Full text of DOI:10.1359/jbmr.2000.15.7.1253

JOURNAL OF BONE AND MINERAL RESEARCH
Volume 15, Number 7, 2000
© 2000 American Society for Bone and Mineral Research
Editorial
Genetics of Fracture: Challenges and Opportunities
TUAN V. NGUYEN1 and JOHN A. EISMAN2
Over the past 30 years or so epidemiological and genetic argued that the observed familial risk of Colles’ fracture
studies have consistently shown that bone mineral density results from the the genetics correlation in BMD.
(BMD), a measure of bone strength, is a primary determi-
However, there is evidence that the influence of family
nant of fracture risk in the general population (1–4); that a history of fracture is partially independent of BMD. For
large component of the variance of BMD (up to 80%) is example, even after adjusting for BMD, the risk of hip
determined by genetic factors(5–9); and that a family history fracture remains increased in those with a history of hip
of fracture confers significant increase in risk of fracture for fracture in their mother (RR 1.5), sister (RR 1.8), or brother
a relative.(10) In this issue, Deng et al.,(11) in an elegant (RR 2.3). Similarly, the risk of wrist fracture is increased by
study, report that approximately 25% of the liability to maternal (RR 1.5) and paternal (RR 2.4) history of wrist
Colles’ fracture is attributable to genetic factors. How fracture.(10) Thus, it seems there is a genetic component in
should we interpret this finding in relation to what we fracture risk independent of those that determine BMD.
already know about the BMD-fracture relationship and the
Factors other than low BMD also contribute to the risk of
genetics of BMD? Is the observed influence of genetic fracture. Among them, quadriceps weakness and postural
factors on Colles’ fracture mediated through or independent instability,(1) aspects of bone quality as assessed by quanti-
from the genetic influence on BMD? What are the implica- tative ultrasound (QUS) measurements(15) have been shown
tions of this finding to the search for genes currently under- to be predictors of fracture risk. Muscle strength and QUS
way in many research groups around the world?
measurements are modulated largely by genetic fac-
Given the extent of genetic contribution to the variation in tors,(16–17) so that familial aggregation of these parameters
BMD and that BMD predicts fracture risk, it is possible to also could contribute to risk for fractures. Therefore, the
estimate the contribution of BMD-related genes to fracture significance of BMD-adjusted family history fracture risk
in the form of familial relative risk (RR) of fracture for a may reflect the genetics of QUS measurements, muscle
given genetic relationship in BMD.(12) Consider a familial strength, and other clinical risk factors for osteoporotic
correlation in BMD of 0.5 (e.g., between mothers and fractures. However, BMD, QUS measurements, and muscle
daughters) and subjects aged 60ϩ years with a relative mass share some common genetic influences(9,16,18); thus,
fracture risk (RR for each SD decrease in BMD) of 2.4 for the familial influences of QUS measurements, BMD, and
hip fracture and 1.8 for upper limb fracture (Nguyen et al, muscle mass may not be independent. This implies that one
unpublished data). It can be estimated that the familial risk set of common genes contributes to variation in intermedi-
of hip and upper limb fractures is 2 and 1.1, respectively. In ate phenotypes such as bone mass, quality, size, and density
the Deng et al. study, the familial RR of Colles’ fracture for and hence to the liability to fracture. Other genes may
a woman with an affected mother and an affected sister was specifically contribute to each of these bone and nonbone
1.3 and 1.9, respectively. Daughters of mothers with hip factors (e.g., quadriceps strength and falls), which also
fracture have lower BMD at the hip and daughters of contribute to the liability to fracture.
mothers with spine fractures lower BMD at the spine.(13–14)
Apart from individual genes contributing to fracture risk,
A similar relationship can be expected at the proximal some bone- and nonbone-related genes may interact such
forearm and humerus sites, because femoral neck BMD that some alleles only have “deleterious” effects in the
predicts fracture risk at these sites.(1) Thus, it could be presence of specific alleles found at other loci. Apart from
1University of New South Wales, The Simpson Centre for Health Service Innovation, Liverpool, BC NSW, Australia.
2Garvan Institute of Medical Research, St. Vincent’s Hospital, Sydney, NSW, Australia.
