Review Article

Abstract

Intelligence — the ability to learn, reason and solve problems — is at the forefront of behavioural genetic research. Intelligence is highly heritable and predicts important educational, occupational and health outcomes better than any other trait. Recent genome-wide association studies have successfully identified inherited genome sequence differences that account for 20% of the 50% heritability of intelligence. These findings open new avenues for research into the causes and consequences of intelligence using genome-wide polygenic scores that aggregate the effects of thousands of genetic variants.

Key points

  • Until 2017, genome-wide polygenic scores derived from genome-wide association studies (GWAS) of intelligence were able to predict only 1% of the variance in intelligence in independent samples.

  • Polygenic scores derived from GWAS of intelligence can now predict 4% of the variance in intelligence.

  • More than 10% of the variance in intelligence can be predicted by multipolygenic scores derived from GWAS of both intelligence and years of education. This accounts for more than 20% of the 50% heritability of intelligence.

  • Polygenic scores are unique predictors in two ways. First, they predict psychological and behavioural outcomes just as well from birth as later in life. Second, polygenic scores are causal predictors in the sense that nothing in our brains, behaviour or environment can change the differences in DNA sequence that we inherited from our parents.

  • Polygenic scores for intelligence can bring the powerful construct of intelligence to any research in the life sciences without having to assess intelligence through the use of tests.

Main

Life is an intelligence test. During the school years, differences in intelligence are largely the reason why some children master the curriculum more readily than other children. Differences in school performance predominantly inform prospects for further education, which in turn lead to social and economic opportunities such as those related to occupation and income. In the world of work, intelligence matters beyond educational attainment because it involves the ability to adapt to novel challenges and tasks that describe the different levels of complexity of occupations. Intelligence also spills over into many aspects of everyday life, such as the selection of romantic partners and choices about health care1. This is why intelligence — often called general cognitive ability2 — predicts educational outcomes3, occupational outcomes4,5 and health outcomes6 better than any other trait. It is also the most stable psychological trait, with a Pearson correlation coefficient of 0.54 from 11 to 90 years of age7. Box 1 describes what intelligence is and how it is assessed.

During the past century, genetic research on intelligence was in the eye of the storm of the nature–nurture debate in the social sciences8,9. In the 1970s and 1980s, intelligence research and its advocates were vilified10,11,12. The controversy was helpful in that it raised the quality and quantity threshold for the acceptance of genetic research on intelligence. As a result, bigger and better family studies, twin studies and adoption studies have amassed a mountain of evidence that consistently showed substantial genetic influence on individual differences in intelligence13. Meta-analyses of this evidence indicate that inherited differences in DNA sequence account for about half of the variance in measures of intelligence14.

These studies and applications in neuroscience15 were already pushing intelligence research towards rehabilitation when it was thrust to the forefront of the DNA revolution 4 years ago by genome-wide association studies (GWAS) focused on a very different variable — years of education. In this Review, we discuss early attempts to find the inherited genetic differences that account for the substantial heritability of intelligence, and a twist of fate involving GWAS on years of education, before discussing the results of recent large GWAS on intelligence. The second half of this Review focuses on genome-wide polygenic scores (GPSs) for intelligence that aggregate the effects of thousands of DNA variants associated with intelligence across the genome (see Box 2 for how GPSs are constructed). We illustrate how GPSs for intelligence will transform research on the causes and consequences of individual differences in intelligence before ending with a discussion of societal and ethical implications. We do not discuss other important research related to intelligence, such as evolutionary research16,17 and neuroscience research15, in order to focus on the role of GPSs in the new genetics of intelligence.

Finding the heritability of intelligence

Similar to results for many other complex traits, early results for intelligence were disappointing for more than 100 candidate gene studies18 and for seven GWAS19,20,21,22,23,24,25. From the 1990s until 2017, no replicable associations were found. GPSs from these early GWAS, which we refer to collectively as 'IQ1', predicted only 1% of the variance of intelligence in independent samples. It became clear that the problem was power: the largest effect sizes of associations between individual single-nucleotide polymorphisms (SNPs) and intelligence were extremely small, accounting for less than 0.05% of the variance of intelligence. The average effect size of the tens of thousands of SNPs needed to explain the 50% heritability of intelligence is of course much lower. If the average effect size is 0.005%, 10,000 such SNP associations would be needed to explain the 50% heritability of intelligence. To achieve sufficient power for detecting such small effect sizes (that is, power of 80%, P = 0.05, one-tailed), sample sizes greater than 250,000 are required. Early IQ GWAS had sample sizes from 18,000 to 54,000, which seemed large at the time but were not sufficiently powered to detect such small effects.

Breakthrough for years of education. A breakthrough for intelligence research came from the unlikely variable of the number of years spent in full-time education, often referred to as educational attainment. Because 'years of education' is obtained as a demographic marker in nearly every GWAS, it was possible to accumulate sample sizes with the necessary power to detect very small effect sizes26. Its relevance to intelligence is that years of education is highly correlated phenotypically (0.50) and genetically (0.65) with intelligence27.

In 2013, a meta-analytic GWAS analysis of years of education yielded three genome-wide significant SNP associations in a sample of 125,000 individuals from 54 cohorts28. These associations could be replicated in independent samples29. The largest effect size associated with an individual SNP accounted for a meagre 0.02% of the variation, equivalent to about 2 months of education. Although individual SNPs of such minuscule effect size are fairly useless for prediction, a GPS based on all SNPs regardless of the strength of their association with years of education predicted 2% of the variance in years of education in independent samples28,29. We refer to this GPS as 'EA1' (where EA stands for educational attainment).

