399Twin Research Volume 6 Number 5 pp. 399± 408
Amajor component of variation in body height is due to
genetic differences, but environmental factors have a substantial
contributory effect. In this study we aimed to analyse
whether the genetic architecture of body height varies between
affluent western societies. We analysed twin data from eight
countries comprising 30,111 complete twin pairs by using the
univariate genetic model of the Mx statistical package. Body
height and zygosity were self-reported in seven populations and
measured directly in one population. We found that there was
substantial variation in mean body height between countries;
body height was least in Italy (177 cm in men and 163 cm in
women) and greatest in the Netherlands (184 cm and 171 cm,
respectively). In men there was no corresponding variation in
heritability of body height, heritability estimates ranging from
0.87 to 0.93 in populations under an additive genes/unique environment
(AE) model. Among women the heritability estimates
were generally lower than among men with greater variation
between countries, ranging from 0.68 to 0.84 when an additive
genes/shared environment/unique environment (ACE) model
was used. In four populations where an AE model fit equally
well or better, heritability ranged from 0.89 to 0.93. This difference
between the sexes was mainly due to the effect of the
shared environmental component of variance, which appears to
be more important among women than among men in our study
populations. Our results indicate that, in general, there are only
minor differences in the genetic architecture of height between
affluent Caucasian populations, especially among men.
Body height is a classic example of a polygenic inherited
trait, which is also substantially influenced by environmental
factors during fetal life, childhood, and adolescence
(Sinclair, 1989). Recently, interest in body height as an indicator
of childhood living conditions has increased
remarkably in the field of epidemiology. In his seminal work
Forsdahl (1977) suggested that poor living conditions in
early life could be an important risk factor of cardiovascular
disease. Later, Barker (1998) developed this hypothesis
further and presented that environmental factors acting
during fetal life and early childhood can induce permanent
changes in metabolism, which predispose humans to cardiovascular
disease. A number of studies confirm an inverse
association between body height and cardiovascular disease,
and these results are interpreted as offering support for the
role of childhood environmental factors in the aetiology of
these diseases (Davey Smith et al., 2000; Jousilahti et al.,
2000), but the findings are not consistent and this explanation
is unproven (Silventoinen et al., 2003a). In light of
recent discussion about the possible reasons behind the
association between body height and cardiovascular disease,
research to improve understanding of factors affecting body
height can also yield important insights into the cause of
cardiovascular disease.
As well as being a measure of early life material circumstances,
body height has also been used as a societal level
indicator of the standard of living. The important focus
here is on the dramatic increase in body height throughout
the world during the 20th century. The increase in body
height has been uniform with the possible exceptions of the
two world wars, and it seems to be a universal process
occurring worldwide in both affluent societies and in the
developing world (Silventoinen, 2003). In Europe, mean
body height has increased about 1 cm per decade during
the 20th century (Cavelaars et al., 2000). This increase is
very similar across European countries, and the differences
in mean body height have generally remained stable over
Heritability of Adult Body Height:
A Comparative Study of Twin Cohorts
in Eight Countries
Karri Silventoinen1
, Sampo Sammalisto2
, Markus Perola2
, Dorret I. Boomsma5
, Belinda K. Cornes3
,
Chayna Davis7
, Leo Dunkel10
, Marlies de Lange8
, Jennifer R. Harris6
, Jacob V.B. Hjelmborg4
,
Michelle Luciano3
, Nicholas G. Martin3
, Jakob Mortensen4
, Lorenza Nisticò9
, Nancy L. Pedersen7
,
Axel Skytthe4
, Tim D. Spector8
, Maria Antonietta Stazi9
, Gonneke Willemsen5
, and Jaakko Kaprio1
1
Department of Public Health,University of Helsinki, Finland
2
National Public Health Institute,Helsinki, Finland
3
Queensland Institute of Medical Research, Brisbane,Australia
4
Institute of Public Health,Epidemiology,University of Southern Denmark,Odense, Denmark
5
Dept of Biological Psyschology,Vrije University,Amsterdam,the Netherlands
6
Norwegian Institute of Public Health,Norway
7
Karolinska Institute,Stockholm, Sweden
8
Twin Research Unit,St.Thomas’s Hospital,London, UK
9
Istituto Superiore di Sanità, Rome, Italy
10
Dept of Pediatrics,University of Helsinki
Address for correspondence: Jaakko Kaprio, M.D., Ph.D., Dept. of
Public Health, University of Helsinki, PO Box 41, FIN-00014
Helsinki, Finland. Email:jaakko.kaprio@helsinki.fi
400 Twin Research October 2003
Karri Silventoinen et al.
birth cohorts. However, one study (Silventoinen et al.,
2001b) has reported that the difference in mean body
height between Finland and Sweden has diminished in the
cohorts born after the Second World War simultaneously
when the Gross National Product in Finland, remarkable
lower than in Sweden before the Second World War, started
to increase rapidly and eventually reached the level of
Sweden. This upward secular trend is probably due to
reduction of environmental stress and may reflect a general
increase in affluence in the world. This hypothesis is supported
by strong aggregate level correlations between the
increasing trends in body height and increase in the Gross
National Product (Steckel, 1995) as well as reduction in
menarcheal age in girls, widely accepted to be due to
improved nutrition (Hauspie et al., 1996).
The genetic factors affecting the variation in body
height has also been under intensive scrutiny. The earliest
estimate for the heritability is available in the data used
in the first population genetic study of body height by
Karl Pearson and Alice Lee (1903) carried out in British
families at the beginning of the last century. Heritability
of body height in these data was estimated to be 0.79 by
a later study (Crow & Kimura, 1970). A recent review
(Silventoinen, 2003) identified 24 twin, 7 adoption, and 22
other family studies which reported familial correlations of
body height. Many of those studies are problematic due to
data collection issues or small sample sizes, but there are also
a few well-conducted studies for which heritability estimates
for body height varied from 0.60 (Husén, 1959) to 0.80
(Stunkard et al., 1986) in men.
Previous studies have shown substantial variation in mean
body height between European populations. In this study,
we examined whether there is corresponding variability in
phenotypic variation and in the genetic architecture of
body height. We chose to examine twin pairs within the age
range of 20 to 40 years, so that growth would have been
completed but decreases in body height with aging would
not yet have occurred.
Data Sources
This study is based upon existing data in the GenomEUtwin
twin cohorts. Where several twin cohorts were
available in same country, we selected the latest born to
minimize the cohort effect. The participating GenomEUtwin
data sets are summarized in Table 1.
