Word count: 3296
Genetics and Exercise Performance
Department of General Education,
Osaka Institute of Technology
Asahi-ku, Osaka, 535, Japan
FAX. 06-952-9976 or 06-957-2137
Factors affecting exercise performance can be grouped into genetic and
environmental. Environmental factors, such as training, nutrition and lifestyle, have been
studied extensively. For top athletes, however, genius for exercise performance is a
matter of primary concern. Because genetic factors cannot be controlled at the
individual's free will, when top athletes wish to attain a high level of exercise
performance, most of them considers to what extent exercise performance is genetically
determined. There is insufficient information about this topic because of the difficulty of
Genetic approaches to the physiological, psychological and biomechanical
characteristics in athletes are needed. However, since this encyclopedia deals primarily
with physiological factors affecting performance and health, this essay will summarize
the knowledge about genetics and exercise performance from a physiological
In general, there are large inter-individual differences in exercise performance . As
exercise performance can vary as a consequence of both genetic and environmental
effects, many studies of exercise performance have been concerned with the genetic
contribution. The genetic approach to exercise performance seems to have been based
mainly on variance analysis, as established by Falconer(6). Falconer defined the total
variance as phenotypic variance, which is the sum of the separate components. This
phenotypic variance can be broadly divided into two types of variance, genotypic
variance and environmental variance. That is, the variance based on genetic background
and that produced by the environment. The relationship between these components can
be shown briefly as follows:
To attempt quantitative analysis of the genetic contribution to phenotypic variance,
"heritability in the broad sense(h2B)" has been calculated in general terms. The
calculation can be shown as follows;
The h2B is the extent to which inter-individual variation of a population is genetically
determined. Based on this basic definition, a determination of the genetic contribution to
exercise performance has been attempted.
Aerobic characteristics are among the most important factors determining
exercise performance, so many studies on the subject have been made to date. In these
studies, the genetic contribution to inter-individual variation in aerobic characteristics has
been extensively investigated. To determine the genetic contribution to inter-individual
variation, h2B was estimated by the classical method of comparing twins in the early
As monozygotic twins are genetically identical, intra-pair variance is mainly
environmental in origin. On the other hand, intra-pair variance in dizygotic twins consists
of environmental and genotypic variance:
Vmz = intra-pair variance in monozygotic twins
Vdz = intra-pair variance in dizygotic twins
So, h2B can be estimated as follows:
Based on this classical twin method, Klissouras(9) estimated h 2B in relation to maximal
oxygen consumption( O2max), which is generally considered the best single measure of
aerobic characteristics. The estimated h2B was 93.4%, which means that the variance
of O2max is largely determined by heredity. In this case, in calculating h 2B, two basic
assumptions were made. First, that environmental influences were comparable for both
twins. Second, that no genetic-environmental interaction occurred.
However, it has become clear, as a result of recent findings, that these
assumptions cannot be justified. Bouchard(2) reported that intra-pair similarity in
dizygotic twins is greater than that in brothers. As both kinds of siblings share about 50%
of their genes, he suggested that the difference may be attributed to a more similar
environment in dizygotic twins, leading to suspicion that the monozygotic twins could be
more alike than dizygotic twins depend on not only an increased genetic similarity but
also as a result of more similar environment. Moreover, Bouchard(1) indicated that
genetic-environmental interaction, such as trainability in endurance training, occurs with
respect to several indicators of aerobic work metabolism. In short, there are considerable
inter-individual differences in the response of aerobic characteristics to training.
Furthermore, Bouchard pointed out that experimental conditions and the physical
characteristics of the subjects must be appropriately controlled. In fact, inconsistent
estimates of the genetic contribution to O2max have been reported, ranging from as low
as zero to more than 90%. To explain this discrepancy, Bouchard also suggested that
twin sample size, methodological error, laboratory methods, age, sex, pretraining level,
training experience, and body fat affect the genetic contribution to a considerable extent.
At present, from a result of Bouchard's work controlling these factors, h2B in O2max is
thought to be moderate or low.
