elucidation of monogenic forms of obesity both in rodents and in humans [1–4].
This development has had crucial implications for our understanding of childhood
obesity, because most of these genetically determined forms typically
become manifest in infancy or early childhood. In clinical terms, the most
important monogenic form of obesity is due to mutations in the melanocortin-4
receptor gene (MC4R), which can be detected in 2–4% of all extremely obese
children [5, 6]. It is nevertheless safe to state that only a very minor fraction of
those genes involved in body weight regulation have become known. Largescaled
efforts are currently being undertaken to detect additional genes
involved in weight regulation and particularly in obesity. The sequencing of the
human genome having been mostly completed, the upcoming years will
undoubtedly witness the identification of several genes either located within
chromosomal linkage regions and/or identified via association studies. In contrast
to the genes, underlying monogenic forms of obesity, these future findings
will have implications for larger subgroups of the obese population. At the same
time, the effect sizes of such gene variants will prove to be substantially smaller
than in the monogenic forms of obesity. The complexity of gene-gene and geneenvironment
interactions needs to be tackled. The respective molecular genetic
findings are bound to not only have an impact on the scientific community, but
also on clinical practice and even society as a whole. Due to the public interest
in this phenotype the respective research results will find rapid entrance into
modern day society necessitating a communication process between researchers,
clinicians and the public to delineate potential implications. In the medium term
we forsee that obesity will be viewed as a more genetically determined condition
than is currently the case. Because research is also being conducted in
children and has diagnostic implications for this age group, ethical issues need
to be considered. In the light of these current and future developments we
will review the current status and attempt to point out the implications of both
past and future genetic research for our understanding of childhood and
adolescent obesity.
Implications of Heritability Estimates
Twin studies [4, 7] have produced the most consistent and highest heritability
estimates in the range of 0.6–0.9 for body mass index (BMI; kg/m2).
These high estimates apply to twins reared both together and apart. It should,
however, be noted that only single and comparatively small studies exist for
twins reared apart in contrast to the vast amount of studies pertaining to twins
reared together, some of which included thousands of twin pairs. Heritability
estimates of this magnitude indicate that the genetic component for body
weight is almost as high as that for body height.
Except for the newborn period for which a lower heritability of 0.4 has
been calculated [8] age does not affect heritability estimates to a substantial
degree. Evidently, the influence of the intrauterine environment on birth weight
is strong. It is well known that particularly in monozygotic twins other anthropometric
measurements, e.g. body height, correlate less well in infancy than in
childhood. The feto-fetal transfusion syndrome contributes to this phenomenon,
as it substantially reduces the effect of genetic factors at birth and during
infancy. Subsequently, however, genetic factors are able to exert their influence,
thus accounting for the fact that both height and BMI become more similar. In
school-age children high heritabilities already apply. It has been suggested that
the heritability of BMI is maximal ( 0.9) during late childhood and adolescence
[9]. The genes relevant for weight regulation in childhood presumably
only partially overlap with those operative in adulthood [10], thus partially
explaining why intraindividual correlations of BMI during childhood and adulthood
are considerably lower than between adolescence and adulthood [11].
In comparison to the twin studies adoption and family studies have mostly
derived at considerably lower heritability estimates [3, 4, 7]. However, a single
large family study [7] has also come up with a heritability estimate of 0.67,
which is in the same range as those derived from twin studies. In their family
study Maes et al. [7] discussed potential reasons why the high heritability estimates
in twin studies may be better than those obtained in other types of family
studies including a better control for age effects.
For an adequate interpretation of the high heritability estimates obtained in
twin studies it is noteworthy to point out that both direct and indirect genetic
effects are subsumed under the genetic component [4]. To illustrate this aspect
let us assume that both infant twins of a monozygotic pair are frequently irritable
due to a biologically driven increased hunger (direct genetic effect; such
as a mutation in the leptin gene leading to leptin deficiency). This hunger
rapidly induces frequent feedings by the caretaker irrespective of his or her
background. Thus, even if the twins are separated at birth, the caretakers can be
expected to respond similarly as soon as they have learned that bottle feeding
soothes the child. This indirect genetic effect actually represents the response of
the environment to the genetically based excessive hunger. It is readily evident
that in this case the indirect genetic effect is crucial for the development of early
onset obesity in both twins. If one of the caretakers of the reared apart twins
systematically curtails the infant’s energy intake (implying a willingness to
deal with the resulting irritability), early onset obesity would ensue in only the
other twin.
The illustrated indirect genetic effect might be particularly important in
infants and young children who totally rely on their caretakers for food supply.
As the twins grow older their relentless hunger would undoubtedly entail that
they can increasingly by themselves seek and obtain food. In addition, whereas
the initial weight gain might be perceived as an indication of good health,
the caretakers themselves could become aware of the obesity and potentially
attempt to reduce the twins’ energy intake. It is evident, that different patterns
of complex child-parent interactions can ensue; indeed, every therapist familiar
with the treatment of childhood and adolescent obesity is aware of the different
strategies that both obese children and parents pursue depending on the age of
the child. Some children might experience that their overeating is socially not
acceptable and resort to eating secretly. In this situation the indirect genetic
effect is accounted for by modern day society with its ready availability of a
large variety of highly palatable foods (and only few requirements for physical
activity); the fact that foods, snacks and candies are inexpensive, implies that
most children can afford to buy them even without parental permission or
knowledge.