Australia
The Australian data were derived from two different
studies: the first, conducted during 1980–1982 was of
twins enrolled in the Australian Twin Register and born
prior to 1964 (Healey et al., 2001); the second was conducted
1989–1990 and was of a younger cohort of twins
from the Australian Twin Register born 1964–1971 (Kirk
et al., 2000). Both studies were conducted as mailed questionnaires,
which included self-reported items on height
and weight. Combining twin pairs from both study cohorts
gives a total of 3484 twin pairs between the ages of 20 and
40, with self-reported height for both cotwins and known
zygosity (which was determined by a combination of traditional
diagnostic questions plus blood grouping and
genotyping). Between 1993 and 1998, subsets of the older
cohort of twins were involved in one or more of a range of
studies involving a clinical examination. In this examination
height was measured with a stadiometer, resulting in
standardised clinical measures of adult height for 2308
individuals from whom the reliability of self-report height
Table 1
Characteristicsof the Twin Populationswith Data on Height at ages 20–40 years in the GenomEUtwin Study
Australia Denmark Finland Italy Netherlands Norway Sweden United Kingdom
Birth cohort 1940–1971 1962–1982 1975–1979 1983 1953–1982 1967–78 1926–1958 1954–1981
Measurement 1980–82, 2002 2000–2002 2003 1991, 1993, 1992,1998 1973 1994–2001
occasions 1988 1995, 1997, 2000
Response 75, 78 75 88 enrolment 69, 44, 60, 57, 53 75 , 63 83 media
rate % ongoing enrolment
N complete 3484 4682 2002 697 2118 3637 8747 578
twin pairs
within the
age group
of 20–40
years
% female twins 63 58 55 59 61 55 53 100
% MZ twins 47 36 33 46 46 37 40 30
N opposite 803 1350 674 169 522 1142 none none
sex pairs
Self reported/ Self Self Self Self Self Self Self Measured
measured
Measurement cm cm cm cm cm cm cm cm
unit
Zygosity Q Q Q Q Q Q Q genotyping
Note: Q means that zygosity has been diagnosed by a validated questionnaire method.
could be estimated (see Healey et al., 2001, for details of
zygosity and clinical measures).
Denmark
The height data were collected from the 2002 twin survey
of the Danish Twin Registry (Skytthe et al., 2002). A total
of 14745 respondents from the birth cohorts 1962–1982
were selected. Respondents with unknown or uncertain
zygosity (851 twins) were excluded. Also excluded were
4068 twins with responses from only one twin in a pair,
and 231 pairs were excluded due to missing height in one
of the twins. The final sample thus includes a total of 4682
twin pairs.
Finland
Height was assessed in the FinnTwin16 fourth wave questionnaire
in 2000 to 2002 (Kaprio et al., 2002). At the
time of analyses, there were 4657 respondents, of which
269 twins with uncertain zygosity were excluded, as well as
370 with responses from only one twin in the pair. There
remained a total of 2009 twin pairs with both twins
included and of known zygosity, aged 23–27 years. In seven
pairs, height was missing on one or both twins, leaving a
final sample size of 2002 pairs.
Italy
In Italy 4827 potential twin pairs born in 1983 (mean age =
19.8 years) residing at the same parental address were contacted
by questionnaires mailed in April–May 2003 (Stazi et
al., 2002). At the time of the analysis (end of June 2003)
603 letters were not delivered (untraceable addressee). Out
of 874 replies, 38 claimed not to be twins, 96 twin pairs
refused to enter the study, and 12 replies were from only
one twin in a pair. Thirteen were of unknown zygosity, and
so 715 pairs were used for the modelling analysis. In 18
pairs, height measure was missing in one twin. Nearly all
(~93%) of twin pairs were actually living together.
The Netherlands
As part of a longitudinal study on health and lifestyle, questionnaire
booklets were sent in 1991, 1993, 1995, 1997,
2000 to twins, family members, and partners registered
with the Netherlands Twin Register (Boomsma et al., 2000;
Boomsma et al., 2002). In total, 16,609 participants completed
a questionnaire at one or more of these surveys. At
each survey participants were asked for their height. For the
purpose of this study, we selected all height data reported
when participants were aged from 20 to 40 years, being
born between 1953 and 1982. If data were available for
more than one questionnaire, the latest completed questionnaire
data were used, while giving preference to surveys
in which both twins participated. The total sample included
5018 twins. In 2118 cases, both twins of a twin pair participated.
In 765 cases, only 1 member of the twin pair
participated. Zygosity was determined from DNA for 535
twin pairs, in all other cases zygosity was determined from
questionnaire data. In 10 cases, 3 complete and 7 incomplete
twin pairs zygosity was missing.
Norway
The twins participating in the program of research at the
Norwegian Institute of Public Health were born from 1967
through 1979, and were identified through information
about plural births contained in the National Medical Birth
Registry. The height data were collected from two questionnaire
studies conducted in 1992 and 1998 (Harris et al.,
2002). The combined first and second questionnaire
samples include 9478 twins who responded to at least one
of the questionnaires.
Sweden
Data are from the 1973 cohort of the Swedish Twin
Registry, which includes the largest sample of twin pairs
with height data who were from 20 to 40 years of age at the
time of testing (Lichtenstein et al., 2002; Pedersen et al.,
2002). The 1973 cohort includes 13197 monozygotic
(MZ) and same-sex dizygotic (DZ) twin pairs (26394 individuals)
between 15 and 47 years of age. An additional 319
twin pairs with uncertain zygosity were excluded. A final
sample size of 8747 MZ and same-sex DZ twin pairs
(17494 individuals) between 20–40 years of age were
included in the analyses.
United Kingdom
Twins from the St Thomas’ UK Adult Twin Registry were
ascertained from the general population through national
media campaigns in the United Kingdom (Spector &
MacGregor, 2002). A large proportion of the twins have
visited the Twin Research Unit at St Thomas’ Hospital
during which a large range of phenotypes was collected for
different studies. Height, weight, and other general characteristics
were measured. Fasting blood samples were taken
for genotyping.
Measurement of Height
In seven countries, adult height measures were based on
self-report from questionnaires, while in the UK it was
based on measured height using a stadiometer. Information
from validation studies of self-reported body height was
available from Australia, Finland, and the Netherlands.
Among Australian twins, the reported correlation between
self-reported and measured height was 0.92 for both
females (1525 subjects) and males (783 subjects). In a subsample
of the Finnish data set, the correlation between
self-reported height and later measured height was 0.97 in
men (N = 96) and 0.96 in women (N = 116), with a mean
overall difference of 0.28 cm. In the Netherlands, among
subjects with self-reported height, who completed the 1997
questionnaire, and who also had their height measured
soon afterwards (N = 466 twins) the correlation was 0.93.