However, is the genetic contribution to O2max in reality moderate or low? It is
difficult to place an upper limit on the effect of physiological characteristics which are
easily modified by environmental factors such as training. So, in practice, h 2B in O2max
has been estimated regardless of whether there is additional adapted capacity. If h 2B in
O2max is estimated in subjects who have attained the maximum level of adaptation in
O2max, the estimate may be high because inter-individual variation in O2max reflects
the limit of functional adaptation. In other words, the h 2B component of O2max may
depend largely on the associated functional adaptation level. As most studies have been
carried out with subjects known to be sedentary, the estimates of h 2B may be moderate
Another important question about the genetic contribution to the trainability in
aerobic training remains. From reliable data, Bouchard and Lortie(3) concluded that
inter-individual differences existed in the sensitivity to endurance training, and that this
sensitivity was largely genotype-dependent. These studies revealed that 70 to 80% of
the variance in the training response was genotype-dependent. This estimate is quite
high in comparison with h2B in O2max. One real possibility is that adaptive response in
substrate availability and utilization is determined by genetic background to a
considerable extent. Després et al.(4) reported that sensitivity of stimulated lipolysis to
endurance training is largely genotype-dependent. Although additional supporting data
are needed, substrate availability and utilization may be important in explaining the high
genetic contribution to adaptive response in aerobic training.
Muscular power is also one of the important factors determining exercise
performance. Muscular power generally has been thought to be affected by muscle fiber
composition and mass. Therefore, many sports scientists interested in the genetic
contribution to these physiological characteristics have tried to estimate the degree of
the genetic contribution to total variation.
In particular, there is a fair amount of information about the genetic contribution to
muscle fiber composition. Komi et al.(10) have reported that monozygous twins were
almost identical in the percentage of type I fibers, while dizygous twins were variable and
the h2B was 99.5% and 92.8% for males and females, respectively. This implied that
muscle fiber composition was largely determined by heredity and was not influenced by
environmental factors, such as training, nutrition, or lifestyle.
To date, numerous studies have shown that muscle fiber composition is not a
plastic physiological trait. In short, the absence of a progressive shift between Type I and
Type II fibers with training has been demonstrated. These studies simultaneously
supported the view that muscle fiber composition was genotype-dependent.
However, as discussed in connection with the genetic contribution to O2max.,
several authors have pointed out that twin studies with cross-sectional observations were
not always easy to interpret and the data were often affected by methodological
problems. Lortie(12) suggested that the intra-class coefficient of the percentage of type I
fibers was only 0.55 in monozygotic twins and other factors contributed to the remaining
45% of the total inter-individual variation. In addition, he concluded that about 30% of
total variance can be accounted for by non genetic factors, based on the data obtained
in his laboratory which showed that approximately 12% of the total inter-individual
variation in percentage of type I fibers was due to experimental error.
In fact, several authors have cast doubt on the existence of a high h 2B
contribution to muscle fiber composition. From results obtained at autopsy, Fugl-Meyer
et al.(7) demonstrated muscle fiber composition asymmetry in the extensor carpi radialis
brevis and proposed that this asymmetry was attributable to right/left muscle
morphological asymmetry. From this observation, they concluded that functional
demands were also important determining factors in the development of muscle
structure. Furthermore, with recent advances in histochemistry, it has become clear that
an intermediate fiber type is present (type IIC). This fiber type is thought to appear during
transition between Type I and Type II fibers, leading to speculation that a progressive
shift between Type I and Type II may occur.
In addition, recent studies have revealed that muscle fiber composition drastically
changes under non-physiological condition. Pette et al.(16) have shown that artificial
electrical stimulation induces considerable contractile and metabolic changes in skeletal
muscles. From the biochemical point of view, long-term low-frequency stimulation
transforms the fast type myosin characteristics of these muscles into a slow type that is
similar to that of slow soleus muscles. Although the mechanism by which an electrical
stimulus affects protein isoforms is unknown, it is clear that the mRNA of the slow type
myosin light chain is induced by chronic electrical stimulation. However, transcriptional
regulation of myosin expression is so complicated. Izumo(8) demonstrated that
expression of the myosin heavy chain gene family was transcriptionally regulated by
thyroid hormone in adult rats. That is to say, it was concluded that hyperthyroidism either
increases the mRNA levels of the fast type myosin heavy chain isoform or decreases the
mRNA levels of the slow type myosin heavy chain isoform. Interestingly, it was found
that the same myosin heavy chain gene can be regulated in different ways, depending
on the tissue in which it is expressed, and this transcriptional regulation can even occur
in opposite directions. Izumo's data indicate that one of the important factors determining
muscle fiber composition is thyroid hormone, and that the regulation of this hormonal
effect is complex.