Spurred on by this success, in 2016, a second meta-analytic GWAS analysis with a sample size of 294,000 identified 74 significant loci30. This analysis produced a GPS, EA2, that predicted 3% of the variance in years of education on average in independent samples30. Surprisingly, GPSs for years of education predicted more variance in intelligence than they predicted for the GWAS target trait of years of education27. For example, the EA2 GPS predicts 3% of the variance in years of education but it predicts 4% of the variance in intelligence30. A third GWAS currently in progress includes more than 1 million participants, making it the largest GWAS for any trait to date. Preliminary results from this GWAS have identified more than 1,000 significant associations and a GPS, EA3, that predicts more than 10% of the variance in years of education in independent samples31 (P. D. Koellinger, personal communication). Hence, the EA3 GPS is expected to predict more than 10% of the variance in intelligence. The effect size of the EA3 GPS for predicting intelligence is likely to rival that of family socio-economic status, which is indexed by parents' years of education. Across studies, the correlation value for parents' education with children's intelligence is 0.30, implying that it accounts for 9% of the variance in children's intelligence32. However, this association is confounded by genetics because children inherit the DNA differences that predict their intelligence from their parents. Furthermore, parental phenotypes, such as education, estimate only an average association for offspring, whereas GPSs predict intelligence for each individual.

Large-scale genome-wide association studies of intelligence. In 2017, the largest GWAS meta-analysis of intelligence, which included 'only' 78,000 individuals, yielded 18 genome-wide significant regions33. A GPS (IQ2) derived from these GWAS results finally broke the 1% barrier of previous GWAS of intelligence by predicting 3% of the variance of intelligence in independent samples. However, IQ2 still has less predictive power than the 4% of the variance explained by the EA2 GPS.

A follow-up GWAS meta-analysis reached a sample size of 280,000 with the inclusion of cognitive data from the UK Biobank. This GWAS analysis increased the number of identified genome-wide significant regions from 18 to 206 (Ref. 34). A GPS derived from these GWAS analyses, IQ3, predicts about 4% of the variance of intelligence in independent samples34. Other meta-analytic GWAS using the UK Biobank data, which were released in June 2017 and are publicly available, yield similar results35.

These IQ and EA GPS results are summarized in Fig. 1. It might seem disappointing that the increase of the intelligence GWAS sample sizes from 78,000 to 280,000 boosted the predictive power of the IQ GPS only from 3% to 4%. However, this result is parallel to GWAS results for years of education: after increasing sample sizes from 125,000 to 294,000, the variance in years of education predicted by the EA GPS grew only from 2% to 3%. Note that the predictive power of the EA GPS jumped to more than 10% of the variance in preliminary analyses of the latest meta-analytic GWAS (EA3) with a sample size of more than 1 million31 (P. D. Koellinger, personal communication). We can expect a similar jump in the predictive power of the IQ GPS when the sample size for GWAS meta-analyses of intelligence exceeds 1 million. However, it is more difficult to obtain very large sample sizes for intelligence, which has to be tested, than for years of education, which can be assessed with a single self-reported item.

Figure 1: Variance explained by IQ GPSs and by EA GPSs in their target traits as a function of GWAS sample size.
Figure 1

Genome-wide polygenic score (GPS) prediction of intelligence (IQ) and educational attainment (EA) increased linearly with sample size. The predictive power of GPSs derived from genome-wide association studies (GWAS) of intelligence has increased in the past 2 years from 1% to 4%. The latest EA GPS, EA3, predicts more than 10% of the variance in intelligence (P. D. Koellinger, personal communication) — more than twice as much as the latest IQ3 GPS. Extrapolating from the results of EA3 with a sample size of more than 1 million, we predict that more than 10% of the variance in intelligence will be predicted from IQ GPSs derived from a GWAS of intelligence with a sample size of 1 million. IQ1 (Ref. 22): n = 54,000, r2 = 0.01. IQ2 (Ref. 33): n = 78,000, r2 = 0.03. IQ3 (Ref. 34): n = 280,000, r2 = 0.04. EA1 (Ref. 28): n = 125,000, r2 = 0.02. EA2 (Ref. 30): n = 294,000, r2 = 0.03. EA3 (Ref. 31): n = 1,100,000, r2 >0.10.

Missing heritability. It is possible to use multiple GPSs to boost the power to predict intelligence by aggregating GPSs in a way analogous to aggregating SNPs to produce GPSs (Box 3). Including the EA2 GPS, IQ2 GPS and other GPSs in this multivariate way can already predict up to 7% of the variance in intelligence36,37. Multivariate GPS analyses that incorporate multiple GPSs in addition to the EA2 GPS and IQ2 GPS will explain substantially more than 10% of the variance in intelligence, which is more than 20% of the 50% heritability of intelligence.

Nonetheless, 10% is a long way from the heritability estimate of 50% obtained from twin studies of intelligence14. This gap is known as 'missing heritability', which is a key genetic issue for all complex traits in the life sciences38 (Box 4). The current limit for the variance that can be predicted by GPSs is SNP heritability, which estimates the extent to which phenotypic variance for a trait can be explained by SNPs across the genome without identifying specific SNP associations. For intelligence, SNP heritability is about 25%34,39. It is safe to assume that GPSs for intelligence using current SNP chips can approach the SNP heritability limit of 25% by amassing ever-larger GWAS samples and by using multitrait GWAS that include traits related to intelligence, such as years of education. However, breaking through this ceiling of 25% SNP heritability to the 50% heritability estimated from twin studies — assuming that twin studies yield accurate estimates of the total variance explained by inherited DNA differences — will require different technologies, such as whole-genome sequencing data that include rare variants, not just the common SNPs used on current SNP chips.

