May/June 2020

Genetics of Osteoporosis
By Jamie Santa Cruz
Today’s Geriatric Medicine
Vol. 13 No. 3 P. 10

Genetics research has significant implications for clinical care.

In the last few years, a series of large, well powered studies have dramatically increased scientists’ understanding of specific genetic factors that influence osteoporosis. This information is highly significant because it influences the ability to predict who’s at risk for the disease and it also can enable better treatment of those who already have osteoporosis. So what exactly is known about the genetics of osteoporosis, and how can that information aid clinical practice? Here, Today’s Geriatric Medicine reviews the latest research.

Bone Mineral Density and Fracture Risk: Highly Heritable Traits
Osteoporosis is characterized by two features: low bone mineral density (BMD) and increased risk of fracture. Both of these features have a strong genetic component. For BMD, 60% to 80% of phenotypic variance is explained by genes. Osteoporotic fracture has a lower heritability, but it is still high: 50% to 70% of the risk of osteoporotic fracture is attributable to genetic factors.1 Other bone parameters such as geometry and even bone microstructure are also known to be highly heritable.2,3

There are a number of forms of bone fragility that are monogenic, meaning they’re caused by mutation of a single gene influencing skeletal biology. Examples of such conditions include Paget’s disease and osteogenesis imperfecta. But these diseases are relatively rare, and monogenic forms of bone fragility explain only a very small portion of osteoporosis risk overall. Most cases of osteoporosis are polygenic—meaning they have a relatively complex etiology stemming from the interaction of multiple genes.2 Environmental factors, such as physical inactivity, poor nutrition, smoking, or alcohol consumption, also play a role in increasing risk.

The Age of GWAS: A Revolution in the Study of the Genetics of Osteoporosis
Early studies on the genetics of osteoporosis were dominated by candidate gene-association studies. Candidate-gene approaches focused on a limited number of prespecified genes that scientists suspected might have an impact on osteoporosis. Early candidate genes included ESR1 (encoding estrogen receptor), VDR (encoding vitamin D3 receptor), and COL1A1 (encoding collagen alpha-1(I) chain), among others.4 The key weakness of candidate-gene studies is that they focus on specific hand-picked genes, meaning that such studies may completely miss genes that have a large effect size on osteoporosis risk just because researchers don’t know to look at those genes.

In the last few years, however, use of genomewide association studies (GWAS) has transformed the study of the genetics of disease in general and osteoporosis in particular. Unlike candidate-gene association studies, GWAS examine genetic variants across the entire genome.

In 2008, the first meaningful GWAS related to osteoporosis used a sample of 8,557 participants to identify two genes—LRP5 and TNFRSF11B (OPG)—that were associated with lumbar spine and femoral neck BMD in the general population. The first of these genes was also shown to be associated with osteoporotic fractures.5 Later that same year, a second GWAS with a sample of 13,786 individuals from Iceland identified three new loci associated with both BMD and osteoporotic fracture.6 The same research group in Iceland later expanded sample size to 15,375 and identified several additional variants.7 One of the loci identified in the second Icelandic study, SP7 (osterix), was then confirmed in the first GWAS of BMD in children.8

These early GWAS sample sizes were comparatively small—limited to about 15,000 participants. This meant relatively low statistical power, which in turn meant that these earlier studies could only identify loci that had relatively large effect sizes.1

As GWAS sample sizes became larger, the analyses could pick up on loci with smaller effect sizes, and there were dramatic increases in the discovery of loci associated with BMD.

Of particular value was research released by the GEnetic Factors for OSteoporosis (GEFOS) Consortium, which began publishing a series of GWAS meta-analyses starting in 2009. The first of these meta-analyses included 19,195 participants and identified 13 new loci associated with BMD.9 Three years later, the consortium published another meta-analysis that included 32,961 participants and identified 32 new loci (in addition to confirming loci that had already been identified). Fourteen of these new loci were also associated with osteoporotic fractures.10

In 2018, the GEFOS Consortium published yet another meta-analysis with 66,628 participants. This study identified 80 total loci associated with total-body BMD, of which 36 were previously unidentified.11 It had an added significance in that it included an age-specific analysis. The results showed that one locus in particular (ESR1) had an age-dependent effect on BMD, but most identified loci do not appear to have an age-dependent effect. This is important because it suggests that most of the genes associated with BMD affect peak BMD acquisition and that these effects are observable throughout the life course.2