Statistical Methods
Genetic modeling was conducted separately by sex, using
scripts available on the GenomEUtwin Mx-script library.1
Standard univariate twin modeling based on linear structural
equations is used in this study (Neale and Cardon,
1992). The model assumes that phenotypic variation can
be due to variation in additive genetic (A), dominant
genetic (D), shared environmental (C), and specific environmental
(E) components. The correlations between the
environmental components are expected to be the same
within all twin pairs regardless of zygosity (1 for shared and
0 for specific environment), but the correlations in the
genetic components differ between MZ (1 for the both
401Twin Research October 2003
Heritability of Adult Body Height
402 Twin Research October 2003
Karri Silventoinen et al.
genetic components) and DZ twins (0.5 for additive and
0.25 for dominant genetic component). Our data, which
include only twin pairs but not adopted twins or other relatives,
do not allow us to estimates simultaneously C and D
components. In modeling, we primarily selected the model,
which explained the observed covariance matrix best (i.e.,
provided the best c2
-value for each study population separately
for men and women). In order to provide
comparability, we present, however, both ACE and AE
models for each population. Age was adjusted for if a significant
age effect was present. We did not use opposite-sex
pairs in the modeling, because not all centers had data on
such pairs.
Results
In men, mean height ranged from 177 cm to nearly 184
cm and in women from 163 cm to nearly 171 cm by
country (Table 2). Mean body height was consistently
slightly lower in MZ compared to same sex dizygotic
(SSDZ) twins in men and women. The only exception was
the female population in Norway where both MZ and DZ
twins had the same stature. Within countries, there were
generally no variance differences between MZ and SSDZ
pairs. Standard deviations were very similar between countries
and between zygosity groups.
Twin correlations for MZ male pairs were uniformly
high in all countries, ranging from 0.87 to 0.94 in male
and from 0.84 to 0.94 in female MZ pairs (Table 3). The
corresponding DZ correlations were generally slightly more
than one-half of the MZ correlations, ranging from 0.42 to
0.60. In six countries, correlations for opposite sex (OS)
pairs were available and were generally slightly smaller than
the corresponding same-sex pair correlations, with the
exception of Italy, where the OS DZ pair correlation was as
low as 0.30.
Table 4 gives the results for the genetic modeling for
men and women. Among men, an AE model fitted the data
best for each country except Sweden. In Sweden, the ACE
model provided the best model fit and a significant shared
environmental component was observed. The total phenotypic
variance varied from 38.9 in Sweden to 48.3 in Italy,
but the proportion of variance ascribed to genetic factors
was quite similar in all AE models (0.87 to 0.94), but was
expectedly lower (0.77) in Sweden under the ACE model.
Among women, the ACE model generally fitted the
data better than the AE model, which provided the best fit
in Italy, Norway, Sweden and the Netherlands. Total phenotypic
variance varied in a similar fashion as in men, but
estimates of heritability showed more variation. The lowest
heritability was in the UK (0.68) and highest in the
Netherlands (0.84) under an ACE model. Heritabilities
under the AE models ranged from 0.89 to 0.93.
Discussion
We found substantial geographical variation in mean body
height between our study populations; among both men
and women the average body height was lowest in Italy and
greatest in the Netherlands. A similar geographic pattern
has also been found in previous comparative studies, which
have shown that mean body height is lowest in Southern
Europe and is generally greater in Northern Europe.
Similarly, earlier research has also reported that the Dutch
people are the tallest people in Europe (Cavelaars et al.,
2000; Schmidt et al., 1995). Reasons for this geographic
pattern are not fully understood, and both environmental
and genetic differences between the populations offer
potential explanations. However, since the increasing trends
in body height during the last century have generally been
very similar in European populations, it is probable that
genetic factors are most important when explaining these
geographic differences in height. If environmental factors
were the main explanation, it is likely that the differences in
body height between European countries would have
diminished after the Second World War simultaneously
with the increasing standard of living in Europe, as
occurred in body height differences between Sweden and
Finland (Silventoinen et al., 2001b). We also found that
unlike the geographic variation in mean body height, there
was no corresponding variation in the variance of body
height distribution; rather the variances were very similar in
the different populations. This is in accordance with previous
findings that there is remarkable world wide variation
in mean body heights but little variation in phenotypic
variance of body height (Eveleth & Tanner, 1990).
An important question is how well twin data represent
the general population with respect to body height. Previous
literature has shown that twins are shorter at birth and have
lower birth weight than singletons, but this difference disappears
during childhood (Wilson, 1986). In the UK,
Table 2
Mean (and SD) Height by Country and Zygosity Group
Australia Denmark Finland Italy Netherlands Norway Sweden UK
MZm 178.4 (6.92) 180.9 (6.62) 179.3 (6.73) 176.5 (7.46) 182.6 (7.21) 180.4 (6.47) 176.9 (6.87) n.a
DZm 178.9 (6.67) 181.7 (6.65) 179.5 (6.68) 178.0 (7.15) 183.1 (6.89) 181.0 (6.65) 177.2 (6.94) n.a.
DOSm 179.4 (6.47) 181.8 (6.83) 179.7 (6.31) 177.1 (7.23) 183.6 (7.29) 181.3 (6.61) n.a. n.a.
MZf 166.5 (6.83) 167.8 (6.07) 165.3 (5.80) 162.8 (5.99) 169.3 (6.26) 167.3 (6.15) 164.2 (5.59) 164.2 (6.20)
DZf 166.8 (6.92) 168.2 (6.56) 166.3 (5.95) 163.4 (5.31) 170.0 (6.36) 167.3 (6.13) 164.6 (5.67) 163.7 (6.42)
DOSf 166.3 (6.79) 168.5 (6.15) 166.2 (5.49) 164.5 (6.20) 170.9 (6.84) 168.4 (6.00) n.a. n.a.
Note: MZm = male monozygotic twins, DZm = male dizygotic twins, DOSm = male dizygotic twins in opposite sex twin pairs,
MZf = female monozygotic twins, DZf = female dizygotic twins, DOSf = female dizygotic twins in opposite sex twin pairs
SDs for Finland interchanged by sex
403Twin Research October 2003
Heritability of Adult Body Height
comparisons of means and variances for height between
twins and singleton women from a population-based study
revealed no difference in the means or variances between the
MZ and the DZ twins or between the twins and the singletons
(Andrew et al., 2001). Furthermore, a previous Finnish
study compared mean body height in the twin cohort used
in this study when the participants were aged 16–17 years
to that of Finnish singletons at the same age and found no
differences (Pietiläinen et al., 1999).