Shared and Non-Shared Environment
Another interesting and important aspect of formal genetic studies has
been the observation that non-shared environment explains considerably more
variance of the quantitative phenotype BMI than shared environment. In the
large twin study of Stunkard et al. [12], which encompassed adult twin pairs
reared together or apart, shared environment did not explain variance; instead
non-shared environment totally explained the environmental component estimated
at 30%. This finding is not compatible with our intuitive approach to
obese children and their families: we are inclined to readily attribute both the
child’s obesity and that of other family members to familial eating and dieting
patterns and the familial level of physical activity. However, based on the relative
unimportance of shared environment the fact that several members within
such a family are obese cannot be attributed to the shared environment; instead
genetic factors would account for the familial loading. Taken one step further,
it is not important what the mother buys to eat or what she serves. Instead, variance
can only be accounted for by assessing what and how much each family
member actually eats [13]. As a consequence of the unimportance of the shared
environment we would need to focus on what makes BMIs of twins or sibs
dissimilar. A list of candidate non-shared environmental experiences believed
to promote obesity in children has been outlined [14]. Therapeutic strategies
aiming at influencing the shared environment are not incorporating the information
stemming from these formal genetic studies. We need to devise ways
and means to therapeutically target the non-shared environment.
A word of caution is required. Recent studies indicate that the shared environment
might play a more substantial role after all [15]. In addition, it has been
argued that past research may have underestimated common environmental
effects on BMI because the designs lacked the power or ability to detect them
[15]. Using a very large nationwide dataset of Swedish military conscripts,
male familial correlations in body mass index (BMI) showed highly significant
correlations for BMI for all biological family relations (r 0.28 for father-son
pairs; 0.36 for full-brothers, 0.21 for maternal half-brothers, and 0.11 for paternal
half-brothers). Both the significantly higher correlation for maternal than
for paternal half-brothers (maternal half-brothers are more likely to grow up
within the same family than paternal half-brothers) and the significant correlation
(r 0.06) found for non-biological quasi father-son relations [16] can be
interpreted as evidence for the effect of the shared environment. Clearly, further
research is required to pinpoint the potentially age-dependent contributions of
both the shared and non-shared environment to the variance of BMI.
Gene X Gene and Gene X Environment Interactions
Formal genetic studies indicate that both additive and non-additive gene
effects are important [8]. In their twin study Stunkard et al. [12] estimated
that in males 57 and 17% of the variance in BMI are due to non-additive and
additive gene effects, respectively. In females the respective percentages were
estimated at 37 and 31%. The discrepancy between the higher and lower heritability
estimates calculated in the twin and family studies, respectively, is presumably
also partially due to the strong effect of non-additive factors, which
can be assessed more reliably in twin studies. The importance of the genetic
background including both additive and non-additive effects has repeatedly
been documented in inbred mice [17, 18].
Whereas it is widely accepted that several different genes contribute to
obesity, the respective implications appear less clear. Evidently, these genes
would in some way influence energy intake and/or expenditure. The complexity
of the genetic basis of obesity applies both from a metabolic and behavioral
perspective. Behavioral genetic research has convincingly demonstrated that
approximately 50% of the variance of diverse complex quantitative behaviors is
genetically determined [19]. Accordingly, eating a high fat diet or exercising
too little cannot only be viewed as having an environmental basis. Instead, gene
variants predisposing an individual to choose such a diet or predisposing to
physical inactivity need to be considered. Indeed, both the macronutrient intake
[20] and activity levels [21] have been shown to be genetically co-determined.
Whereas many different studies have addressed the potential consequences of
TV watching for childhood obesity [22], there is an absolute scarcity of studies
addressing the extent to which TV watching is heritable [23]. It appears that this
phenotype, like many other behavioral phenotypes, has an – albeit small –
heritable component.
In the light of genotype X environment interactions the recent obesity
epidemic is of considerable interest. Because the gene pool of a population cannot
change within a generation, environmental changes are presumed to be of
eminent importance [24]. Nevertheless, it should again be pointed out that these
changes can only have a major impact because according to the thrifty genotype
hypothesis our genotypes render us especially obesity prone [25]. Irrespective
of the assumed environmental basis of secular trends for both height and weight
there is no indication that heritability estimates in family studies are actually
declining in the light of the obesity epidemic.
Despite the constancy of the gene pool a genetic contribution to epidemic
obesity cannot be totally dismissed. Based on the high rate of parental obesity
observed among parents of extremely obese children we have hypothesized that
the recent increase of social stigmatization of obese individuals might actually
have led to an increase of assortative mating [26]. This mechanism could contribute
to epidemic obesity particularly by affecting the upper tail of the BMI
distribution. In this context it is worthwhile pointing out that the most dramatic
secular BMI increments in children and adolescents have been detected in the
overweight and obese range [27, 28].
Recently, epigenetic phenomena have been invoked to contribute to the
obesity epidemic. Indeed, it is conceivable that modern day living might affect
methylation patterns of specific genes which in turn increase the risk of obesity.
On a speculative basis, transmission of such patterns via the germ line cannot
be ruled out. Cloned mice share many characteristics consistent with obesity,
which are however not transmitted to the offspring [29].
0 komentar:
Posting Komentar