Recent studies have been revealed the physiological plasticity of muscle fiber
composition. Even under standardized condition in mice, Nimmo(15) showed that the
h2B component involved in determining the percentage of Type I fibers is relatively
low(71.2% for males; 80.0% for females) - less than that implied by Komi's data. This
study may indicate that, under nonstandardized conditions such as those that exist in
human society, the dependence of muscle fiber composition on genotype is only
Strictly speaking, h2B in exercise performance does not mean the degree of
"inheritance" of exercise performance. In other words, the extent to which inter-individual
variation in exercise performance is genetically determined does not mean the extent to
which exercise performance are determined by the genes transmitted from parents.
Falconer(6) suggested that genotypic variance must be further divided into additive
variance, dominance variance and interaction variance. Additive variance means
breeding value, which is explained as variance transmitted from the parents; dominance
variance means dominance deviations at each locus; and interaction variance means
interaction deviations among multiple loci. The relationship between these components
can be shown briefly as follows:
Based on this variance analysis, in addition to h 2B, Falconer demonstrated another
form of heritability, which he called "heritability in the narrow sense(h 2N)". The
calculation of h2N can be shown as follows:
As the numerator is not additive variance but genotypic variance, h 2N has been
defined as the extent to which phenotypes are determined by the genes transmitted from
the parents. However, to my knowledge, very few studies of h 2N in exercise
performance have been published. In practice, genetic analysis of the relationship
between exercise performance of parents and offspring is confined to correlation
In fact, several studies of familial relationship have calculated a correlation
coefficient between the exercise performance of parent and offspring. These studies
have dealt with the problem of the parent-offspring relationship affecting aerobic power
and provided a correlation coefficient for O2max . In 1978, Montoye and Gayle(13)
reported a father-son correlation of 0.34 for 93 pairs. Later, Lotie(11) criticized the rather
modest sample size in their study and calculated a relatively low value(r=0.17) for
parent-child correlation in O2max, based on a large number of subjects(170 pairs).
Considering the methodological difference in sample size, the correlation coefficient
published in these two reports may be considered as being substantially identical.
However, Bouchard(1) indicated that correlation coefficients calculated for directly
measured values of O2max are lower than those obtained for estimated values of
O2max. Using to his directly measured data, obtained in a treadmill test, quite small
correlation coefficients were obtained, except for that between mother and child (r=0.03
for parent-child; r=-0.01 for father-child; r=0.28 for mother-child). It was concluded that
the difference between the correlation coefficients obtained for the mother-child and
father-child pairs of measured O2maxvalues could be explained by the existence of a
maternal effect in this phenotype. Bouchard suggested that this result may be evidence
of cytoplasmic inheritance and the involvement of mitochondrial DNA genes. In fact, a
very recent study has indicated that sequence variation in mitochondrial DNA may
contribute to individual differences in O2max(5).
Theoretically, Falconer(6) suggested h2N can be estimated from correlation
coefficients(r) between parent(s) and offspring and 2 x r can be considered as h 2N.
According to this definition, h2N of O2max between mother and offspring can be
approximately estimated as 60% from the data in Bouchard's report(1). This means that
60% of inter-individual differences in O2max are transmitted to the next generation. As
the correlation coefficient for father-child is thought to be about zero, this moderate
degree of inheritance may be attributed to a function of the mitochondrial DNA. As
mitochondrial DNA is the only genetic material transmitted solely by the mother, no
dominance or epistatic effect can be expected. Therefore, the residual variance, about
40%, can be considered as environmental variance or genetic-environmental
interaction variance. However, on the other hand, one must not overlook the rather small
sample size in Bouchard's study; he used only 49 pairs in calculating the correlation
coefficient for mother and child. Further research on this topic is clearly needed.
