Genetic determinants of bone mineral content at the spine and radius: A twin study

Bone, 8, 207-209

(1987)

Printed in the USA. All rights reserved

Copyright

8756-3282187 $3.00 + .OO 0 1987 Pergamon Journals Ltd.

Genetic Determinants of Bone Mineral Content at the Spine and Radius: A Twin Study J. DEQUEKER,

J. NIJS, A. VERSTRAETEN,

The Dinsion of Rheumatology, K.U. Leuven, U.Z. Pellenberg, Address

for correspondence

P. GEUSENS, and G. GEVERS

Department of lntemal Medicine, 5-3047 Pellenberg, Belgium. and reprint: Prof. Dr. J. Dequeker,

and Arthritis and Mefabok

Division of Rheumatology,

Abstract

B-3041 Pellenberg,

Belgium

Materials and Methods

the spinal bone mass in the age group younger than 25 years. The heritability index h2 was 0.75 for cortical BMC and 0.88 for axial BMC. Such a genetic determinant could not conclusively be demonstrated in adult twins for the spine and in youngsters for the cortical bone, suggesting that environmental factors may play a more dominant role in growth of cortical bone during adolescence and diminution of axial bone during adult life. Mineral

U.Z. Pellenberg,

Unit,

the study of twins. The magnitude of the differences between the within-pair differences of one-egg and two-egg twins can provide an estimate of the relative importance of hereditary and environment. This formulation of naturenurture studies in twins, which is still the usual method and objective of twin studies, assumes that there are two distinct and distinguisable types of twins, monozygotic (oneegg) and dizygotic (two-egg), and that the environmental forces which affect within-pair character differences are comparable in the two types of twins (Kempthorne and Osborne 1961; Christian 1979).

The possible role of genetic and/or environmental factors in determining bone mass has been investigated in 30 pairs of twins (16 monozygotic and 14 dizygotic) divided in two age groups (below and above 25 years of age). Bone mineral content was evaluated by single- and dual photon absorptiometry at the distol third of the radius for peripheral cortical bone and in the lumbar spine for the axial bone. The “within pair” variance has been used as an index of genetic influence. A significant (p < 0.01) genetic determinant was found for the bone mass of the radius in adults and for

Key Words: Heredity-Bone -Trabecular-Twin Study.

Bone Disease Research

The twin pairs selected for the present study were obtained from a population study on bone mass recruited by a press appeal. A total of 65 pairs older than 10 years volunteered for this study Only same-sexed pairs were considered for this investigation. The criteria for inclusion were that they knew their zygosity and that they had or had had no disease or drugs which might affect the skeletal system Two pairs did not know for sure their zygosity Sixteen pairs had to be eliminated because of diseases. Sacroiliitis, gout, gastrectomy, diabetes mellitus and Marfan syndrome was present in both members of four monozygotic twin pairs and one dizygotic twin pair. The following conditions were present in one of the twin numbers: lumbar stenosis, hypercalcemia, breast cancer, trauma (twice), hypophysectomy, Reiter’s and Behcet’s disease, corticosteroids for a neurological disorder, diabetes mellitus and Legg Calve Perthes disease. The study group consisted finally of 30 pairs of twins (age range 1O-72), -16 monozygotic (MZ) -pairs and 14 dizygotic (02) pairs. Seventeen twins (8 MZ. 9 DZ) were vounaer than 25 vears and 15 (8 MZ, 5 DZ) dlder than 2i years Figure 1 shows the actual age distribution of the twin pairs. The decision to split the age groups at the age of 25 was based on the assumption that until that age most twins in our society will be in a relatively uniform environment as long as the two partners of a pair live at home, with augmenting environmental differences when they are separated later in life

Content-Cortical

Introduction The study of factors determining bone mass is of particular importance for the detection of patients at risk for osteoporosis and prevention of fracture due to minimal trauma. The amount of bone which remains in old age is a result of the interaction between the amount present at maturity and subsequent rates of loss. Different environmental and genetic factors may influence each of these determinants to produce the end result: low bone mass. A genetic component responsible for determining the amount of cortical bone at the peripheral skeleton has been reported (Smith et al., 1973, Moller et al., 1978). At present there are no such studies for the amount of bone at the axial skeleton, the major site of pathological fracture. A simple and efficient method for appraising the role of heredity and/or environment in determining bone mass, particularly with respect to quantitative inheritance, is by