Although previous studies suggest that body heights
in twins and singletons are comparable, in this study
we found some differences between zygosity groups; in
most of our study populations DZ twins were slightly taller
than MZ twins. A similar difference is reported in a previous
study of older Finnish twin cohorts (Silventoinen et al.,
2000b). The reasons for this body height difference are not
clear. It is possible that the placentation and pregnancy
factors special to MZ pairs might be important, and that
the shorter MZ twins are those who are monochorionic
(Derom et al., 1995). However, even though a lower birth
weight has been found among monochorionic MZ twins
than among other MZ or DZ twins, and low birth weight
predicts lower adult height in twins (Pietiläinen et al.,
2001), it has been reported that this difference in body size
between monochorionic twins and singletons disappears
during childhood (Falkner & Matheny, 1995). The evidence
that mothers of DZ twins are, on average, taller than
mothers of MZ twins (MacGillivray et al., 1988) offers a
further possible explanation.
In spite of the substantial variation in mean body
height, especially among men, there was little variation in
heritability estimates of body height. The estimates were
also quite similar to those from an earlier, large male twin
study in the USA; using military register data it was found
that the heritability was 0.80 (Stunkard et al., 1986).
In Finland, the heritability estimates were slightly lower in
older birth cohorts (0.77) compared to this study. We have
previously shown that heritability increased in Finland
since the beginning of the last century among men, probably
indicating a role of environmental factors (Silventoinen
et al., 2000b). Among men, the estimates were lowest
in Norway and Sweden, partly because ACE models fitted
Table 3
Twin Correlationsfor Height by Country and Zygosity Group
Australia Denmark Finland Italy Netherlands Norway Sweden UK
MZm 0.87 0.89 0.92 0.94 0.89 0.87 0.89 n.a.
DZm 0.42 0.47 0.53 0.57 0.47 0.49 0.56 n.a.
MZf 0.84 0.89 0.87 0.94 0.90 0.89 0.89 0.88
DZf 0.49 0.55 0.53 0.49 0.49 0.49 0.49 0.56
DOS 0.46 0.50 0.49 0.30 0.43 0.44 n.a. n.a.
Note: MZm = male monozygotic twins, DZm = male dizygotic twins, MZf = female monozygotic twins,
DZf = female dizygotic twins, DOS = opposite sex twin pairs
Table 4
Estimates of VarianceComponentsand Heritabilitiesfor Height
Men Women
Model Va Vc Ve Vp h2
Dc2
1
Model Va Vc Ve Vp h2
Dc2
1
Australia ACE 40.26 0.00 6.30 46.60 0.87 — ACE 33.80 6.00 7.60 47.40 0.71 —
AE 40.26 6.30 46.60 0.87 0.00 AE 39.27 7.47 46.74 0.84 4.86*
Denmark ACE 37.20 1.90 5.00 44.20 0.84 — ACE 29.50 6.60 4.20 40.30 0.73 —
AE 38.80 5.00 43.80 0.89 0.55 AE 35.20 4.20 39.40 0.89 13.38***
Finland ACE 34.28 6.98 3.77 45.03 0.76 — ACE 24.30 6.10 4.30 34.70 0.70 —
AE 40.10 3.71 43.81 0.89 3.51 AE 29.61 4.25 33.86 0.87 5.25*
Italy ACE 37.48 12.90 3.29 53.67 0.70 — ACE 25.57 4.39 2.31 32.27 0.79 —
AE 48.31 3.25 51.56 0.94 2.60 AE 29.34 2.30 31.64 0.93 0.99
Netherlands ACE 38.71 5.49 5.66 49.86 0.78 — ACE 33.49 2.23 3.94 39.65 0.84 —
AE 43.66 5.62 49.28 0.89 1.20 AE 35.50 3.92 39.42 0.90 0.58
Norway ACE 33.32 4.47 5.47 43.26 0.77 — ACE 30.00 3.34 4.28 37.66 0.79 —
AE 37.17 5.40 42.57 0.87 2.47 AE 32.95 4.25 37.19 0.89 2.33
Sweden ACE 29.80 4.10 5.00 38.90 0.77 — ACE 25.94 1.08 3.48 30.50 0.85 —
AE 33.32 4.94 38.26 0.87 14.44*** AE 26.90 3.50 30.40 0.89 1.87
UK ACE ACE 26.96 8.16 4.56 39.68 0.68 —
AE AE 33.89 4.34 38.23 0.89 8.21**
Note: Va = additive genetic variance, Vc = shared environmental variance, Ve = specific environmental variance, Vp = total phenotypic variance,
h2
= heritability estimate, Dc2
= change in the c2
-values between AE and ACE models
*p < .05, **p < .01, ***p < .001
404 Twin Research October 2003
Karri Silventoinen et al.
better than AE models and may be less comparable.
In Sweden this may be because the Swedish data represented
older birth cohorts than other data sets; it was also a
large data set with more power to detect small effects.
A previous study on the Swedish Twin Registry data on
middle-aged and older twins has shown that heritability of
body height increased over the period from 1963 to 1984
from 0.69 to 0.84 (Harris, 1998), indicating that heritability
has increased in the Swedish population as it has in
Finland. This is further supported by comparisons to previous
literature since in a Swedish study based on a military
register data collected in 1948–1958, the heritability estimate
was even lower (i.e., 0.60) (Husén, 1959; heritability
estimated by Silventoinen, 2003).
The heritability estimates were highest in Italy both
among men and women. This may be partly due to
ongoing data collection in Italy focused on pairs living at
the same address, and both MZ and DZ correlations are
probably higher in the current data than they will be after
data collection has been completed. However, the Italian
data set also includes the youngest individuals of all the
cohorts studied. Nevertheless, even if the effect of environmental
factors in Italy is slightly underestimated, the high
heritability of body height suggests that the smaller average
body height in Italy compared to the other countries is not
due to environmental factors but rather due to genetic differences
between Italian and other European populations.
Another, although less likely, explanation is that this difference
in mean body height is due to environmental factors
with equal effects on the whole population, thus limiting
environmental variation. A systematic review of previous
twin and family studies on body height showed substantial
Table 5
LOD Scores greater than 2 from Genome Scans of Human Stature.