1253

1254
NGUYEN AND EISMAN
the complexity of gene-gene interactions, genes also may in such a large pool is undoubtedly a difficult task, and one
interact with nongenetic factors to convert vulnerability into of the major challenges at present is how to design powerful
actual fracture outcomes. Nongenetic factors often are genetic studies to search for these genes. Some relevant and
equated with “environmental” factors, but there are other important questions in study design include the following:
types of nongenetic factors that act during development, What are appropriate sampling plans? What analytical strat-
including epigenetic factors, such as stochastic or chance egies maximize the statistical power to detect genes? What
factors that act during bone development. Indeed, even in phenotypes should be considered?
monozygotic (MZ) twins, who share 100% of their DNA, a
Currently, there are essentially two strategies for search-
difference in phenotypes such as fingerprints is still ob- ing genes that influence a complex trait such as osteo-
served, likely because of such stochastic factors. Thus, porosis: candidate genes and genome screening. In the
fracture does not necessarily arise from a simple biological candidate-gene approach, individual genes can be tested
“defect” but rather represents an event along a continuum of directly for their association with or linkage to a phenotype.
phenotypes.
In a genome search approach, all genes are systematically
It must, therefore, be concluded that the liability to frac- screened using panels of microsatellite DNA markers “uni-
ture is genetically complex. The exact nature of the genetic formly” distributed throughout the entire genome. The
component in fracture may be a combination of polygenic genomewide search approach can only identify region(s)
effects and gene-gene and genetic-environmental interac- where genes are located, not the genes themselves. Any
tions. Under this supposition, no single genetic locus will by region identified by genome scanning typically spans 5–10
itself cause the event or even, for that matter, determine the centimorgans (cM) and localization has to be progressively
major share of risk of fracture or any osteoporosis pheno- refined until a gene actually can be identified. Actually,
type. It is likely that multiple alleles at multiple loci within there is another strategy of “positional candidates” for the
the genome interact to produce vulnerability to fracture. search of genes. In this approach, any identified region of
Also, the genes may act at different times in the develop- several centimorgans is searched intensively for candidate
ment of bone and perhaps even in different bone envelopes. genes using established databases and those coming from
From the functional point of view, there are two kinds of the Human Genome Project.
genes, namely, “osteoporosis” genes that actually determine
In each of the candidate gene and genome search ap-
an abnormality in bone metabolism and hence contribute to proaches, genes can be identified based on a demonstration
the risk of the fracture event per se and “innocent by- of a significant association and/or linkage relationship. As-
stander” genes that are completely unrelated to these phe- sociation analysis compares the frequency of alleles be-
notypes but lie so close to and/or in linkage disequilibrium tween affected (e.g., fracture or low BMD) and unaffected
with the osteoporosis genes that they can be used to track individuals; a given allele is considered to be associated
them. There are several good reasons to search for both with the disease if that allele occurs at a significantly higher
kinds of genes. In particular the definite causes of osteopo- frequency among affected individuals. On the other hand,
rosis are not completely known and known environmental linkage analysis tests whether the genes show correlated
factors such as lifestyle and dietary habit account for only a transmission within pedigrees.
modest proportion of variance of the liability to fracture.
The majority of genetic studies in osteoporosis in the past
Thus, genetics may prove to be the single most powerful have been based on association analysis of candidate genes,
tool for gaining mechanistic insights into the pathophysiol- and results of these studies often have been inconsistent and
ogy of osteoporosis, as has been the case for genes in sometimes difficult to interpret. A major disadvantage of the
familial aggregations of breast cancer. Once an osteoporosis candidate gene approach is that if multiple genes are in-
gene has been identified, researchers can elucidate its func- volved (as is expected in the case of osteoporosis), analysis
tions, study its interactions with developmental and envi- of each candidate gene in isolation of the others may
ronmental factors, examine its ability to predict fracture in amount to testing every gene on the human genome, an
an individual, and use this understanding to devise novel endeavor fraught with statistical problems relating to false-
treatments. For example, most drugs work by binding to a positive results. It is, of course, possible to correct for these
specific protein, altering its activity to achieve a therapeutic associations by statistical adjustment, but by definition,
outcome; thus, identification of a gene and hence its protein candidate genes have a reasonable prior probability of being
product can lead to new targets for new therapeutic agents. involved in disease susceptibility; therefore, correction for
In the future, such knowledge of gene action may even open the number of genes tested should be conservative or it
an opportunity for gene therapy. In terms of public health, could be counterproductive. It is important to recognize that
identification of genes can lead to new diagnostics for the a significant association between a marker and a trait could
early detection of individuals with high risk of fracture and arise from three major possibilities: the “marker” locus is
also help identify responder/nonresponder groups for par- the disease locus; the marker locus has no direct effect on
ticular therapies.