GPSs in intelligence research

A bottom-up approach to intelligence focused on specific genes will be difficult for three reasons. First, genetic effects are extremely pleiotropic. Second, many hits are in intergenic regions, which means that there are no 'genes' to trace through the brain to behaviour. Third, the biggest hits have minuscule effects — less than 0.05% of the variance — which means that hundreds of thousands of SNP associations are needed to account for the 50% heritability estimated by twin studies. A systems biology approach to molecular studies of the brain is needed that is compatible with this extreme pleiotropy and polygenicity40.

By contrast, the top-down approach of GPSs that aggregate thousands of these tiny effects is already transforming research on intelligence41. Unlike quantitative genetic studies that require special samples, such as twins, or GWAS that require very large samples in the hundreds of thousands, GPSs can be used to add a genetic dimension to any research with modest sample size. For example, a GPS for intelligence that predicts 10% of the variance needs a sample size of only 60 to detect its effect with 80% power (P = 0.05, one-tailed).

GPSs are unique predictors in the behavioural sciences. They are an exception to the rule that correlations do not imply causation in the sense that there can be no backward causation when GPSs are correlated with traits. That is, nothing in our brains, behaviour or environment changes inherited differences in our DNA sequence. A related advantage of GPSs as predictors is that they are exceptionally stable throughout the lifespan because they index inherited differences in our DNA sequence. Although mutations accrue in the salivary and blood cells used to collect DNA, these mutations would not be expected to systematically change the thousands of inherited SNPs that contribute to a GPS.

In other words, GPSs derived from GWAS of any trait at any age would be expected to have a correlation near 1.0 when GPSs are constructed from DNA obtained at birth and in adulthood for the same individual, although we are not aware of any empirical evidence relevant to this prediction. If GPSs for individuals do not change during the lifespan, a GPS derived from GWAS of intelligence in adulthood will predict adult intelligence just as well from DNA obtained at conception or birth as from DNA obtained in adulthood. By contrast, intelligence tests at birth cannot predict intelligence at age 18 years. At 2 years of age, infant intelligence tests predict less than 5% of the variance of intelligence in late adolescence32,42.

GPSs are unbiased in the sense that they are not subject to training, faking or anxiety. They are also inexpensive, costing less than US$100 per person. This expense would not be incurred specifically to predict intelligence; the same SNP chip genotype information used in GWAS can be used to create GPSs for hundreds of disorders and traits, one of which is intelligence.

GPSs for intelligence will open new avenues for research into the causes and consequences of intelligence. Three examples are developmental change and continuity, multivariate links between traits and gene–environment (GE) interplay. A critical requirement for capitalizing on these opportunities is to make the ingredients for GPSs publicly available — that is, GWAS summary-level statistics (Box 5).

Developmental research. One of the most interesting developmental findings about intelligence is that its heritability as estimated in twin studies increases dramatically from infancy (20%) to childhood (40%) to adulthood (60%), whereas age-to-age genetic correlations are consistently high43,44. What could account for this increasing heritability despite unchanging age-to-age genetic correlations? Twin studies suggest that genetic effects are amplified through GE correlation as time goes by45. That is, the same large set of DNA variants affects intelligence from childhood to adulthood, resulting in high age-to-age genetic correlations, but these DNA variants increasingly have an impact on intelligence as individuals select environments correlated with their genetic propensities, leading to greater heritability of intelligence.

Developmental hypotheses about high age-to-age genetic correlations and increasing heritability can be tested more rigorously and can be extended through the use of GPSs. Does the variance explained by GPSs for intelligence increase from childhood to adolescence to adulthood? Are the correlations between GPSs at these ages consistently high?

High age-to-age genetic correlations for intelligence imply that GWAS of adults should predict intelligence in childhood. The EA2 GPS30, currently the best genetic predictor of intelligence until the EA3 GPS becomes available, was derived from a GWAS meta-analysis of years of education in adults who had completed their education. Nonetheless, the EA2 GPS predicts 2% of the variance in intelligence at age 7 years, 3% at age 12 years and 4% at age 16 years in a longitudinal study46.

Multivariate genetic research. Multivariate genetic research focuses on the genetic covariance between traits rather than the variance of each trait. A specific multivariate question for intelligence research is why EA GPSs predict twice as much variance in intelligence as do GPSs for intelligence itself. This question raises interesting methodological and conceptual issues (Box 6).

Multivariate genetic research is especially important for intelligence because genetic effects in the cognitive domain have been shown in twin studies to be general. That is, genetic effects correlate highly across most cognitive abilities, such as verbal and spatial abilities, as well as most educational skills, such as reading and mathematics47. A recent multivariate finding is that the EA2 GPS predicts 5% of the variance in comprehension and efficiency of reading48. This is by far the most powerful GPS predictor of reading ability because there have as yet been no large GWAS of reading with replicable results49. EA GPSs are also likely to predict other educational skills, such as mathematics, and other cognitive abilities, such as spatial ability.

