Our genes influence more than just our raw intelligence and problem-solving abilities.
The latest discoveries in the genetics of intelligence are showing how our DNA shapes how we learn, our educational attainment, occupational success, and even our health outcomes.
Genetic research is revealing the deep connections between inherited cognitive capacities and multiple aspects of life.
Key Facts:
- Inherited DNA differences account for 50% of differences in intelligence test scores. This is known from decades of twin and family studies.
- Recent large genome-wide association studies (GWAS) have identified common DNA variants that account for 10-20% of the genetic influence on intelligence.
- Polygenic scores aggregate the effects of thousands of DNA variants to predict individual differences in intelligence from DNA alone. The latest polygenic scores predict 4% of variance in intelligence, and over 10% when combined with polygenic scores for years of education.
- Polygenic scores for intelligence predict outcomes from early childhood to old age, and across diverse cognitive tests, educational achievements, occupation, and health.
Source: Nat Rev Genet.
The Complex Genetics of Intelligence
Intelligence is complex.
It emerges from the interplay of inherent capacities to learn, reason, and solve novel problems with environmental experiences and exposures.
Twin studies estimate that about half of the differences between individuals in IQ test performance can be attributed to genetic factors.
But pinpointing which genes are responsible has been challenging.
Early attempts at finding intelligence genes faltered.
The effect sizes of individual genetic variants are tiny, accounting for less than 0.05% of variance each.
Tens of thousands of DNA differences are likely involved.
Genome-wide association studies (GWAS) were a game-changer.
By scanning markers across the entire genomes of large samples, GWAS could detect very small effects that add up.
This approach works for any complex trait.
One twist of fate gave intelligence research an unlikely boost – huge GWAS of a correlated trait, years of educational attainment.
The Genetic Overlap Between Education and IQ
Years of education is a simple demographic variable collected in many health studies.
GWAS of this proxy for intelligence yielded hits that also predict cognitive test performance.
The genes affecting how long someone stays in school overlap substantially with genes that affect intelligence more directly.
In 2013, the first large GWAS of educational attainment found 3 significant genetic variants.
A “polygenic score” based on all variants, not just the significant ones, explained 2% of variance in years of education.
In 2016, a larger GWAS found 74 significant variants, and a polygenic score predicting 3% of variance.
An astounding new GWAS with over 1 million participants has identified over 1000 significant genetic associations for education.
Its polygenic score explains more than 10% of variance in educational attainment.
These genetic results for an “educational achievement polygenic score” also predict intelligence test performance even better than they predict years of education itself.
The 2016 polygenic score predicts 4% of variance in intelligence – more than twice the 3% of variance it predicts in the education years GWAS target trait.
The huge new GWAS is expected to yield over 10% prediction for intelligence, which accounts for 20% of the 50% heritability from twin studies.
Hitting the Limits of SNP Heritability
But why can polygenic scores only predict 25-50% of twin study heritability estimates?
There are two types of “missing heritability.”
Polygenic scores are limited by SNP heritability, the variance explained by measured DNA variants on genotyping arrays used in GWAS.
For intelligence, SNP heritability is about 25%.
The difference between the 4% prediction from current polygenic scores and the 25% SNP heritability can be closed by larger GWAS samples and multivariate methods.
However, breaking through the 25% ceiling to the 50% twin study heritability will require incorporating rare genetic variants, epistasis (gene-gene interactions), and gene-environment interplay.
Whole genome sequencing and new multivariate methods will be needed to fully characterize the complex genetics of intelligence.
But polygenic scores have already yielded ample predictive power to transform intelligence research.
Missing heritability gaps for intelligence:
- Twin study heritability is 50%.
- SNP heritability from genome-wide genotypes is 25%.
Current polygenic score prediction is 4-10%.
Transforming Intelligence Research with Polygenic Scores
Polygenic scores are game changers for studying the genetics of intelligence and its impacts on human lives.
Scores derived from GWAS of adult IQ can predict test performance from early childhood with equivalent accuracy.
The scores are stable lifelong markers of inherited cognitive propensities.
Polygenic scores also unlock powerful multivariate genetic analyses.
For example, despite general effects across cognitive abilities, genetic contributions to verbal, spatial, memory and other specific cognitive skills can be parsed out.
Polygenic scoring provides a breakthrough for studying gene-environment interplay.
DNA differences between individuals predict differences in their environmental exposures – smart, motivated kids will seek out enrichment and advance education.
Polygenic scores can index genetic factors in order to isolate environmental effects.
For example, school achievement gaps between low and high socioeconomic status students are not moderated by polygenic scores, indicating equivalent genetic influences.
Real-World Impacts: IQ & Genetics
Now that polygenic scores reliably predict academic achievement and intelligence, researchers are examining their impacts in the real world.
For instance, students at elite private schools have higher polygenic scores on average than those at public schools, indicating genetic factors influence access to privileged education.
Intergenerational social mobility also reflects polygenic score differences.
As DNA tests become routine consumer products, our growing knowledge about genetic contributions to individual cognitive differences raises ethical challenges.
How can we respect inherent DNA differences between children growing up in the same family?
How do we avoid genetic determinism or discrimination?
Policy will need to catch up with the science.
The future is bright for illuminating the deep connections between our genomic blueprint and the functioning of the remarkable human mind.
Polygenic scores for intelligence will become more powerful and find wider applications as bigger genome-wide datasets are assembled.
We face an exciting journey of discovery into how inherited DNA variations shape learning abilities, educational attainment, career success, and longevity – the very trajectories of our lives.
References
- Study: The new genetics of intelligence
- Authors: Plomin & von Stumm (2018)