Single

photonabsorptiometry

radius

Measurements of bone mineral content were made at the distal third of the radius at 8 cm from the distal end on the left arm, with use of the single photon absorptiometry techntque (SPA) Norland Cameron Bone Mineral Analyser Model 189, as described by Cameron and Sorenson (1964). This method utilizes a monochromatic beam from a 125l source The forearm is wrapped in “tissue” 207

208

J Dequeker

et al

: Bone mineral content: Twin study

N 5

4 3 2 1 L

i

.. :. :. :. :. :. :. :. :. *:. .:. *:. .:. .:. *:. -:. *:. :.

lo

15

20

25

30

Fig. 1. The age dtstnbution

35

40

of the monozygotic

equivalent material to give a constant soft tissue thickness, and IS scanned transversily Four linear scans are made at each site and the results are averaged The reproducrbility of the method is In our laboratory 1 8%

Dual photonabsorptiometry

lumbar spine

Bone mrneral content (BMC) of the ax/al skeleton was determined by dual photonabsorptiometry (DPA), as described by Krolner and Nielsen (1980) The DPA technique for measurrng bone mineral content IS based on measurements of radiation transmission of two separate photon energres through a medium consisting of two different materials bone and soft trssue. The drchromatlc beam from a ls3Gd source has photon electric peaks at - 44 and 100 kV BMC of the spine was assessed from scans of the L, to L, rearon usina Novo BMC Lab 22a ,Total bone mineral content IS calculated by summrng up BMC values within the regron of interest L,-L, and is expressed after calibration tn units of grams hydroxyapatite (gHA). The reproducibrlity of the measurements of the lumbar spine In our laboratory IS 1 8% All the measurements were performed by a laboratory technrclan who was unaware of the zygosity diagnosis Twin data have been analysed according to the method descrtbed by Chrtstian (1979) The first step in the analysrs IS a test of the differences between the means of the twin types. the

Table I. Analysts of variances

45

50

55

60

N Parrs Mean BMC radrus Among Parrs Mean squares Within Pairs Mean squares

8 0 925 1434

9 0 899 8101 5

1 19

2 23

AGE

second step IS the test for assocration of total variance with twin type and the thtrd step IS only possible when there are no srgnifrcant differences found In the first two steps When In one of them a srgnifrcant difference IS found, a serious bias may occur The heritahrlity index described by Holzinger (7) was calculated as h2 = Drzygotlc (DZ) rntrapair varianceMonozygotrc (MZ) intraparr variance/DZ intraparr variance

Results are presented in difference between the mean BMC and the within-pair mean squares. The among pairs mean squares were significantly different in the twin group younger than 2.5 years but not in the group older than 25 years. For this reason it is not proper to utilize the “wlthrn pairs” variance comparrsons as a test of the genetic effect In the younger group, but the among pair variance compansons have to be used (Christian, 1979). No significant difference was found in the group younger than 25 years. The estimation of genetic variance of the within-pair was srgnificantly different (p -c 0.01) between monozygotic and dizygotrc twins older than 25 years Results

from

the

BMC

of the disiradius

Table I. There was no significant

and dlzygotrc

twins

Twins ~25 Years Drzygotrc

70

(MZ) and Drzygotrc (DZ) twin Pairs studred

of cortical bone mineral content tn monozygotic

Monozygotic

65

Twins ~25 Years P

NS P
Dizygotrc

P

8 1 129 69 89

5 1 020 90 23

NS NS

4 56

17997

NS

Monozygotrc

Estimate of Genetic Variance Among Pairs Within Pairs

F ratio = 0 017

NS F ratio = 3 94

P
J Dequeker

et al

209

: Bone mineral content: Twin study

Table II. Analysis of variances of lumbar spine bone mineral content in monozygotic

and bizygotic

twins

Twins ~25 Years Monozygotic N Pairs Mean Lumbar BMC Among Pairs Mean squares Withln Pairs Mean squares

Dizygotic

8 39.3 415.4 1 08

9 35.1 1954 8.93

Twins >25 Years P

MonozygotIc

NS NS

8 41 6 1136

NS

104

Dizygotic

P

4 37.2 27.12

NS N S.