EA GPSs are correlated genetically with a wider range of variables than any other GPS50. This pervasive genetic influence of EA GPSs extends to a negative genetic correlation with schizophrenia and positive genetic correlations with height51, with myopia52 and, surprisingly, with autism53. Linkage disequilibrium (LD) score regression analysis54, which uses summary GWAS statistics rather than GPSs for individuals, finds a similar pattern of results for intelligence using the IQ2 GWAS: the negative genetic correlation with schizophrenia (−0.20) and the positive genetic correlations with height (0.10) and autism (0.21)33. The same LD score regression analysis33 found that intelligence significantly correlated genetically with many other traits, including Alzheimer disease (−0.36), smoking cessation (−0.32), intracranial volume (0.29), head circumference in infancy (0.28), depressive symptoms (−0.27), attention-deficit–hyperactivity disorder (−0.27), having ever smoked (−0.23), longevity (0.22) and, of course, years of education (0.70).

Despite this evidence for ability-general genetic effects, genetic correlations across cognitive abilities and educational skills are not 1.0, which implies that there are ability-specific SNP associations. An important direction for research is to identify ability-specific GPSs derived from large GWAS analyses focused on specific cognitive abilities independent of general intelligence. Preliminary analyses of this sort would be possible using existing GWAS of intelligence because most of these studies assessed multiple measures of specific cognitive abilities, which were combined to index intelligence. These data could be reanalysed in meta-analytic GWAS that focus on specific abilities included in multiple studies. However, what is needed are large GWAS focused on well-measured specific cognitive abilities, such as verbal, spatial and memory abilities and specific cognitive skills taught in schools, for example reading, mathematics and language. The pay-off from these studies will be GPSs that predict specific abilities independent of general intelligence. These ability-specific GPSs could be used to create profiles of genetic strengths and weaknesses for individuals who could be targets for personalized prediction, prevention and intervention.

In addition to investigating links between different traits, multivariate genetic research can examine genetic links between dimensional and diagnostic measures of the 'same' domain. For example, the EA2 GPS predicts reading disability just as much as reading ability, from slow readers to speed readers48. Because GPSs are always normally distributed, they will show that there are no aetiologically distinct common disorders, only continuous dimensions55. This is also true for very low and for very high intelligence46. Even extremely high intelligence is only quantitatively, not qualitatively, different genetically from the normal distribution56,57. The exception is severe intellectual disability, which is genetically distinct from the rest of the distribution of intelligence58 and affected by rare, often de novo mutations with large effects59.

Research on gene–environment interplay. The high heritability of intelligence should not obscure the fact that heritability is considerably less than 100%. Research using genetically sensitive designs has led to one of the most important findings about environmental influence on intelligence. Intelligence has always been known to run in families, but it was assumed that this family resemblance was due to nurture, called 'shared family environmental influence'. That is, siblings were thought to be similar in intelligence because they grew up in the same family and attended the same schools. Twin and adoption studies consistently support this assumption, but only until adolescence. After adolescence, the effect of shared family environmental influence on intelligence is negligible, which means that family environments have little effect on individual differences in the long run45,60. Family resemblance for intelligence is due to nature rather than nurture, although it should be emphasized that we are referring to the normal range of environmental influence, not the extremes such as neglect or abuse. However, little is known about the specific environmental factors that make children growing up in the same family different14.

The importance of both genetics and environment for cognitive development is a recommendation for investigating the interplay between them. GPSs for intelligence will greatly facilitate this research because they offer, for the first time, the possibility of directly assessing genetic propensities of individuals to investigate their interplay with aspects of the environment. GE interplay refers to two different concepts — GE interaction and GE correlation.

GE interaction denotes a conditional relationship in which the effects of genes on intelligence depend on the environment. For example, some twin research suggests that the heritability of intelligence is lower in low-socio-economic-status family environments and higher in high-socio-economic-status family environments61. This hypothesis predicts that GPSs for intelligence will correlate less with intelligence in environments of low socio-economic status than in those of high socio-economic status. The first test of this hypothesis using the EA2 GPS found no evidence for such an interaction46. That is, the EA2 GPS were correlated with intelligence in low-socio-economic-status just as much as in high-socio-economic-status family environments. GPSs provide a particularly powerful approach to test for GE interaction compared with twin studies62.

In contrast to GE interaction, GE correlation refers to the correlation between genetic propensities and experiences. GE correlation is the reason why most environmental measures used in the behavioural sciences show genetic influence in twin studies63. Associations between environmental measures and behavioural traits such as intelligence are also mediated in part by genetic differences. Research using GPSs is beginning to confirm these twin study findings about the 'nature of nurture' by showing, for example, that EA GPSs correlate with social mobility64 and capture covariation between environmental exposures and children's behaviour problems and educational achievement65. GE correlation provides a general model for how genotypes become phenotypes — how children select, modify and create environments correlated with their genetic propensities. GPSs will greatly advance research on GE correlation by providing an individual-specific index of the 'G' of GE interplay. GPSs will also make it possible to assess environmental influences on intelligence while controlling for genetic influences.

Implications for society

The most exciting aspect of GPSs is their potential for addressing novel, socially important questions, which we will illustrate with three recent examples from our own research. First, children in public and private schools differ in their EA2 GPSs because private schools select pupils on the basis of genetic differences in intelligence66. Second, intergenerational educational mobility reflects EA2 GPS differences67. Finally, the EA2 GPS predicts twice as much variance in educational attainment and occupational status in the post-Soviet era as in the Soviet era in Estonia, a finding compatible with the hypothesis that heritability is an index of equality of opportunity and meritocracy68.