When comparing aerobic performance, the relationships between parents and
offspring in muscle characteristics have not been considered to date. As described
before, the genetic contribution to individual variation in muscle fiber composition is
thought to be moderate or high. However, it is doubtful whether a significant correlation
exists between parents and offspring in muscle fiber composition. In other words, are
individual variations in muscle fiber composition mainly transmitted from parents? At this
point, it may be useful to introduce Nakamura's report(14) based on a selection
experiment in rats. Nakamura carried out successive selection for a high percentage of
slow twitch fibers(%ST) over four generations and showed relatively low realized
heritability(0.17), which is the value of h2N obtained by selection experiment, but it was a
statistically significant. This means that about 20% of inter-individual variance in muscle
fiber composition in rats is determined by genes transmitted from the parents. Under the
same standardized conditions, Nimmo(15) reported that h 2B in muscle fiber composition
in mice was 71.2% for males and 80.0% for females. Provided that there is no extreme
difference in the mechanisms determining muscle fiber composition in these two
species, it may be concluded that there is a large difference between h 2B and h2N in
muscle fiber composition. In other words, 50 to 60% of individual variance in muscle fiber
composition may be derived from dominance and interaction effects. It was also
recognized by Nimmo that significant dominance in the direction of increasing the Type I
percentage is present in mice to the extent of about 25%. These facts may mean that
dominance and interaction effects are very important, and the additive effect is relatively
unimportant, in determining muscle fiber composition.
In this way, quantitative genetics have revealed the correlation between genetic
and exercise performance. Finally, a simplified model of factors associated with the
exercise performance is shown in Fig. 1. Further studies about the extent with which
these factors associated are needed. At the same time, it is important to decide a
essential genetic determinant to exercise performance. Moreover, the time has arrived to
launch main studies into the inheritance of the genetic determinant.
1. Bouchard, C. Genetics of aerobic power and capacity. In. Sport and Human
Genetics, R. M. Malina and C. Bouchard(Eds.). Champaign, IL: Human Kinetics, 1986,
2. Bouchard, C., R. Lesage, G. Lortie, et al. Aerobic performance in brothers, dizygotic
and monozygotic twins. Med. Sci. Sports Exerc. 18:639-646, 1986.
3. Bouchard, C. and G. Lortie. Heredity and endurance performance. Sports Med. 1:38-
4. Després, J. P., C. Bouchard, R. Savard, D. Prud'homme, L. Bukowiecki and G.
Thériault. Adaptive changes to training in adipose tissue lipolysis are genotype
dependent. Int. J. Obesity 8:87-95, 1984.
5. Dionne, F. T., L. Turcotte, M.-C. Thibault, M. R. Boulay, J. S. Skinner and C.
Bouchard. Mitochondrial DNA sequence polymorphism, O2max, and response to
endurance training. Med. Sci. Sport Exerc. 23:177-185, 1991.
6. Falconer, D. S. Introduction to Quantitative Genetics. New York: Ronald Press, 1960
7. Fugl-Meyer, A. R., A. Eriksson, M. Sjöström and G. Söderström. Is muscle structure
influenced by genetical or functional factors?: A study of three forearm muscles. Acta
Physiol. Scand. 114:277-281, 1982.
8. Izumo, S., B. Nadal-Ginard and V. Mahdavi. All members of the MHC multigene family
respond to thyroid hormone in a highly tissue-specific manner. Science 231:597-600,
9. Klissouras, V. Heritability of adaptive variation. J. Appl. Physiol. 31:338-344, 1971.
10. Komi, P. V., J. T. Viitasalo, M. Havu, A. Thorstensson, B. Sjödin and J. Karlsson.
Skeletal muscle fibers and muscle enzyme activities in monozygous and dizygous twins
of both sexes. Acta Physiol. Scand. 100:385-392, 1977.
11. Lortie, G., C. Bouchard, C. Leblanc, et al. Familial similarity in aerobic power. Human
Biol. 4:801-812, 1982.
12. Lortie, G., J.-A. Simoneau, M. R. Boulay and C. Bouchard. Muscle fiber type
composition and enzyme activities in brothers and monozygotic twins. Sport and Human
Genetics, R. M. Malina and C. Bouchard(Eds.). Champaign: Human Kinetics, 1986, pp.
13. Montoye, H. J. and R. Gayle. Familial relationships in maximal oxygen uptake.
Human. Biol. 50:241-249, 1978.
14. Nakamura, T., S. Masui, M. Wada, H. Katoh, H. Mikami and S. Katsuta. Heredity of
muscle fibre composition estimated from a selection experiment in rats. Eur. J. Appl.
Physiol. 66:85-89, 1993.
15. Nimmo, M. A., R. H. Wilson and D. H. Snow. The inheritance of skeletal muscle fiber
composition in mice. Comp. Biochem. Physiol. 81:109-115, 1985.
16. Pette, D. and G. Vrbova. Invited review: Neural control of phenotypic expression in
mammalian muscle fibers. Muscle and Nerve 8:676-689, 1985.