fracture but it is in linkage disequilibrium with the disease
It has been estimated that there are between 60,000 and locus; and there is an artifactual association caused by
70,000 genes located within 3 billion base pairs of DNA in population admixture or other population phenomenon.
the human genome.(19) However, the number of genes Population admixture can occur, for example, in a case-
thought to be involved in osteoporosis (either directly or control study conducted in a mixture of two subpopulations,
indirectly) is much smaller, probably less than 100 and only one of which has both a higher fracture prevalence and a
some of these will have major effects. Finding these genes higher marker frequency. Moreover, even with close link-

GENETICS OF FRACTURE
1255
age, low frequency and modest effect of a locus may lead to error). Nevertheless, with recent advances in statistical ge-
lack of association, that is, the absence of an association netics, particularly the variance component approach,(23)
does not exclude linkage.
this problem (of nonindependence) can be resolved by max-
To minimize those artifacts in population-based associa- imizing the likelihood of a sibship jointly on all members in
tion studies, family based association studies on related the sibship. In fact, with the variance component method, it
individuals have been proposed, such as the transmission is possible to increase the power of a study by including all
linkage disequilibrium test (TDT).(20) The sampling unit in individuals in any extended pedigree in the analysis.(24)
these methods consists of two parents with an affected Todorov et al. (1997)(25) have further shown that large
child; parental alleles not transmitted to affected children sibships can be a cost-effective alternative to the use of
are used as controls. Thus, the TDT considers affected sibling pairs.
children of heterozygous parents at a marker locus and
The ultimate and clinically relevant consequence of os-
simply tests whether these children have received this locus teoporosis is fracture. Thus, it could be argued that fracture
with a probability different from 0.5, the value expected is the most relevant trait for a study of osteoporosis. For a
under random segregation. The TDT test can be powerful. discrete categorical trait such as fracture, Risch (1990))(26)
However, in the more common situation in which the developed the use of the ratio ␭ of the risk to relatives of a
marker locus is different from the disease locus, the power proband to the population prevalence of the trait. In this
is highly dependent on the disease-marker allele frequencies analysis, the power for detecting linkage essentially de-
and the strength of their linkage disequilibrium.
pends on the risk ratio ␭ and the recombination fraction ␪;
Linkage studies are aimed at tracing cosegregration and with power decreasing rapidly as recombination fraction
recombination phenomena between observed marker alleles increases. When ␪ ϭ 0 and ␭ ϭ 2, a sample size of only 200
and unobserved putative-trait influencing alleles among sib pairs would be required to have an 80% power to detect
members of families or pedigrees. Although the association linkage at a LOD score Ͼ 3. Considering this approach in
analysis of candidate genes has been productive in osteo- relation to the data of Deng et al.,(11) with RR for female
porosis, it can only lead to identification of allelic associa- first-degree relatives of 1.3–1.9, the search for osteoporosis
tions for genes known to be involved in bone biology. A genes based on the dichotomous phenotype of fracture
genomewide search using linkage analysis offers the poten- could be useful. However, the potential for using fracture as
tial to identify a range of known (and unknown) genes the phenotype for genetic studies is limited by the phenom-
involved in determination of osteoporosis. The genome enon of phenocopies, that is, environmentally produced
screening approach has several advantages, but perhaps the phenotypes that mimic the genetically produced pheno-
most important is that it requires no assumptions about types. An example of this would be fracture of “nonfragile”
possible pathophysiological mechanisms. One recent anal- bones caused by environment-related falls. In fact, any
ysis of “osteopenia” in 37 members of 7 families suggested fracture event is the result of a number of factors including
monogenetic inheritance on the basis of a possibly bimodal reduced BMD, bone quality, and nonbone factors such as
inheritance.(21) This group analyzed 143 members and con- quadriceps weakness and postural instability. Hence, it
cluded that loci other than those associated with the colla- seems inherently more reliable and informative to use these
gen I␣1 and I␣2 and vitamin D receptor (VDR) genes are traits as surrogate phenotypes (rather than the global phe-
responsible for the low bone density observed in affected notype of fracture) for genetic studies of osteoporosis.
members. By contrast, Johnson and colleagues have map-
Despite some advantages of economies of scale and
ped an apparently autosomal dominant gene associated with model independence in studying multiple fracture-related
very high bone mass in just 28 members of an extended traits in genetic studies, it would be suboptimal to examine
family pedigree.(22) This mapping was facilitated by the each phenotype separately, because of the increase in type I
extreme degree of difference in bone density but it is not error rate. Although procedures for correction such as the
clear to what extent this infrequent trait contributes to bone Bonferroni adjustment can be applied, the results may be
mass or density in the overall population. These limitations too conservative, because the phenotypes are likely corre-
notwithstanding, the successes of these studies support the lated. Thus, instead of adjusting the significance level of
power of linkage studies in extended family pedigrees over univariate analyses of each phenotype separately, a multi-
association studies.
variate analysis of multiple phenotypes would be preferable.