A few other GWAS have begun to look at sex-specific effects. One specific variant at FAM9B has shown an effect in BMD on men only, and several loci have been identified as being significantly associated with BMD, fracture risk, or other bone parameters specifically in women.12-14

The Biobank Project: A Dramatic Step Forward in Sample Sizes
Although the GEFOS Consortium studies brought a vast increase in the number of loci that have been associated with BMD, the combination of all the known loci was still able to explain only a small amount of variation in BMD—about 6%.1 However, that changed as a result of the UK Biobank project, which has an enormous number of participants from the general population of the United Kingdom and which includes data on the genotype of each participant as well as their BMD (measured via ultrasound of the heel bone).

Since 2017, there have been several new GWAS of BMD using data from the UK Biobank project.12,15,16 The UK Biobank studies involve approximately 420,000 participants and are the most significant studies to date on the genetics of osteoporosis.15,16 These studies have been able to identify a total of approximately 500 to 600 loci associated with heel BMD, of which 301 are new loci that had not been identified previously. With these new loci, it’s now possible to explain approximately 20% to 25% of phenotypic variance in BMD.1,16

According to David Karasik, PhD, an associate scientist in the Hinda and Arthur Marcus Institute for Aging Research, as well as an associate professor in Bar Ilan University (Israel), the major weakness of the UK Biobank is that it’s limited to individuals from a single country. In addition, participants in the UK Biobank are relatively young (45 to 65 years of age, which means that the studies using these data don’t give much insight into osteoporosis in the geriatric population), and they are mostly white. Research into ethnic minorities “is totally underdeveloped—we don’t have Hispanics, we don’t have blacks,” Karasik says. “We really cannot apply the genetic findings to ethnic minorities.”

Still, the findings from this data bank underscore the sheer number of genetic factors that influence osteoporosis. In the earlier stages of research on the genetics of osteoporosis, the assumption was that there would be only a few genes that would determine osteoporosis risk, but the UK Biobank studies demonstrate that the actual number is quite large. This is striking, given that few other diseases or traits have so many identified loci that are associated at genomewide significance.15

Whole-Genome Sequencing: The Push to Identify Less Common Variants
GWAS analyses have dramatically increased our understanding of genes that play a role in BMD. However, the variants identified through GWAS still explain only about 20% of the variance in BMD. This is likely due in part to the fact that GWAS generally focus on variants in the genome that are comparatively common, whereas rare variants are known to have larger effect sizes than do common variants.2 It’s likely that less common variants would account for some of the genetic basis of BMD that’s currently unexplained.17

To help uncover less common variants that could affect BMD and fracture, a number of recent studies have begun to rely on whole-genome sequencing rather than on GWAS. “When we are doing whole-genome sequencing, we are literally testing every genetic variant in the subject’s genome,” says Hui Shen, PhD, an associate professor in the School of Public Health and Tropical Medicine at Tulane University, as well as associate director of the university’s Center for Bioinformatics and Genomics. “So we can pick up not only those common variants that are highly frequent in the population but also so-called rare variants.”

A strategy that’s been employed in some recent research is to use whole-genome sequencing to examine individuals with either very low or very high BMD to identify rare variants that may have large effect sizes. Such studies have identified several rare mutations at LGR4 and at COL1A2 that are all associated with low BMD.18,19 One particularly powerful whole-genome sequencing study identified a noncoding variant at EN1 that also has large effects on BMD.20

Zeroing in on Fracture Risk
So far, most genetic research has focused on BMD rather than on fracture risk, on the grounds that it’s difficult to recruit large numbers of study participants with fracture. However, the genes that affect BMD are not necessarily the same genes that influence fracture risk. “BMD is a strong risk predictor for osteoporotic fracture, [and] both BMD and osteoporosis have strong heritability,” Shen says. “But the genetic correlation—meaning the shared genetic architecture between them—is relatively low. So that tells [me] that in the future we should emphasize more studies [that look directly at] osteoporotic fracture.”

