Professional Documents
Culture Documents
Deary Intelligence
Deary Intelligence
R673
Primer
Intelligence
Ian J. Deary
for young children are typically lower, increasing the sample sizes has Education
whereas estimates for adults are brought substantially more genome- People who score better on intelligence
higher (up to 70–80%). There might wide significant hits, and these are tests tend to stay longer in education,
be some small decline in heritability accounting for ever-higher proportions to gain higher-level qualifications, and
in old age, but it remains high. Some of the phenotypic variance. Also being to perform better on assessments
studies that include many diverse tests pursued are strategies whereby the of academic achievement. Some of
of mental abilities have computed the GWAS of phenotypes linked with the correlations between intelligence
heritabilities of g and the more specific intelligence are being used to create, scores at the end of primary school
domains of intelligence. They find that in separate samples, polygenic risk and academic results some years
g is highly heritable and that there is scores for the phenotype. These later are high, suggesting that it is not
less genetic influence that is specific to polygenic risk scores are then tested just a matter of education boosting
each domain. There is some evidence, for association with intelligence. For intelligence. Also, educational
though mixed, that the heritability of example, using GWAS data from the attainment has a moderately high
intelligence is higher among more Psychiatric GWAS Consortium, a heritability, and a strong genetic
affluent people when compared with polygenic risk score was created for correlation with intelligence. On the
more deprived individuals. schizophrenia in separate samples of other hand, there is also evidence
Within the range of normal cognitive older people with intelligence data, that education can provide a boost to
abilities — that is, apart from studies none of whom had suffered from scores on tests of complex thinking,
of learning difficulties — the molecular schizophrenia. Those with higher and some of these increments last into
genetic study of intelligence is still polygenic risk for schizophrenia old age. Therefore, there is probably
quite new. There have been almost tended to have lower cognitive ability a bidirectional causal association
no well-replicated associations in old age and also had larger relative between intelligence and education.
between candidate genetic variants declines in cognitive ability between
and intelligence. An exception is the childhood and old age. Further studies Social status and social mobility
APOE gene: people with one or two e4 will extend such analyses to other People who score better on intelligence
alleles of this gene tend to have lower phenotypes with which intelligence is tests tend to go into more professional
cognitive ability in old age, and tend associated, to establish the extent of occupations (typically those with
to decline more in cognition across their genetic correlations. higher status) and to perform better
their lifetimes, than those who lack With regard to the environment, twin in the workplace. There is a positive
e4. Genome-wide association studies studies suggest that the contribution association between intelligence test
(GWAS) of intelligence have not yet of shared environment to intelligence scores in childhood and social position
revealed any additional variants with differences is small, even negligible, by later in life: people who score higher
genome-wide significance. adulthood, and that that which is non- tend to be in more professional jobs,
There is, though, molecular genetic genetic is largely due to non-shared to live in less deprived areas, and to
evidence that some variance in environment and measurement error. have higher incomes. The association
intelligence differences is detected is not perfect. Results show that, when
by common single nucleotide Brain correlates of intelligence it comes to attained social position
polymorphisms (SNPs). The application differences in maturity, intelligence, education
of the genetic complex trait analysis Beyond the general finding that and parental background all count to
(GCTA) method to intelligence there is some genetic variation in some extent. That is, there is some
differences in childhood and older age intelligence, there is a modest (~0.30) meritocracy and intelligence-driven
found that between a quarter and a half correlation between intelligence social mobility, and there is also some
of the variance could be accounted for test scores and overall brain size. social inertia. There is some evidence
by variants in linkage disequilibrium As yet, it is not understood what for a so-called gravitational hypothesis:
with common SNPs. GCTA creates it is about bigger brains that is that intelligence in youth relates more
a genetic relationship matrix among associated with being brighter. strongly to occupational and social
unrelated individuals and calculates There is a similar-sized correlation position later rather than earlier in
the association between this genetic between general intelligence and the adulthood.
similarity and phenotypic similarity. general integrity of the brain’s white
This analysis did not identify the causal matter, as measured using diffusion Health, illness, and death
variants; it suggested that intelligence tensor magnetic resonance brain People who score better on intelligence
is highly polygenic, with large numbers imaging. This correlation is largely tests tend to make healthier lifestyle
of variants of small effect sizes. accounted for by people’s differences and dietary choices, to have better
Bivariate GCTA analysis has shown in speed of processing. There is health, to be less likely to have chronic
that the genetic correlation between some support from brain imaging and illnesses like cardiovascular disease,
intelligence measured in childhood and electroencephalographic research that and to live longer. These findings have
old age in the same individual is high; cleverer brains are more efficient. been the result of the field of cognitive
to a substantial extent, the same genes epidemiology, which is little more than
cause higher intelligence in childhood Consequences of intelligence a decade old. Some of the studies in
and older age. differences this field have been heroic in size and
Current research is accumulating A reason to take intelligence differences duration: some are of sample sizes of
larger sample sizes for larger GWAS seriously is that scores on intelligence over 1,000,000 people, and some have
of intelligence; for the complex traits tests are associated with a number of intelligence test data in childhood and
of height and obesity, for example, important life outcomes. then data on health information up to
Current Biology Vol 23 No 16
R676
more than 60 years later. Just focusing on the variance shared by all tests, and These tend to show that boys and girls
on the intelligence–death associations, also some specific effects in addition have about the same mean level of
this applies to mortality from all causes, to that, principally on the domains of intelligence. They also show that, for g,
to cardiovascular deaths, to suicides processing speed and memory. the boys have greater variance: there
and homicides, and to accidental Not everyone experiences the same is an excess of boys at the lower and
deaths, but probably not to deaths rate of cognitive decline, and there is upper ends of the intelligence scale.