6 64

N.S

Estimate of Genetlc Variance Withln Pairs

F ratio = 8 268

The results of the spme BMC are presented in Table II. One pair in the dizygotic adult group had to be eliminated because of local spondylosis. No significant differences in mean lumbar BMC, nor In among- or within-pair mean squares was disclosed. The F ratio of the within-pair comparisons is significant In the group younger than 25 years and not in the twins older than 25 years, The numerical estimate of heritability has been calculated usrng the hz index. Ranges from 0 to 1 indicate the range from no heritability to full heritability. The h2 value for cortrcal bone BMC was 0.47 for the younger group and 0.75 for the older group. The respective hz values for the spine BMC were 0.88 and - 0.57.

P < 0.01

F ratio = 0.640

N.S

The heritability index found In this study for cortical bone (0.75) is very similar to that found by Moller et al. (1978) in elderly twins for metacarpal cortical thickness (0.78) and by Smith et al. (1973) in adult twins for the radius (0.75). The h2 value found for the axial bone in this study in young twins is higher than the value for peripheral bone. Since no genetic determinant could conclusively be demonstrated for the spine in adults and for the peripheral skeleton in youngsters, the present data suggest that environmental factors have a differential effect on bone mineral content at the axial or peripheral skeleton according to the age period. Environmental factors may play a more dominant role In growth of cortical bone during adolescence and diminution of axial bone during adult life.

Discussion We have demonstrated by SPA and DPA that the genetic determinant of BMC In the spine is consrderable in younger individuals while in the radius the genetic effect IS more manifest In adults. When the twin model is used, it is assumed that the populatron means of the variables are equal and that total variance within the population of monozygotics and dozygotrcs is also equal (Christian, 1979). The means and total variances were equal in lumbar BMC groups and In radius BMC groups older than 25 years. This within-pair component could be used as an index of genetic influence and a significant effect was observed. The total variance in the radius BMC group younger than 25 years was significantly different (P < 0.01) and here the among-twin parr comparison was used and was not found to be significant. The present results confirm the findings of Smith et al. (1973) and Moller et al. (1978) that there is a genetrc effect on cortical bone mass in adults. In young juvenile twins. however, Smith et al. (1973) could not demonstrate conclusively a genetic effect on cortical bone, a result very similar to our findings tn the age group younger than 25 years In this study for the first time a genetic effect on the development of amount of axial bone mass could be demonstrated in twins younger than 25 years old. Such a genetic determinant for axial bone could not conclusively be demonstrated in adult twins.

References Cameron J R and Sorenson J Measurement of bone mIneral In v~vo An Improved method Science 142 230-232 1964 ChrIstIan J C Testing twin means and estlmatlng genetic variance Base methodology for the analyss of quant tatlve !wln data Acta Genet Med Gemeiloi 2835-40 1979 Holzlnger. K J The relative effect of nature and nurture nfluences on !win differences j Due Psycho/ 20 241-248 1929 Kempthorne 0 and Osborne R H The nterpreta!ion of twin data Am i Hum Genet 13320-339 1961 Krolner B and Nielsen P Measurement of bone mInera content (BMC) of !he lumbar sp ne I Theory and appllcatlon of a new two-dlmens(onal dual-photon attenuation method Stand J C/in Lab invesi 40 653663 1980 Moller M Horsman A Harvald B Hauge M Hennlngsen K and Nordin B E C Metacarpal morphometry in monozygotIc and d zygotic tw ns Caic TISS Res 25 197-201 1978 Smith D M Nance W C Kang K W ChrIstIan J C ana Johnston C C Genetlc factors in cietermlnlng bone mass d C/in invesf 52 2800-2808 1973

Recerved October 28 1966 Revised January 27 1987 Accepted February 9 1987