Chromosomal LOD Score Population Reference
region
1p21 2.25 African American (Wu et al., 2003)
2q11 2.23 Caucasian(Finnish/Botnia) (Hirschhorn et al., 2001)
3p14 2.31 Caucasian(Finnish) (Hirschhorn et al., 2001)
3p26 3.17 Caucasian(British/Irish) (Wiltshire et al., 2002)
3p26 2.06 EuropeanAmerican (Wu et al., 2003)
4q25 2.28 Caucasian(Finnish/Botnia) (Hirschhorn et al., 2001)
5q31 2.14 Caucasian(North American) (Deng et al., 2002)
5q31 2.26 EuropeanAmerican (Wu et al., 2003)
6q12 2.66 EuropeanAmerican (Wu et al., 2003)
6q25 3.85 Caucasian(Finnish/Botnia) (Hirschhorn et al., 2001)
6q25 3.06 Caucasian(Dutch) (Xu et al., 2002)
7q11-21 2.26 Caucasian(British/Irish) (Wiltshire et al., 2002)
7q31 2.46 Multiple (Wu et al., 2003)
7q35 3.40 Caucasian(Swedish) (Hirschhorn et al., 2001)
7q36 2.91 Caucasian(Finnish) (Perola et al., 2001)
8q24 2.52 Caucasian(Finnish) (Hirschhorn et al., 2001)
9p1 2.09 Caucasian(Dutch) (Xu et al., 2002)
9q21 2.01 Caucasian(Finnish/Botnia) (Hirschhorn et al., 2001)
9q34 2.61 Caucasian(Finnish) (Perola et al., 2001)
12p13 2.07 Caucasian(Finnish) (Hirschhorn et al., 2001)
12q13 3.35 Caucasian(Finnish) (Hirschhorn et al., 2001)
13q33 3.56 Caucasian(Finnish) (Hirschhorn et al., 2001)
14q23 3.67 EuropeanAmerican (Wu et al., 2003)
17q21 2.69 Caucasian(Finnish/Botnia) (Hirschhorn et al., 2001)
20p12 3.00 Pima Indians (Thompson et al., 1995)
20q13 2.51 Caucasian(Finnish/Botnia) (Hirschhorn et al., 2001)
Table 6
Heritabilitiesfor Heightin PublishedGenome Scan Datasets
Reference Population Heritability
(Hirschhorn et al., 2001) Finnish/Botnia ~80%
Finnish >95%
Saguenay-Lac-St.-Jean ~70%
Swedish ~80%
(Xu et al., 2002) Dutch 78%
(Deng et al., 2002) Caucasian(North American) 73%
(Wu et al., 2003) Multiple 75–98%
(Perola et al., 2001) Finnish 69%
(Wiltshire et al, 2002) Caucasian(British/Irish) 89%
variation in the heritability estimates (Silventoinen, 2003).
Our results suggest that this variation between previous
studies is more likely to be due to poor data quality and
small sample sizes of many previous twin data sets rather
than actual substantial differences in the genetic architecture
of body height itself.
Variation in the heritability estimates of body height
was larger between our study populations in women compared
to men. It has previously been suggested that growth
in women is more resistant to environmental stress than
that in men (Stini, 1969), but our results do not seem to
support this hypothesis. A previous Finnish study has
shown that the heritability of body height has decreased
among women from younger birth cohorts, but a similar
trend was not found in men (Silventoinen et al., 2001a).
This greater temporal variation among women indicates
that there are specific sources of environmental variation
among women, which may also change over time. The
source of this female-specific environmental variation is not
clear but could be associated with the relatively high
current prevalence of dieting and eating disorders, which
may also explain the geographic variation in female populations
in the present study. Other possible relevant factors
are changes in maternal health and family size, which may
affect women differently than men in the next generation.
Shared environmental factors contributed significantly
to variation in body height only in one of the seven male
populations, whereas among women it was more important
and had an estimable effect in four of the eight female populations.
The small effect of common environmental
factors on body height in most of our male populations is
slightly surprising since previous studies have shown substantial
socioeconomic variation in body height both
among men and women, which is considered to indicate
the effect of childhood material environment on growth
and adult body height (Silventoinen et al., 1999). In a previous
Finnish twin study we also found that body height
differences by educational level were mainly due to the
effect of shared environmental factors (Silventoinen et al.,
2000a). It is possible that since most of our study populations
represent cohorts born in 1970s and 1980s, the
standard of living among these cohorts in Europe has been
so high that the role of social background is not detectable.
The only male population for which we found evidence of
shared environmental effects in this study (i.e., Sweden)
represented older cohort than other populations, supporting
the hypothesis that the genetic architecture may have
changed over time.
We also found that the role of shared environment was
more important in the female populations, and the stronger
effect of environmental factors among women was mainly
due to differences in the shared environmental factors. It
remains an open question which specific factors related to
the environment shared by female twins explain this difference.
A previous Finnish study showed that shared
environmental factors affecting body height are common
not only to co-twins but also to their spouses (Silventoinen
et al., 2003b). Thus, these environmental factors need not
to be necessarily related only to family background but they
may also be part of wider childhood environment.
It is possible that assortative mating may contribute to
the results. Previous studies (for a review see Spuhler, 1982)
have shown that assortative mating by body height is a universal
feature of human behaviour. Assortative mating may
inflate the effect of shared environment, since if it is due to
the direct effect of the phenotype on spouse selection (i.e.,
phenotypic assortment) it violates the assumption of
random mating presumed in the genetic modelling used in
this study. In a previous Finnish study of twins and their
spouses, we found that the spousal correlation in body
height was partly due to phenotypic assortment, but also to
the common background of spouses (social homogamy)
contributed to this correlation (Silventoinen et al., 2003b).
It is possible that the stronger effect of shared environment
among women may be associated with differences in assortative
mating between men and women. However, this is
not a likely explanation since when we compared male and
female twin pairs in this previous study, we did not find
any major differences in assortative mating.
It is important to recognise that studies of MZ and DZ
twins reared together are incapable of resolving all the
components of variance that may be of interest. The negative
confounding of dominance and shared environment is
the most obvious shortcoming. Since phenotypic assortative
mating for a trait of high heritability can substantially
increase the additive genetic variance between both MZ
and DZ pairs to the same degree, this extra genetic variance
will be completely confounded with common
environment, and negatively confounded with dominance
effects. The only way to resolve all components is in an
extended twin design, which includes parents, siblings,
spouses and offspring of both MZ and DZ twin pairs.
Such an analysis for body height in 30000 subjects from
twin families has been carried out in the Virginia Twin
Register by Eaves and colleagues (1999). They estimated
the following proportions of variance: additive genetic
56% for males, 60% for females; assortment 16%, 17%;
dominance 9%, 7%; total genetic 84%, 87%; shared environment
5%, 7%; unique environment including error
15%, 13%; genotype-environment covariance –1%, –8%.
These results are from a sample in which malnutrition and
major childhood infections have probably not affected
growth. It is probable therefore that what in some countries
in our study is estimated as common environment is
in fact a combination of dominance, assortment, and possibly
rather a small amount of genuine shared environment.