from cancer. The associations between a growing interest in the genetic and There continues to be discussion
intelligence in youth and health and environmental (biological and social) about the so-named Flynn effect,
survival into old age are not explained determinants of people’s differences whereby the absolute scores on
by parental social class. There is some in age-related cognitive changes. Not intelligence tests have been rising
statistical mediation of the association many of these determinants are well since testing started in the early-to-mid
by education and the person’s own replicated. Some of the more solid 20th century. The extent of the rise, its
social class in mid-life, though it is not evidence exists for the following being geographical distribution in the world,
clear whether this is informative about cognitively protective: not having the and especially its causes are all still
the mechanistic pathways involved or APOE e4 allele, being physically more being studied. Some hypothesise that
whether education and occupational active and fit, and not smoking. better nutrition might explain some of
social class are, in part, acting as For biologists, apart from helping the increase, and others put it down
proxies for intelligence. psychologists to seek more to society’s making more accessible
Currently, there are four possible determinants of healthy cognitive and emphasizing the skills tested by
accounts of the intelligence versus ageing, there are two theoretical ideas intelligence tests.
health/death associations: that they that might be attractive in the study There is interest in finding
are associated because, even in youth, of the ageing of intelligence. The first interventions that might boost
intelligence is capable of indexing is the idea that some people have intelligence. On the biological side there
some general, underlying bodily system cognitive reserve such that their brains is research showing that breast feeding
integrity; that the intelligence test are better able to withstand the insults is associated with a sizeable advantage
scores detected some pathology even of age and illness. Researchers in this in intelligence later in childhood.
in youth; that intelligence is associated area write about the possibility of there However, there is also some evidence
with later health choices and lifestyles; being passive brain reserve and active that this is explained by the higher
and that intelligence acts as a selection reserve. Passive reserve refers to the intelligence scores of the mothers
variable into safer occupational and possibility of there being some aspects who tend to breastfeed. There is still
social environments. These are being of brain structure — maybe even as unresolved researching and discussion
tested at present. prosaic as just having a bigger brain — of the possible social boosters of
that provides the reserve. Active intelligence. For example, adoption
The ageing of intelligence reserve refers to the possibility that, in from a deprived to a more affluent
Intelligence has an important place response to an insult, some people’s setting is reported to be associated
in the world’s changing demographic brains are more flexible in reorganizing with an intelligence advantage. There
structure. Especially in so-called networks to regain or retain cognitive is still debate about the effectiveness
developed countries, populations functions. The other idea in research of intensive intervention programmes
have a growing proportion of older on the ageing of intelligence is the early in life, and whether any cognitive
people, greater absolute numbers of common cause hypothesis. This is built advantages last or whether advantage
them, and people are living longer. upon some empirical findings which accrues to social rather than cognitive
As well as bodily changes with age, suggest that the age-related decline of skills.
cognitive capabilities decline too. different bodily systems is correlated; Human intelligence is important; it
There are declines in cognitive function that is, people who are experiencing matters in our lives. Understanding the
even among people who do not faster cognitive declines might also be biology of intelligence differences could
develop dementia. Not all cognitive experiencing faster declines in sensory help to ameliorate declines in cognitive
functions decline at the same rate. and some physical functions. In so function.
Some cognitive functions — often far as this occurs, researchers have
referred to as markers of crystallized sought possible common causes that Further reading
intelligence — hold up well with age. might provide the mechanisms or at Deary, I.J. (2012). Intelligence. Annu. Rev. Psychol.
63, 453–482.
These include vocabulary and general least indicators of them. They have, Deary, I.J., Penke, L., and Johnson, W. (2010). The
and specific knowledge. The cognitive for example, considered inflammation, neuroscience of human intelligence differences.
Nat. Rev. Neurosci. 11, 201–211.
functions that tend to decline are oxidative stress, telomere length, and Hunt, E. (2011). Human Intelligence (Cambridge:
called fluid intelligence. These tend the hypothalamic-pituitary-adrenal axis. Cambridge University Press).
to involve on-the-spot thinking with Nisbett, R. E., Aronson, J., Blair, C., Dickens, W.,
Flynn, J., Halpern, D. F., and Turkheimer, E.
novel materials, and in situations in Other issues in intelligence research (2012). Intelligence: new findings and theoretical
which past knowledge is of limited There are still findings and developments. Am. Psychol. 67, 503–504.
Salthouse, T.A. (2010). Major Issues in Cognitive
help. This includes abstract reasoning, controversies about sex and race Ageing (Oxford: Oxford University Press).
spatial abilities, processing speed, and differences in research. With respect to
working and other types of memory. the former, the better data come from
Centre for Cognitive Ageing and Cognitive
The empirical data show that, when the massive samples of children tested
Epidemiology, Department of Psychology,
various fluid-type cognitive functions in school settings, where there has University of Edinburgh, 7 George Square,
are studied, we find the hierarchy been either full-population testing or Edinburgh EH8 9JZ, Scotland, UK.
again. That is, age has a negative effect the testing of representative samples. E-mail: i.deary@ed.ac.uk