One of the key issues in any study aiming at identify It has been estimated that linear combination of many traits
genes is the problem of power. If the power of a study is into factor scores is more powerful than the use of univar-
low, there is a good chance that the study’s results will be iate or mean phenotypic data.(27–29) Therefore, some
inconclusive. Among strategies for increasing power, such progress in the genetic dissection of osteoporosis will come
as the selection of sampling unit, other strategies relating to from the integrated analysis of these surrogate phenotypes
data analysis also are worth consideration. Past linkage alone or even perhaps in combination with fracture as a
studies largely have been based on the sib-pair design, in phenotype.
which a set of two related individuals is analyzed. However,
The Human Genome Project will provide crucial oppor-
it is intuitive that larger sibships can provide more informa- tunities for these studies, although some difficulties will
tion than two-individual sib pairs. For a sibship with m likely remain.(30) With all human genes cloned and the
siblings, it is possible to evaluate m (m Ϫ 1)/2 sib pairs. allelic variants identifiable, probably in the form of single
However, the fact that only m Ϫ 1 is independent can lead nucleotide polymorphisms (SNPs), it will be possible to
to a falsely inflated estimate of significance level (type I perform powerful and efficient family based association and

1256
NGUYEN AND EISMAN
linkage studies, for example, using chip technologies.(31)
Thus, the ongoing Human Genome Project, the new molec-
ular technologies, and the new statistical methods will col-
lectively enhance the search for osteoporosis genes and
hence translate the prediction of genetically complex osteo-
porosis into the realm of the possible.
rotic Fractures Research Group. Arch Intern Med 157:629–
634.
16. Howard GM, Nguyen TV, Harris M, Kelly PJ, Eisman JA
1998 Genetic and environmental contributions to the associa-
tion between quantitative ultrasound and bone mineral density
measurements: A twin study. J Bone Miner Res 13:1318–
1327.
17. Arden NK, Spector TD 1997 Genetic influences on muscle
strength, lean body mass, and bone mineral density: A twin
study. J Bone Miner Res 12:2076–2081.
18. Seeman E, Hopper JL, Young NR, Formica C, Goss P, Tsaka-
mandris C 1996 Do genetic factors explain associations be-
tween muscle strength, lean mass, and bone density? A twin
study. Am J Physiol 270:E320–E327.
19. Fields C, Adams MD, White O, Venter JC 1994 How many
genes in the human genome? Nat Genet 7:345–346.
20. Schaid DJ 1998 Transmission disequilibrium, family controls
and great expectation. Am J Hum Gent 63:935–941.
21. Spotila LD, Caminis J, Devoto M, Shimoya K, Sereda L, Ott
J, Whyte MP, Tenenhouse A, Prockop DJ 1996 Osteopenia in
37 members of seven families: Analysis based on a model of
dominant inheritance. Mol Med 2:313–324.
22. Johnson ML, Gong G, Kimberling W, Recker SM, Kimmel
DB, Recker RB 1997. Linkage of a gene causing high bone
mass to human chromosome 11. Am J Hum Genet 60:1326–
1332.
23. Almasy L, Blangero J 1998 Multipoint quantitative-trait link-
age analysis in general pedigrees. Am J Hum Genet 62:1198–
1211.
24. Williams JT, Blangero J 1999 Comparison of variance com-
ponents and sibpair-based approaches to quantitative trait link-
age analysis in unselected samples. Genet Epidemiol 16:113–
134.
25. Todorov AA, Province MA, Borecki IB, Rao DC 1997 Trade-
off between sibship size and sampling scheme for detecting
quantitative trait loci. Hum Hered 47:1–5.
26. Risch N 1990 Linkage strategies for genetically complex traits.
II. The power of affected relative pairs. Am J Hum Genet
46:229–241.