It would be useful in the future to obtain estimates of the
marital correlation for height in each country to estimate
the proportion due to assortment.
We had information about opposite sex twin pairs
available in six study populations, and thus were also able
to consider the role of sex-specific genetic factors in body
height. It is clear that genes on the Y-chromosome increase
body height as seen by the sex difference in mean body
height as well as the taller stature of XYY men compared to
XY men (Ratcliffe et al., 1992), but it is not self evident that
these genes affect variation of body height. Sex-specific
genetic effects are indicated by lower phenotypic correlation
among opposite-sex than same-sex siblings. Our results
show only slight differences between correlations of opposite
405Twin Research October 2003
Heritability of Adult Body Height
sex twins and same sex DZ pairs. This suggests that sex-specific
genetic effects only weakly contribute to variation in
body height. This confirms our previous results in a Finnish
population, where we found no evidence for sex-specific
genetic effects (Silventoinen et al., 2001a). The current
results suggest that the sex-specific effect has only a small
role in the genetic variance of body height, and the genes
on the Y-chromosome affect mainly mean body height.
The next step in the study of the genetic architecture of
body height is to try to identify the specific genes. In the
last two decades success in molecular genetic studies
to identify Mendelian disorders has been considerable.
A fruitful approach to look for genetic variants for complex
traits may be to screen genes known to cause Mendelian
syndromes, which have the trait of interest as one of the
phenotypes associated with the disorder. For body height,
the picture is somewhat more complicated. There are hundreds
of Mendelian traits for which abnormal body size
characterizes the phenotype (OMIM, 2003). Many
Mendelian traits are serious disorders affecting most phases
in the development of the individual. Thus, it is plausible
to assume that the growth disorder is, in most cases, secondary
to the severe systemic effect the inherited disease
has on the individual rather than due to a gene also affecting
height in the normal population. Therefore, a better
strategy would be to exclude such syndromic forms associated
with height from the study population as much as
possible. In addition, there is the more obvious need to
control for other, less genetic syndromes, which affect
growth, such as diabetes or malnutrition.
However, certain Mendelian traits do present interesting
genes and loci for candidate genes even in general
population. Given the trait difference between sexes, an
obvious choice would be to study the sex chromosomes.
The SHOX gene is associated with Turner’s syndrome and
idiopathic short stature (Rao et al., 1997). The gene is
located in the pseudoautosomal region of the sex chromosomes,
which has very high recombination rates and low
linkage disequilibrium even for rare SNPs (May et al.,
2002). It has also been linked to stature in one of the
genome scans listed in Table 5 (Deng et al., 2002). This is
also one of the rare genome scan reports, which has
included non-autosomal regions in the results. It has been
proposed that a RFLP in the non-recombining region of Yq
is associated with male stature (Ellis et al., 2001); the deletion
of this region also causes syndromic growth disorder
(Ogata et al., 1995). Ellis and colleagues (2001) found an
association suggesting a possible epistatic effect with biallelic
polymorphisms in the aromatase gene (CYP19) and
the non-recombining region of the Y-chromosome. Growth
hormone 1 in 17q23.3, and the genes in the cascade associated
with its actions, have a major effect on growth but
mutations in these genes have been regarded as comparatively
rare causes of short stature (OMIM, 2003). Only one
of the published genome scans has linked 17q to stature
(Hirschhorn et al., 2001). Of the genome scans published
(Table 5), 7q has been linked to stature in four independent
studies (Hirschhorn et al., 2001; Perola et al., 2001;
Wiltshire et al., 2002; Xu et al., 2002), which makes this
region clearly the most interesting autosomal region for
candidate gene selection. Interestingly, this region has been
linked to BMI as well (Feitosa et al., 2002), and contains
the locus for the leptin gene. The heritabilities for height
(Table 6) in the published genome scan data are similar in
means to the twin data but show slightly greater variation
being derived from sib-pair and family data, and a wider
range of populations.
In conclusion, phenotypic variation and the genetic
architecture of body height are very similar between populations
in spite of the substantial variation in mean body
height observed. Heritability of body height is lower
among women than among men and there is also greater
geographic variation in the heritability estimates among
women. This is probably due to unidentified environmental
factors specific to women only, which may vary more
in time and between populations. Sex-specific genetic
factors seem to have only little effect on body height.
Specifying the genes generating the genetic variance of
body height in general population is an important challenge
for future research.
Endnote
1 Available from http://www.psy.vu.nl/mxbib/
Acknowledgments
This project was supported by the European Commission
under the program Quality of Life and Management of the
Living Resources of 5th Framework Programme (no.
QLG2-CT-2002-01254). We would like to thank the twins
in all the countries for their participation. Support in participating
centres: The secretarial assistance of Jytte
Duerlund Jensen, Bachelor of business language, at the
Danish Twin Register. This work was supported by the
Netherlands Organization for Scientific Research (NWO
985-10-002, NWO 900-562-137 and NOW 575-25-006).
Data collection in the Finnish Twin Cohort study has been
supported by NIAAA (AA 08315 and AA 00145), the
Academy of Finland (#44069), and the Yrjö Jahnsson
Foundation. The Italian Twin Registry is supported by the
Italian Health Ministry, Project 2000, N°0AB/F”. The
Norwegian Institute of Public Health program of twin
research is supported by grants from The Norwegian
Research Council, The Norwegian Foundation for Health
and Rehabilitation, and the Nordic Academy of Advanced
Study. The Swedish Twin Registry is supported by grants
from the Department of Higher Education, the Swedish
Scientific Council, and AstraZeneca. The UK Twin
Research Unit is supported by grants from the British
Heart Foundation, Arthritis Research Campaign, the
Wellcome Trust and the Chronic Disease Research
Foundation. Many thanks are owed to the research nurses
for skilful data collection of UK twin data. The Australian
Twin Register is supported by the National Health and
Medical Research Council and data collection was supported
by grants from NHMRC and NIH.
References
Andrew, T., Hart, D. J., Sneider, H., de Lange, M., Spector, T.
D., & McGregor, A. J. (2001). Are twins and singletons
406 Twin Research October 2003
Karri Silventoinen et al.
comparable? A study of disease-related and lifestyle characteristics
in adult women. Twin Research, 4, 454–477.
Barker, D. J. P. (1998). Mothers, babies and health in later life (2nd
ed.). Edinburgh: Churchill Livingstone.
Boomsma, D. I., Beem, A. L., van den Berg, M., Dolan, C. V.,
Koopmans, J. R., Vink, J. M., et al. (2000). Netherland’s twin
family study of anxious depression (NETSAD). Twin
Research, 3, 323–334.