Understanding ourselves. IQ GPSs will be used to predict individuals' genetic propensity to learn, reason and solve problems, not only in research but also in society, as direct-to-consumer genomic services provide GPS information that goes beyond single-gene and ancestry information. We predict that IQ GPSs will become routinely available from direct-to-consumer companies along with hundreds of other medical and psychological GPSs that can be extracted from genome-wide genotyping on SNP chips. The use of GPSs to predict individuals' genetic propensities requires clear warnings about the probabilistic nature of these predictions and the limitations of their effect sizes (Box 7).

Although simple curiosity will drive consumers' interests, GPSs for intelligence are more than idle fortune telling. Because intelligence is one of the best predictors of educational and occupational outcomes, IQ GPSs will be used for prediction from early in life before intelligence or educational achievement can be assessed. In the school years, IQ GPSs could be used to assess discrepancies between GPSs and educational achievement (that is, GPS-based overachievement and underachievement). The reliability, stability and lack of bias of GPSs make them ideal for prediction, which is essential for the prevention of problems before they occur. A 'precision education' based on GPSs could be used to customize education, analogous to 'precision medicine'.

A novel, socially important direction for research using IQ GPSs is to understand differences within families. First-degree relatives are on average only 50% genetically similar, which means they are on average 50% genetically different. A major impact of GPSs will be to recognize and respect these large genetic differences within families.

For scores on an intelligence test standardized to have a mean of 100 and a standard deviation of 15, the average difference between pairs of individuals who are selected randomly from the general population is 17 IQ points. The average difference between parents and offspring and between siblings is 13 IQ points69. IQ GPSs might help parents understand why their children differ in school achievement. Because GPSs are probabilistic, a low-IQ GPS does not mean that a child is destined to go no further in education than secondary school. However, it does mean that the child is more likely to find academic learning more difficult and less rewarding than a sibling with a high-IQ GPS.

Ethical implications. Genomic research and studies of intelligence face four principal ethical concerns: the notion of biological determinism; the potential for discrimination and stigmatization; the question of ownership of information; and the emotional impact of knowledge of one's personal genomics and intelligence. These and other ethical issues are explored in detail by the Ethical, Legal and Social Implications (ELSI) Research Programme, which is an integral part of the Human Genome Project70. In addition, recent books discuss ethical as well as scientific issues about personal genomics, specifically in relation to education71 and occupation72. Many of these ethical discussions focus on single-gene disorders, for example, Huntington disease, which has 100% penetrance. By contrast, GPSs are 'less dangerous' because they are intrinsically probabilistic, not hard-wired and deterministic like single-gene disorders. It is important to recall here that although all complex traits are heritable, none is 100% heritable. A similar logic can be applied to IQ scores: although they have great predictive validity for key life outcomes1,2,3,4,5,6, IQ is not deterministic but probabilistic. In short, an individual is always more than the sum of their genes or their IQ scores.

Issues of discrimination and stigmatization have accompanied research into genetics and intelligence from the beginning, typically because findings from both fields of study were applied to justify policies that served sociopolitical ideologies. For example, IQ testing was infamously used to differentiate European immigrants to the United States of America who arrived at Ellis Island in the early 1900s, as well as to guide eugenic ideas about sterilization in Britain and the United States of America throughout the 20th century11. It is important to acknowledge the risk of discrimination that occurs on the back of scientific findings about individual differences. However, it is equally important to realize that research does not lead directly to any policy recommendations. We must be careful not to blame the scientists or entire disciplines when their findings are used wrongly9.

Who 'owns' our genetic information? And who should decide who can access it? The question of ownership of personal data has become pivotal but also increasingly complex in our current age of information. At the same time, understanding and managing the emotional impact that stems from knowledge of our genomics and intelligence have emerged as new societal responsibilities. It is beyond the scope of this paper to elucidate these issues in the depth that they deserve, but we expect that the discussions of ethical issues that surround personal genomics will consolidate the DNA revolution.

Conclusions

Genetic association studies have confirmed a century of quantitative genetic research showing that inherited DNA differences are responsible for substantial individual differences in intelligence test scores. A reachable objective shared with all complex traits in the life sciences is to close the gap between the 10% variance in intelligence scores explained by GPSs and the SNP heritability of intelligence of about 25%. A more daunting challenge is to break through the ceiling of 25% SNP heritability to reach the 50% heritability estimated by twin studies.

Until 2016, GPSs could predict only 1% of the variance in intelligence. Progress has been rapid since then, reaching our current ability to predict 10% of the variance in intelligence from DNA alone. GPSs will soon be available that can predict more than 10% of the variance in intelligence (that is, more than 20% of the 50% heritability of intelligence estimated from twin studies) and more than 40% of the 25% SNP heritability of intelligence. This is an important milestone for the new genetics of intelligence because effect sizes of this magnitude are large enough to be "perceptible to the naked eye of a reasonably sensitive observer" (Ref. 73). With these advances in the past few years, intelligence steps out of the shadows and takes the lead in genomic research.

In addition to investigating traditional issues about development, multivariate links among traits and GE interplay, IQ GPSs will open new avenues for research into the causes and consequences of intelligence. The new genetics of IQ GPSs will bring the omnipotent variable of intelligence to all areas of the life sciences without the need to assess intelligence.

Publisher's note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. 1.

    Why g matters: The complexity of everyday life. Intelligence 24, 79–132 (1997).

  2. 2.

    et al. Genetic contributions to stability and change in intelligence from childhood to old age. Nature 482, 212–214 (2012).