Although such studies are limited, there are a few. The largest and most significant GWAS focusing on fracture in particular is a 2018 meta-analysis of data from 25 cohorts across Europe, the United States, Asia, and Australia. The participants in this meta-analysis included 37,857 individuals with fracture plus 227,116 controls. The study identified 15 genes associated with fracture risk, all of which were also associated with BMD. These results underscore the clear effect of BMD on fracture risk, although it is noteworthy that the identified genes had a smaller effect on fracture risk than they did on BMD.21

Next Steps
The huge advance in the research of the last few years is that scientists have identified hundreds of specific loci associated with BMD. But a key gap in current research is that scientists don’t necessarily know which gene each locus maps to, and in most cases they also don’t know the functions of those loci.17 “One of the limitations of GWAS is that it just gives you a location,” says Charles Farber, PhD, an associate professor of public health sciences in the University of Virginia School of Medicine. “It’s almost like putting a pin down on Google Maps, where you’re marking that location as a region that’s interesting. If we mark a particular region by GWAS, we know that there’s variation there, but it doesn’t tell us anything about what’s going on mechanistically.” According to Farber, the next step is to fill in those gaps by figuring out which variants regulate which genes, as well as the mechanisms by which those genes impact bone. “That’s important downstream as we think about trying to translate the information that we have from GWAS into both a better understanding of bone biology as well as using that to identify potential new therapeutic targets,” Farber says.

Another key limitation of the existing research is that it has focused very heavily on BMD to the exclusion of other bone traits such as bone size and quality. (Bone quality refers to several properties such as tissue strength, fracture toughness, and fatigue strength.) The focus on BMD is understandable in a sense, both because BMD is the single biggest predictor of fracture risk and because it is the easiest bone trait to measure. Yet many people with low BMD do not experience fracture, and many people with normal BMD do fracture. Thus there’s a need for genetic research focusing on other bone traits as well.17

Relevance for Clinical Practice
The new findings on the genetic basis for osteoporosis, BMD, and fracture risk have key implications for clinical practice.

To begin with, genetic information can be used to gauge which individuals are at risk of osteoporosis and fracture. Stuart Kim, PhD, an emeritus professor of developmental biology at Stanford University, is one of the researchers engaged in applying genetics to risk prediction models. Kim, who’s the author of one of the recent UK Biobank studies, has not only identified new variants associated with BMD but also used that genetic information to develop a composite risk score for low BMD (and by extension osteoporosis and fracture).16 When he tested that risk score on a population of approximately 50,000 people, its ability to accurately predict which had low BMD turned out to be remarkably high. “It gets it right 78% of the time,” Kim says. “By comparison, this is more informative than breast cancer testing or Alzheimer’s testing. Those are two things that are well known to work for genetic testing, and this [risk score] works better than that.”

Farber agrees that genetic data will provide a valuable addition to risk prediction models. “We have relatively good prediction tools—things like FRAX,” he says. “But I think that in the next couple of years, there will be a push to include genetic information to make those predictions more accurate. I’m not sure what the timeline will be. It may be five to 10 years from now, but I think, moving forward, these data will begin to be used for prediction and risk stratification.”

In addition to aiding risk prediction, genetic data can also be used to develop new pharmaceutical treatments. “These actually [give] us insight [into] the pathophysiology of the development of osteoporosis. That actually can provide us with new targets for drug development,” Shen says. Five of the eight drugs approved for osteoporosis target genes for which there’s evidence of a contribution to osteoporosis, demonstrating the value of genetic understanding for pharmaceutical development.

Genetic understanding can also help clinicians tailor specific treatments to individuals who will benefit most. “There are more and more indications that those adverse effects [that some patients experience while taking osteoporosis drugs] are genetically predictable,” Karasik says. Research is not yet to the point where scientists can identify based on genetics which patients will respond well to which drugs, but the field probably will be at that point within the next year or two, according to Karasik.

Finally, genetic information could also help prevent osteoporosis before it actually develops. Specifically, genetic testing could be used to alert younger individuals that they are at risk of low BMD, which could in turn encourage preventive lifestyle interventions. “Imagine somebody when they are 80 and they have low BMD,” Kim says. Weight-bearing exercise and diet don’t do anything anymore. You have to take these drugs for osteoporosis, and the drugs are not fun. That same 80-year-old, if they started when they were 20 or 30, then they can improve their BMD by doing weight-bearing exercise and eating calcium and vitamin D. … We could be alerting young people about a disease they can avert 50 years from now.”

All told then, genetics research has significant implications for clinical care. “I doubt [most physicians] see genetics as a useful tool in clinical decisions currently because it’s not part of FRAX, [and] the drugs don’t have labeling to take genetics into account,” Karasik says. But physicians should expect that genetics will soon be impacting practice in multiple ways. “It will come,” he says.