Boomsma, D. I., Vink, J. M., van Beijsterveldt, T. C., de Geus, E.
J., Beem, A. L., Mulder, E. J., Derks, E. M., Riese, H., et al.
(2002). Netherlands Twin Register: A focus on longitudinal
research. Twin Research, 5, 401–406.
Cavelaars, A. E. J. M., Kunst, A. E., Geurts, J. J. M., Crialesi, R.,
Grötvedt, L., Helmert, U., et al. (2000). Persistent variation in
average height between countries and between socio-economic
groups: an overview of 10 European countries. Annals of
Human Biology, 27, 407–421.
Crow, J. F., & Kimura, M. (1970). An introduction to population
genetics theory New York: Harper and Row.
Davey Smith, G., Hart, C., Upton, M., Hole, D., Gillis, C., Watt,
G., & Hawthorne, V. (2000). Height and risk of death among
men and women: Aetiological implications of associations
with cardiorespiratory disease and cancer mortality. Journal of
Epidemiology and Community Health, 54, 97–103.
Deng, H. W., Xu, F. H., Liu, Y. Z., Shen, H., Deng, H., Huang,
Q. Y., et al. (2002). A whole-genome linkage scan suggests
several genomic regions potentially containing QTLs underlying
the variation of stature. American Journal of Medical
Genetics, 113, 29–39.
Derom, R., Derom, C., & Vlietinck, R. (1995). Placentation. In
L. Keith et al. (Eds.), Multiple pregnancy: Epidemiology, gestation
& perinatal outcome. London: The Parthenon Publishing
Group.
Eaves, L. J., Heath, A. C., Martin, N. G., Neale, M. C., Meyer, J.
M., Silberg, J. L., et al. (1999). Biological and cultural inheritance
of stature and attitudes. In C. R. Cloninger (Ed.),
Personality and psychopathology (pp. 269–308). Washington,
DC: American Psychopathological Association.
Ellis, J. A., Stebbing, M., & Harrap, S. B. (2001). Significant
population variation in adult male height associated with the
Y chromosome and the aromatase gene. Journal of Clinical
Endocrinology and Metabolism, 86, 4147–4150.
Eveleth, P., & Tanner, J. M. (1990). Worldvide variation in human
growth (2nd ed.). Cambridge: Cambridge University Press.
Falkner, F., & Matheny, A. (1995). “The long-term development
of twins: Anthropometric factors and cognition. In L. Keith et
al. (Eds.), Multiple pregnancy: Epidemiology, gestation & perinatal
outcome. London: The Parthenon Publishing Group.
Feitosa, M. F., Borecki, I. B., Rich, S. S., Arnett, D. K., Sholinsky,
P., Myers, R. H., et al. (2002). Quantitative-trait loci influencing
body-mass index reside on chromosomes 7 and 13:
The National Heart, Lung, and Blood Institute Family Heart
Study. American Journal of Human Genetics, 70, 72–82.
Forsdahl A. (1977). Are poor living conditions in childhood and
adolescence an important risk factor for arteriosclerotic
disease? British Journal of Preventive and Social Medicine, 31,
91–95.
Harris, J. R. (1998). The etiology of indivudual differences in health
and anthropometric measures: A developmental study of adult
twins. The Pennsylvania State University.
Harris, J. R., Magnus, P., & Tambs, K. (2002). The Norwegian
Institute of Public Health Twin Panel: A description of the
sample and program of research. Twin Research, 5, 415–423.
Hauspie, R. C., Vercauteren, M., & Susanne, C. (1996). Secular
changes in growth. Hormony Research, 41(Suppl. 2), 8–17.
Healey, S. C., Kirk, K. M., Hyland, V. J., Munns, C. F., Henders, A.
K., Batch, J. A., et al. (2001). Height discordance in monozygotic
females is not attributable to discordant inactivation
if X-linked stature determining genes, Twin Research, 4, 19–24.
Hirschhorn, J. N., Lindgren, C. M., Daly, M. J., Kirby, A.,
Schaffner, S. F., Burtt, N. P., et al. (2001). Genomewide
linkage analysis of stature in multiple populations reveals
several regions with evidence of linkage to adult height.
American Journal of Human Genetics, 69, 106–116.
Husén, T. (1959). Psychological twin research. Uppsala: Almqwist
and Wiksells.
Jousilahti, P., Tuomilehto, J., Vartiainen, E., Eriksson, J., & Puska, P.
(2000). Relation of adult height to cause-specific and total mortality:
A prospective follow-up study of 31,199 middle-aged
men and women in Finland. American Journal of Epidemiology,
151, 1112–1120.
Kaprio, J., Pulkkinen, L., & Rose, R. J. (2002). Genetic and environmental
factors in health-related behaviors: Studies on
Finnish twins and twin families. Twin Research, 5, 366–371.
Kirk, K. M., Bailey, J. M., Dunne, M. P., & Martin, N. G. (2000).
Measurement models for sexual orientation in a community
twin sample. Behavior Genetics, 30, 345–356.
Lichtenstein, P., de Faire, U., Floderus, B., Svartengren, M.,
Svedberg, P., & Pedersen, N. L. (2002). The Swedish Twin
Registry: A unique resource for clinical, epidemiological and
genetic studies. Journal of Internal Medicine, 252, 184–205.
MacGillivray, I., Samphier, M., & Little, J. (1988). Factors affecting
twinning. In I. MacGillivray, D. M. Campbell, & B.
Thompson (Eds.), Twinning and twins. Chichester: John
Wiley & Sons.
May, C. A., Shone, A. C., Kalaydjieva, L., Sajantila, A., & Jeffreys,
A. J. (2002). Crossover clustering and rapid decay of linkage
disequilibrium in the Xp/Yp pseudoautosomal gene SHOX.
Nature Genetics, 31, 272–275.
Neale, M. C., Cardon, L. R. (1992). Methodology for genetic
studies of twins and families. Dordrecht: Kluwer Academic
Publisher.
OMIM (TM) Online Mendelian Inheritance in Man. McKusicNathans
Institute for Genetic Medicine, John Hopkins
University (Baltimore, MD) and National Center for
Biotechnology Information, National Library of Medicine
(Bethesda, MD; 2003). Retrieved from http://www.ncbi.nlm.
nih.gov/omim/
Ogata, T., Yoshizawa, A., Muroya, K., Matsuo, N., Fukushima, Y.,
Rappold, G., & Yokoya, S. (1995). Short stature in a girl
with partial monosomy of the pseudoautosomal region distal
to DXYS15: Further evidence for the assignment of the critical
region for a pseudoautosomal growth gene(s). Journal of
Medical Genetics, 32, 831–834.