  3. 3.

    , , & Intelligence and educational achievement. Intelligence 35, 13–21 (2007).

  4. 4.

    & General mental ability in the world of work: occupational attainment and job performance. J. Pers. Soc. Psychol. 86, 162–173 (2004).

  5. 5.

    Intelligence and socioeconomic success: a meta-analytic review of longitudinal research. Intelligence 35, 401–426 (2007).

  6. 6.

    et al. Childhood intelligence in relation to major causes of death in 68 year follow-up: prospective population study. Brit. Med. J. 357, 2708 (2017).

  7. 7.

    , & The stability of intelligence from age 11 to age 90 years: the Lothian birth cohort of 1921. Psychol. Sci. 24, 2361–2368 (2013).

  8. 8.

    [No authors listed] Intelligence research should not be held back by its past. Nature 545, 385–386 (2017). This editorial is a landmark in the acceptance of genetic influence on intelligence, concluding, "it's well established and uncontroversial among geneticists that together, differences in genetics underwrite significant variation in intelligence between people."

  9. 9.

    The Blank Slate: The Modern Denial of Human Nature (Penguin, 2003).

  10. 10.

    & The IQ Controversy: Critical Readings (Pantheon, 1976).

  11. 11.

    The Mismeasure of Man (W.W. Norton, 1982).

  12. 12.

    The Science and Politics of IQ (Routledge, 1974).

  13. 13.

    & Familial studies of intelligence: a review. Science 212, 1055–1059 (1981).

  14. 14.

    , , & Behavioral Genetics. 7th edn (Worth, 2017).

  15. 15.

    The Neuroscience of Intelligence (Cambridge Univ. Press, 2016).

  16. 16.

    Survival of the friendliest: Homo sapiens evolved via selection for prosociality. Annu. Rev. Psychol. 68, 155–186 (2017).

  17. 17.

    & The Evolution of Intelligence (Psychology Press, 2013).

  18. 18.

    et al. Most reported genetic associations with general intelligence are probably false positives. Psychol. Sci. 23, 1314–1323 (2012).

  19. 19.

    et al. Childhood intelligence is heritable, highly polygenic and associated with FNBP1L. Mol. Psychiatry 19, 253–258 (2014).

  20. 20.

    , , & Genome-wide quantitative trait locus association scan of general cognitive ability using pooled DNA and 500K single nucleotide polymorphism microarrays. Genes Brain Behav. 7, 435–446 (2008).

  21. 21.

    et al. Genome-wide association studies establish that human intelligence is highly heritable and polygenic. Mol. Psychiatry 16, 996–1005 (2011).

  22. 22.

    et al. Genetic contributions to variation in general cognitive function: a meta-analysis of genome-wide association studies in the CHARGE consortium (N = 53 949). Mol. Psychiatry 20, 183–192 (2015).

  23. 23.

    et al. Genome-wide association study of cognitive functions and educational attainment in UK Biobank (N = 112 151). Mol. Psychiatry 21, 758–767 (2016).

  24. 24.

    et al. A genome-wide scan of 1842 DNA markers for allelic associations with general cognitive ability: a five-stage design using DNA pooling and extreme selected groups. Behav. Genet. 31, 497–509 (2001).

  25. 25.

    et al. GWAS meta-analysis reveals novel loci and genetic correlates for general cognitive function: a report from the COGENT consortium. Mol. Psychiatry 22, 336 (2017).

  26. 26.

    & Genetics and educational attainment. Sci. Learn. 2, 1–7 (2017).

  27. 27.

    et al. Common genetic variants associated with cognitive performance identified using the proxy-phenotype method. Proc. Natl Acad. Sci. USA 111, 13790–13794 (2014). This study uses EA1 SNPs to predict intelligence, although less than 1% of the variance is predicted.

  28. 28.

    et al. GWAS of 126,559 individuals identifies genetic variants associated with educational attainment. Science 340, 1467–1471 (2013). This is the GWAS origin of EA1, which yields a GPS that predicts 1% of the variance in years of education.

  29. 29.

    et al. Replicability and robustness of genome-wide-association studies for behavioral traits. Psychol. Sci. 25, 1975–1986 (2014).

  30. 30.

    et al. Genome-wide association study identifies 74 loci associated with educational attainment. Nature 533, 539–542 (2016). This is the GWAS origin of EA2 GPS, which increases the prediction of educational attainment from 1% to 3% of the variance.

  31. 31.

    Behavior Genetics Association 47th Annual Meeting Abstracts. et al. GWAS of educational attainment – phase 3: main results [abstract]. Behav. Genet. 47, 699 (2017). This study refers to the largest GWAS of educational attainment (n = 1,100,000), which increases the power of its GPS, EA3, to predict more than 10% of the variance in the targeted trait.

  32. 32.

    & Socioeconomic status and the growth of intelligence from infancy through adolescence. Intelligence 48, 30–36 (2015).

  33. 33.

    et al. Genome-wide association meta-analysis of 78,308 individuals identifies new loci and genes influencing human intelligence. Nat. Genet. 49, 1107–1112 (2017). This is the GWAS origin of IQ2 GPS, which increases the prediction of intelligence from 1% to 3%.

  34. 34.

    et al. GWAS meta-analysis (N = 279,930) identifies new genes and functional links to intelligence. Preprint at (2017). This paper describes the largest GWAS of intelligence to date, which yields a GPS (IQ3) that predicts 4% of the variance in intelligence.

