Behavior Genes
reared together. The pattern, however, is
rather different in adults and, surprisingly,
it is only comparatively recently that much
attention has been focused on this. It turns
out that biologically unrelated individuals
who are raised as siblings have an average
correlation of zero for general intelligence
when measured as adults. Twin studies
similarly show that the heritability of IQ
steadily increases in adolescents compared
with children and in adults compared with
adolescents. At the same time, the effect
of shared environment decreases. Very re-
cent studies of twins in late life show that
the nondemented elderly have the high-
est heritability for general intelligence of
all age groups at around 60% with no
evidence of a shared environmental effect.
In summary, quantitative genetic stud-
ies have recently converged on a consensus
that general intelligence is at least moder-
ately heritable and perhaps highly herita-
ble, particularly in late adulthood. Much
interest has now begun to shift towards
the goal of actually being able to locate
and identify genes involved in cognitive
Molecular studies: searching for ‘‘IQ-QTLs’’
It is broadly accepted from the pattern
of inheritance of general cognitive ability
that it is likely to be a highly polygenic trait.
That is, there is likely to be a contribution
from many genes each of which on its
own has only a very small effect. If
this assumption is correct, then linkage
studies will be very difFcult and therefore
almost all the attention has focused on
allelic association. A pioneering study was
by Plomin and colleagues who in 1995
published results on 100 polymorphisms
selected on the basis that they were at
or near genes that were expressed in the
brain. A comparison of allele distributions
was made in three groups of subjects
showing high, middle, and low scores on
IQ tests and positive results were followed
up on a replication sample. Although
Fndings including a locus on chromosome
6p and an association with a mitochondrial
polymorphism, these failed to replicate on
an independent sample.
The same group then embarked on an
ambitious project attempting to search the
whole genome, chromosome by chromo-
some, for linkage disequilibrium using
a highly polymorphic set of microsatel-
lite markers. Plomin and colleagues also
adopted an innovative approach using
DNA pooling in which the DNA from
very high IQ subjects was combined in
one pool and this was compared with the
DNA from a pool of subjects with average
IQ. The aim here was to achieve a rapid
way of testing many markers. Thus, the
initial study searching for genes having an
effect on IQ used around 2000 markers
on 200 unrelated subjects with high IQ
and 100 with average IQ. This would have
entailed 600 000 individual genotypings,
but with DNA pooling of the high-IQ and
average-IQ groups, the initial genotyping
was reduced to a more manageable 4000.
Positive results were then subsequently
followed up by individual genotyping. To
date, there are no deFnite positive asso-
ciations that have been discovered with
general cognitive ability that have repli-
cated on all samples. However, there have
been some intriguing, suggestive Fnd-
ings, for example, of an association with
a variant in the insulin-like growth fac-
tor receptor 2 gene (
). While these
results may be worth pursuing further,
it should be emphasized that
not, as some popular press reports sug-
gested, the ‘‘gene for’’ IQ. In fact, the
association reported by the group
of Plomin et al. appears at best to account
previous page 604 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online next page 606 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online Home Toggle text on/off