— Jamie Santa Cruz is a health and medical writer in the greater Denver area.

References
1. Yang T, Shen H, Liu A, et al. A road map for understanding molecular and genetic determinants of osteoporosis. Nat Rev Endocrinol. 2020;16:91-103.

2. Trajanoska K, Rivadeneira F. The genetic architecture of osteoporosis and fracture risk. Bone. 2019;126:2-10.

3. Karasik D, Demissie S, Zhou Y, et al. Heritability and genetic correlations for bone microarchitecture: The Framingham Study Families. J Bone Miner Res. 2017;32(1):106-114.

4. Karasik D, Rivadeneira F, Johnson ML. The genetics of bone mass and susceptibility to bone diseases. Nat Rev Rheumatol. 2016;12(6):323-334.

5. Richards JB, Rivadeneira F, Inouye M, et al. Bone mineral density, osteoporosis, and osteoporotic fractures: a genome-wide association study. Lancet. 2008;371(9623):1505-1512.

6. Styrkarsdottir U, Halldorsson BV, Gretarsdottir S, et al. Multiple genetic loci for bone mineral density and fractures. N Engl J Med. 2008;358(22):2355-2365.

7. Styrkarsdottir U, Halldorsson BV, Gretarsdottir S, et al. New sequence variants associated with bone mineral density. Nat Genet. 2009;41(1):15-17.

8. Timpson NJ, Tobias JH, Richards JB, et al. Common variants in the region around Osterix are associated with bone mineral density and growth in childhood. Hum Mol Genet. 2009;18(8):1510-1517.

9. Rivadeneira F, Styrkársdottir U, Estrada K, et al. Twenty bone-mineral-density loci identified by large-scale meta-analysis of genome-wide association studies. Nat Genet. 2009;41(11):1199-1206.

10. Estrada K, Styrkarsdottir U, Evangelou E, et al. Genome-wide meta-analysis identifies 56 bone mineral density loci and reveals 14 loci associated with risk of fracture. Nat Genet. 2012;44(5):491-501.

11. Medina-Gomez C, Kemp JP, Trajanoska K, et al. Life-Course Genome-wide Association Study Meta-analysis of Total Body BMD and Assessment of Age-Specific Effects. Am J Hum Genet. 2018;102(1):88-102

12. Kemp JP, Morris JA, Medina-Gomez C, et al. Identification of 153 new loci associated with heel bone mineral density and functional involvement of GPC6 in osteoporosis. Nat Genet. 2017;49(10):1468-1475.

13. Koller DL, Zheng HF, Karasik D, et al. Meta-analysis of genome-wide studies identifies WNT16 and ESR1 SNPs associated with bone mineral density in premenopausal women. J Bone Miner Res. 2013;28(3):547-558.

14. Wang C, Zhang Z, Zhang H, et al. Susceptibility genes for osteoporotic fracture in postmenopausal Chinese women. J Bone Miner Res. 2012;27(12):2582-2591.

15. Morris JA, Kemp JP, Youlten SE, et al. An atlas of genetic influences on osteoporosis in humans and mice. Nat Genet. 2019;51(2):258-266. Published correction appears in: Nat Genet. 2019;51(5):920.

16. Kim SK. Identification of 613 new loci associated with heel bone mineral density and a polygenic risk score for bone mineral density, osteoporosis and fracture. PLoS One. 2018;13(7):e0200785.

17. Al-Barghouthi BM, Farber CR. Dissecting the genetics of osteoporosis using systems approaches. Trends Genet. 2019;35(1):55-67.

18. Styrkarsdottir U, Thorleifsson G, Sulem P, et al. Nonsense mutation in the LGR4 gene is associated with several human diseases and other traits. Nature. 2013;497(7450):517-520.

19. Styrkarsdottir U, Thorleifsson G, Eiriksdottir B, et al. Two Rare Mutations in the COL1A2 Gene Associate With Low Bone Mineral Density and Fractures in Iceland. J Bone Miner Res. 2016;31(1):173-179.

20. Zheng HF, Forgetta V, Hsu YH, et al. Whole-genome sequencing identifies EN1 as a determinant of bone density and fracture. Nature. 2015;526(7571):112-117.

21. Trajanoska K, Morris JA, Oei L, et al. Assessment of the genetic and clinical determinants of fracture risk: genome wide association and mendelian randomisation study. BMJ. 2018;362:k3225.