REFERENCES
1. Nguyen T, Sambrook P, Kelly P, Jones G, Lord S, Freund J,
Eisman J 1993 Prediction of osteoporotic fractures by postural
instability and bone density. BMJ 307:1111–1115.
2. Hui SL, Slemenda W, Johnston CC 1989 Baseline measure-
ment of bone mass predicts fracture in white women. Ann
Intern Med 111:355–361.
3. Melton LJ III, Atkinson EJ, O’Fallon M, Wahner HW, Riggs
BL 1993 Long-term fracture prediction by bone mineral as-
sessed at different sites. J Bone Miner Res 8:1227–1233.
4. Cummings SR, Black DM, Nevitt MC, Browner W, Cauley J,
Ensrud K, Genant HK, Palermo L, Scott J, Vogt T 1993 Bone
density at various sites for prediction of hip fractures. Lancet
341:72–75.
5. Pocock NA, Eisman JA, Hopper JL, Yeates MG, Sambrook
PN, Eberl S 1987 Genetic determinants of bone mass in adults:
A twin study. J Clin Invest 80:706–710.
6. Young D, Hopper JL, Nowsen CA. Green RM, Sherwin J,
Kaymacki B, Smid M, Guest GS, Larkins RG, Wark JD 1995
Determinants of bone mass in 10 to 26 years old females: A
twin study. J Bone Miner Res 10:558–567.
7. Flicker L, Hopper JL, Rogers L, Kaymacki B, Green RM,
Wark JD 1995 Bone mineral density determinants in elderly
women: A twin study. J Bone Miner Res 10:1607–1613.
8. Smith DM, Nance WE, Kang KW, Christian JC, JohnstonCC
1973 Genetic factors in determining bone mass. J Clin Invest
52:2800–2808.
9. Nguyen TV, Howard GM, Kelly PJ, Eisman JA 1998 Bone
mass, lean mass, and fat mass: Same genes or same environ-
ments? Am J Epidemiol 147:3–16.
10. Fox KM, Cummings SR, Powell-Threets K, Stone K 1998
Family history and risk of osteoporotic fracture. Study of
Osteoporotic Fractures Research Group. Osteoporos Int
8:557–562.
11. Deng HW, Chen WM, Recker S, Stegman MR, Li JL, Davies
KM, Zhou Y, Deng H, Heaney R, Recker RR 2000 Genetic
determination of Colles fracture and differential bone mass in
women with and without Colles fracture. J Bone Miner Res.
15:1243-1252.
27. Allison DB, Thiel B, St Jean P, Elston RC, Infante MC, Schork
NJ 1998 Multiple phenotype modeling in gene-mapping stud-
ies of quantitative traits: Power advantages. Am J Hum Genet
63:1190–201.
28. Schmitz S, Cherny SS, Fulker DW 1998 Increase in power
through multivariate analyses. Behav Genet 28:357–363.
29. Boomsma DI, Dolan CV 1998 A comparison of power to
detect a QTL in sib-pair data using multivariate phenotypes,
mean phenotypes and factor scores. Behav Genet 28:329–340.
30. Chakravarti A 1998 It’s raining SNPs, hallelujah? Nat Genet
19:216–217.
12. Aalen OO 1991 Modelling the influence of risk factors on
familial aggregation of disease. Biometrics 47:933–945.
13. Seeman E, Hopper JL, Bach LA, Cooper ME, Parkinson E,
MacKay J, Jerums G 1989 Reduced bone mass in daughters of
women with osteoporosis. N Engl J Med 320:554–558.
14. Seeman E, Tsalamandris C, Formica C, Hopper JL, McKay J
1994 Reduced femoral neck bone density in the daughters of
women with hip fractures: The roles of low peak bone density
in the pathogenesis of osteoporosis. J Bone Miner Res 9:739–
743.
31. Gerhold D, Rushmore T, Caskey CT 1999 DNA chips: Prom-
ising toys have become powerful tools. Trends Biochem Sci
24:168–173.
Address reprint requests to:
Dr. Tuan V. Nguyen
The Simpson Centre
Liverpool Hospital
Locked Bag 7103
15. Bauer DC, Gluer CC, Cauley JA, Vogt TM, Ensrud KE,
Genant HK, Black DM 1997 Broadband ultrasound attenua-
tion predicts fractures strongly and independently of densitom-
etry in older women. A prospective study. Study of Osteopo-
Liverpool BC NSW 1871
Australia