum Genet. 2016;24(7):1063–70.

    Article  PubMed  Google Scholar 

  17. Allen KD, Thoma LM, Golightly YM. Epidemiology of osteoarthritis. Osteoarthritis Cartilage. 2022;30(2):184–95.

    Article  CAS  PubMed  Google Scholar 

  18. Sun G, French CR, Martin GR, Younghusband B, Green RC, Xie YG, Mathews M, Barron JR, Fitzpatrick DG, Gulliver W, et al. Comparison of multifrequency bioelectrical impedance analysis with dual-energy X-ray absorptiometry for assessment of percentage body fat in a large, healthy population. Am J Clin Nutr. 2005;81(1):74–8.

    Article  CAS  PubMed  Google Scholar 

  19. Martin GR, Loredo JC, Sun G. Lack of association of ghrelin precursor gene variants and percentage body fat or serum lipid profiles. Obesity (Silver Spring). 2008;16(4):908–12.

    Article  CAS  PubMed  Google Scholar 

  20. Aref-Eshghi E, Hurley O, Sun G, Simms A, Godwin M, Duke P, Araee M, Mahdavian M, Asghari S. Genetic associations in community context: a mixed model approach identifies a functional variant in the RBP4 gene associated with HDL-C dyslipidemia. BMC Med Genet. 2018;19(1):205.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Zhang W, Likhodii S, Zhang Y, Aref-Eshghi E, Harper PE, Randell E, Green R, Martin G, Furey A, Sun G, et al. Classification of osteoarthritis phenotypes by metabolomics analysis. BMJ Open. 2014;4(11):e006286.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Werdyani S, Liu M, Zhang H, Sun G, Furey A, Randell EW, Rahman P, Zhai G. Endotypes of primary osteoarthritis identified by plasma metabolomics analysis. Rheumatology (Oxford). 2021;60(6):2735–44.

    Article  CAS  PubMed  Google Scholar 

  23. arc OC, arc OC, Zeggini E, Panoutsopoulou K, Southam L, Rayner NW, Day-Williams AG, Lopes MC, Boraska V, Esko T, et al. Identification of new susceptibility loci for osteoarthritis (arcOGEN): a genome-wide association study. Lancet. 2012;380(9844):815–23.

    Article  Google Scholar 

  24. TwinsUK – The biggest twin registry in the UK for the study of ageing related diseases. https://twinsuk.ac.uk/. Accessed 10 Jun 2023.

  25. PLINK 1.9. https://www.cog-genomics.org/plink/. Accessed 20 Feb 2023.

Download references

Acknowledgements

We thank all the study participants who made this study possible.

Funding

The study was funded by Canadian Institutes of Health Research (CIHR) [132178, 143058, 153298, 175015, and MOP192552], The Research and Development Corporation of Newfoundland and Labrador [Grant number 5404.1423.102], and The Memorial University of Newfoundland Medical Research Fund. The arcOGEN study was funded by a special purpose grant from Arthritis Research UK (grant 18030). TwinsUK is funded by the Wellcome Trust, Medical Research Council, Versus Arthritis, European Union Horizon 2020, Chronic Disease Research Foundation (CDRF), Zoe Ltd, the National Institute for Health and Care Research (NIHR) Clinical Research Network (CRN) and Biomedical Research Centre based at Guy’s and St Thomas’ NHS Foundation Trust in partnership with King’s College London. These funding bodies had no role in the conceptualization, design, data collection, analysis, decision to publish, or preparation of the manuscript.

Author information

Authors and Affiliations

  1. Human Genetics and Genomics, Division of Biomedical Sciences, Faculty of Medicine, Memorial University of Newfoundland, St. John’s, NL, A1B 3V6, Canada

    Robert Gill, Ming Liu & Guangju Zhai

  2. Discipline of Medicine, Faculty of Medicine, Memorial University of Newfoundland, St. John’s, NL, Canada

    Guang Sun & Proton Rahman

  3. Division of Orthopaedic Surgery, Faculty of Medicine, Memorial University of Newfoundland; Office of the Premier, Government of Newfoundland & Labrador, St. John’s, NL, Canada

    Andrew Furey

  4. Department of Twin Research & Genetic Epidemiology, Faculty of Life Sciences & Medicine, King’s College London, London, UK

    Tim Spector

Authors
  1. Robert Gill
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Ming Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Guang Sun
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Andrew Furey
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Tim Spector
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Proton Rahman
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Guangju Zhai
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

R.G. analyzed and interpreted the patient data, and was a major contributor in writing the manuscript. M.L. had a major role in data collection and analysis. G.S. had a role in study conception and data acquisition. A.F. had a role in study conception and data acquisition. P.R. had a role in study conception and data acquisition. T.S. provided control samples in the replication cohort and critical comments on interpretation of the results and manuscript. G.Z. had a role in study conception and design, along with data analysis and interpretation. G.Z. (gzhai@mun.ca) takes responsibility for the integrity of this work as a whole, from inception to the finished article. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Guangju Zhai.

Ethics declarations

Ethics approval and consent to participate

Both NFOAS and CODING study were approved by the Health Research Ethics Authority of Newfoundland and Labrador (HREB; NFOAS—# 2011.311 & CODING—#10.33). The arcOGEN was approved by the Oxfordshire Research Ethics Committee with reference 07/H0606/150. Written informed consent was obtained from all participants.

Consent for publication

Not applicable.

Competing interests

Author P.R. discloses they are a paid consultant for AbbVie, BMS, Eli Lilly, Janssen, Novartis, Pfizer, and UCB. Research support as a PI from AbbVie, Pfizer, Novartis, and Janssen. Payment or honoraria from AbbVie, BMS, Eli-Lilly, Jansen, Pfizer, Novartis, UCB. Leadership or fiduciary role in the Canadian Academy of Health Sciences and Canadian Council of Academies.

Author G.Z. discloses payment or honoraria from The Canadian Rheumatology Society 2022 and 7th World Congress on Controversies, Debates & Consensus on Bone, Muscle & Joint Diseases 2019.

All other authors have no financial conflict of interest to disclose.

Additional information

Publisher’s Note

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

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/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Gill, R., Liu, M., Sun, G. et al. Genomic heterozygosity is associated with a lower risk of osteoarthritis. BMC Genomics 25, 85 (2024). https://doi.org/10.1186/s12864-024-10015-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12864-024-10015-9

Keywords

BMC Genomics

ISSN: 1471-2164

Contact us