Pearson, K., & Lee, A. (1903). On the laws on inheritance in man.
Biometrika, 2, 356–462.
Pedersen, N. L., Lichtenstein, P., & Svedberg, P. (2002). The
Swedish Twin Registry in the third millennium. Twin Research,
5, 427–432.
407Twin Research October 2003
Heritability of Adult Body Height
408 Twin Research October 2003
Karri Silventoinen et al.
Perola, M., Ohman, M., Hiekkalinna, T., Leppavuori, J., Pajukanta,
P., Wessman, M., et al. (2001). Quantitative-trait-locus analysis
of body-mass index and of stature, by combined analysis of
genome scans of five Finnish study groups. American Journal of
Human Genetics, 69, 117–123.
Pietiläinen, K. H., Kaprio, J., Rissanen, A., Winter, T., Rimpelä,
A., Viken, R. J., et al. (1999). Distribution and heritability of
BMI in Finnish adolescents aged 16 y and 17 y: A study of
4884 twins and 2509 singletons. International Journal of
Obesity and Related Metabolic Disorders, 23, 107–115.
Pietiläinen, K. H., Kaprio, J., Räsänen, M., Winter, T., Rissanen,
A., & Rose, R. J. (2001). Tracking of body size from birth to
late adolescence: Contributions of birth length, birth weight,
duration of gestation, parents’ body size, and twinship.
American Journal of Epidemiology, 154, 21–29.
Rao, E., Weiss, B., Fukami, M., Rump, A., Niesler, B., Mertz, A.,
et al. (1997). Pseudoautosomal deletions encompassing a
novel homeobox gene cause growth failure in idiopathic short
stature and Turner syndrome. Nature Genetics, 16, 54–63.
Ratcliffe, S. G., Pan, H., & McKie, M. (1992). Growth during
puberty in the XYY boy. Annals of Human Biology, 19, 579–587.
Schmidt, I. M., Jorgensen, M. H., & Michaelsen, K. F. (1995).
Height of conscripts in Europe: Is postneonatal mortality a
predictor?” Annals of Human Biology, 22, 57–67.
Silventoinen, K. (2003). Determinants of variation in adult bodyheight.
Journal of Biosocial Science, 35, 263–285.
Silventoinen, K., Kaprio, J., Koskenvuo, M., & Lahelma, E.
(2003a). The association between body height and coronary
heart disease among Finnish twins and singletons. International
Journal of Epidemiology, 32, 78–82.
Silventoinen, K., Kaprio, J., & Lahelma, E. (2000a). Genetic and
environmental contributions to the association between body
height and educational attainment: A study of adult Finnish
twins. Behavior Genetics, 30, 477–485.
Silventoinen, K., Kaprio, J., Lahelma, E., & Koskenvuo, M.
(2000b). “Relative effect of genetic and environmental factors
on body-height: Differences across birth cohorts among
Finnish men and women. American Journal of Public Health,
90, 627–630.
Silventoinen, K., Kaprio, J., Lahelma, E., Viken, R. J., & Rose, R.
J. (2001a). Sex differences in genetics and environmental factors
contributing to body-height. Twin Research, 4, 25–29.
Silventoinen, K., Kaprio, J., Lahelma, E., Viken, R. J., & Rose, R.
J. (2003b). Assortative mating by body height and BMI:
Finnish twins and their spouses. American Journal of Human
Biology (in press).
Silventoinen, K., Lahelma, E., Lundberg, O., & Rahkonen, O.
(2001b). Body-height, birth cohort and social background in
Finland and Sweden. European Journal of Public Health, 11,
124–129.
Silventoinen, K., Lahelma, E., & Rahkonen, O. (1999). Social
background, adult body-height and health. International Journal
of Epidemiology, 28, 911–918.
Sinclair, D. (1989). Human growth after birth (5th ed.). London:
Oxford University Press.
Skytthe, A., Kyvik, K., Holm, N. V., Vaupel, J. W., & Christensen,
K. (2002). The Danish Twin Registry: 127 birth cohorts of
twins. Twin Research, 5, 352–357.
Spector, T. D., & MacGregor, A. J. (2002). The St Thomas’ UK
Adult Twin Registry. Twin Research, 5, 440–443.
Spuhler, J. N. (1982). Assortative mating with respect to physical
characteristics. Social Biology, 29, 53–66.
Stazi, M. A., Cotichini, R., Patriarca, V., Brescianini, S., Fagnani,
C., D’Ippolito, C., et al. (2002). The Italian Twin Project:
From the personal identification number to a national twin
registry. Twin Research, 5, 382–386.
Steckel, R. H. (1995). Stature and the standard of living. Journal
of Economic Literature, 33, 1903–1940.
Stini, W. A. (1969). Nutritional stress and growth: Sex difference
in adaptive response. American Journal of Physical Anthropology,
31, 417–426.
Stunkard, A. J., Foch, T. T., & Hrubec, Z. (1986). A twin study of
human obesity. JAMA, 256, 51–54.
Thompson, D. B., Ossowski, V., Janssen, R. C., Knowler, W. C.,
& Bogardus, C. (1995). Linkage between stature and a region
on chromosome 20 and analysis of a candidate gene, bone
morphogenetic protein 2. American Journal of Medical Genetics,
59, 495–500.
Wilson, R. S. (1986). Growth and development of human twins.
In F. Falkner, & J. M. Tanner (Eds.), Human growth: Ecological,
genetic, and nutritional effects on growth (Vol. 3; pp.
197–211). London: Plenum Press.
Wiltshire, S., Frayling, T. M., Hattersley, A. T., Hitman, G. A.,
Walker, M., Levy, J. C., et al. (2002). Evidence for linkage of
stature to chromosome 3p26 in a large U.K. Family data set
ascertained for type 2 diabetes. American Journal of Human
Genetics, 70, 543–546.
Wu, X., Cooper, R. S., Boerwinkle, E., Turner, S. T., Hunt, S.,
Myers, R., et al. (2003). Combined analysis of genomewide
scans for adult height: Results from the NHLBI Family Blood
Pressure Program. European Journal of Human Genetics, 11,
271–274.
Xu, J., Bleecker, E. R., Jongepier, H., Howard, T. D., Koppelman,
G. H., Postma, D. S., & Meyers, D. A. (2002). Major recessive
gene(s) with considerable residual polygenic effect
regulating adult height: Confirmation of genomewide scan
results for chromosomes 6, 9, and 12. American Journal of
Human Genetics.