  35. 35.

    et al. Ninety-nine independent genetic loci influencing general cognitive function include genes associated with brain health and structure (N = 280,360). Preprint at (2017).

  36. 36.

    et al. Multi-polygenic score approach to trait prediction. Mol. Psychiatry (2017). This study employs a multiple-GPS approach and finds that 81 GPSs derived from well-powered GWAS predict 5% of the variance in intelligence.

  37. 37.

    , , , & A combined analysis of genetically correlated traits identifies 107 loci associated with intelligence. Preprint at (2017). This study employs multiple-trait analysis of GWAS for intelligence and finds that educational attainment and income predict 7% of the variance in intelligence in an independent sample.

  38. 38.

    et al. Finding the missing heritability of complex diseases. Nature 461, 747–753 (2009).

  39. 39.

    et al. Common DNA markers can account for more than half of the genetic influence on cognitive abilities. Psychol. Sci. 24, 562–568 (2013).

  40. 40.

    , & An expanded view of complex traits: from polygenic to omnigenic. Cell 169, 1177–1186 (2017).

  41. 41.

    Blueprint: How DNA Makes Us Who We Are (Allen Lane/Penguin, in the press). This book describes genetic research on behaviour from twin studies to the DNA revolution and its implications for science and society.

  42. 42.

    , & The stability of mental test performance between two and eighteen years. J. Exp. Educ. 17, 309–324 (1948).

  43. 43.

    et al. A twin study of the genetics of high cognitive ability selected from 11,000 twin pairs in six studies from four countries. Behav. Genet. 39, 359–370 (2009).

  44. 44.

    & Genetics and intelligence differences: five special findings. Mol. Psychiatry 20, 98–108 (2015). This article highlights five genetic findings that are special to intelligence differences, including one not mentioned in this Review — assortative mating is much greater for intelligence than for other traits.

  45. 45.

    & Explaining the increasing heritability of cognitive ability across development: a meta-analysis of longitudinal twin and adoption studies. Psychol. Sci. 24, 1704–1713 (2013).

  46. 46.

    et al. Predicting educational achievement from DNA. Mol. Psychiatry 22, 267–272 (2017). This study shows that EA2 predicts 9% of the variance in tested educational achievement at age 16, which was the strongest GPS prediction of a behavioural trait at that time.

  47. 47.

    & Generalist genes and learning disabilities. Psychol. Bull. 131, 592–617 (2005).

  48. 48.

    et al. Genome-wide polygenic scores predict reading performance throughout the school years. Sci. Stud. Read. 21, 334–349 (2017).

  49. 49.

    et al. Evaluation of results from genome-wide studies of language and reading in a novel independent dataset. Genes Brain Behav. 15, 531–541 (2016).

  50. 50.

    et al. Phenome-wide analysis of genome-wide polygenic scores. Mol. Psychiatry 21, 1188–1193 (2015).

  51. 51.

    et al. Common genetic variants explain the majority of the correlation between height and intelligence: the generation Scotland study. Behav. Genet. 44, 91–96 (2014).

  52. 52.

    et al. Phenotypic and genotypic correlation between myopia and intelligence. Sci. Rep. 7, 45977 (2017).

  53. 53.

    et al. Age-dependent pleiotropy between general cognitive function and major psychiatric disorders. Biol. Psychiatry 80, 266–273 (2016).

  54. 54.

    et al. LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat. Genet. 47, 291–295 (2015).

  55. 55.

    , & Common disorders are quantitative traits. Nat. Rev. Genet. 10, 872–878 (2009).

  56. 56.

    et al. A genome-wide analysis of putative functional and exonic variation associated with extremely high intelligence. Mol. Psychiatry 21, 1145–1151 (2016).

  57. 57.

    et al. A genome-wide association study for extremely high intelligence. Mol. Psychiatry (2017). This GWAS of intelligence uses a novel strategy to increase power — a case–control design in which the subjects were individuals with extremely high IQ from the top 0.0003 of the population (mean IQ of 170).

  58. 58.

    et al. Discontinuity in the genetic and environmental causes of the intellectual disability spectrum. Proc. Natl Acad. Sci. USA 113, 1098–1103 (2016).

  59. 59.

    , & Genetic studies in intellectual disability and related disorders. Nat. Rev. Genet. 17, 9–18 (2016).

  60. 60.

    & Why are children in the same family so different from one another? Behav. Brain Sci. 10, 1–16 (1987).

  61. 61.

    & Large cross-national differences in gene × socioeconomic status interaction on intelligence. Psychol. Sci. 27, 138–149 (2016).

  62. 62.

    et al. Socioeconomic status (SES) and children's intelligence (IQ): in a UK-representative sample SES moderates the environmental, not genetic, effect on IQ. PLOS ONE 7, e30320 (2012).

  63. 63.

    & The nature of nurture: genetic influence on "environmental" measures. Behav. Brain Sci. 14, 373–386 (1991).

  64. 64.

    et al. The genetics of success. Psychol. Sci. 27, 957–972 (2016).

  65. 65.

    et al. Widespread covariation of early environmental exposures and trait-associated polygenic variation. Proc. Natl Acad. Sci. USA 114, 11727–11732 (2017).

  66. 66.

    et al. Differences in exam performance between pupils attending different school types mirror the genetic differences between them. NPJ Sci. Learn. (in the press).

  67. 67.

    , , & Genetic influence on intergenerational educational attainment. Psychol. Sci. 28, 1302–1310 (2017). This paper describes both twin analyses and EA2 GPSs that show genetic influence on intergenerational EA.

  68. 68.

    Behavior Genetics Association 46th Annual Meeting Abstracts. , , , & Genetic influence on educational attainment and occupational status during and after the Soviet era in Estonia [abstract]. Behav. Genet. 46, 803 (2016).

  69. 69.

    & Genetics and intelligence: recent data. Intelligence 4, 15–24 (1980).

  70. 70.

    et al. The ethical, legal, and social implications program of the National Human Genome Research Institute: reflections on an ongoing experiment. Annu. Rev. Genom. Hum. Genet. 15, 481–504 (2014).

  71. 71.

    , , & Genetics, Ethics and Education (Cambridge Univ. Press, 2017).

  72. 72.

    & The Genome Factor: What the Social Genomics Revolution Reveals about Ourselves, our History, and the Future (Princeton Univ. Press, 2017).

  73. 73.

    Statistical Power Analysis for the Behavioral Sciences (Lawrence Erlbaum Associates, 1977).

  74. 74.

    Mainstream science on intelligence. Wall Street Journal (13 December 1994).

  75. 75.

    Human Cognitive Abilities: A Survey of Factor-Analytic Studies (Cambridge Univ. Press, 1993).

  76. 76.

    'General Intelligence' objectively determined and measured. Am. J. Psychol. 15, 201–292 (1904).

  77. 77.

    The g Factor: The Science of Mental Ability (Praeger, 1998).

  78. 78.

    Intelligence. Annu. Rev. Psychol. 63, 453–482 (2012). This article is an authoritative overview of intelligence research.

  79. 79.

    et al. Stability and change in intelligence from age 11 to ages 70, 79, and 87: the Lothian Birth Cohorts of 1921 and 1936. Psychol. Ageing 26, 232–240 (2011).

  80. 80.

    Developmental Influences on Adult Intelligence: The Seattle Longitudinal Study (Oxford Univ. Press, 2005).

  81. 81.

    & Schooling in adolescence raises IQ scores. Proc. Natl Acad. Sci. USA 109, 425–430 (2012).

  82. 82.

    Does the raising IQ–raising g distinction explain the fadeout effect? Intelligence 56, 65–71 (2016).

  83. 83.

    , & How can we boost IQs of "dull children"?: a late adoption study. Proc. Natl Acad. Sci. USA 96, 8790–8794 (1999).

  84. 84.

    & Is working memory training effective? A meta-analytic review. Dev. Psychol. 49, 270–291 (2013).

  85. 85.

    et al. Head Start Impact Study Final Report. Administration for Children and Families (2010).

  86. 86.

    & The future of genomics for developmentalists. Dev. Psychopathol. 25, 1263–1278 (2013).

  87. 87.

    & Dissecting the genetics of complex traits using summary association statistics. Nat. Rev. Genet. 18, 117–127 (2017).

  88. 88.

    et al. Modeling linkage disequilibrium increases accuracy of polygenic risk scores. Am. J. Hum. Genet. 97, 576–592 (2015).

  89. 89.

    et al. PRSice: polygenic risk score software. Bioinformatics 31, 1466–1468 (2015).

  90. 90.

    et al. Molecular genetic contributions to social deprivation and household income in UK Biobank. Curr. Biol. 26, 3083–3089 (2016).

  91. 91.

    et al. MTAG: Multi-Trait Analysis of GWAS. Preprint at (2017).

  92. 92.

    et al. LD Hub: a centralized database and web interface to perform LD score regression that maximizes the potential of summary level GWAS data for SNP heritability and genetic correlation analysis. Bioinformatics 33, 272–279 (2017).

  93. 93.

    et al. Concepts, estimation and interpretation of SNP-based heritability. Nat. Genet. 49, 1304–1310 (2017).

  94. 94.

    et al. Psychiatric genomics: an update and an agenda. Am. J. Psychol. (2017).

  95. 95.

    Sharing extended summary data from contemporary genetic studies is unlikely to threaten subject privacy. PLOS ONE 12, e0179504 (2017).

  96. 96.

    et al. Multivariate genetic analyses of cognition and academic achievement from two population samples of 174,000 and 166,000 school children. Behav. Genet. 42, 699–710 (2012).

  97. 97.

    et al. Molecular genetic contributions to socioeconomic status and intelligence. Intelligence 44, 26–32 (2014).

  98. 98.

    , & Variation in the heritability of educational attainment: An international meta-analysis. Soc. Forces 92, 109–140 (2013).

  99. 99.

    et al. The high heritability of educational achievement reflects many genetically influenced traits, not just intelligence. Proc. Natl Acad. Sci. USA 111, 15273–15278 (2014).

  100. 100.

    , & Twins Early Development Study (TEDS): a genetically sensitive investigation of cognitive and behavioral development from childhood to young adulthood. Twin Res. Hum. Genet. 16, 117–125 (2013).

Download references

Acknowledgements

The authors gratefully acknowledge the ongoing contribution of the participants in the Twins Early Development Study (TEDS) and their families. TEDS is supported by a programme grant to R.P. from the UK Medical Research Council (MR/M021475/1 and previously G0901245), with additional support from the US National Institutes of Health (AG046938). The research reported here has also received funding from the European Research Council (ERC) under the European Union's Seventh Framework Programme (FP7/2007-2013) grant agreement 602768 and ERC grant agreement 295366. R.P. is also supported by a Medical Research Council Professorship award (G19/2). S.v.S. is supported by a Jacobs Foundation Research Fellowship award (2017–2019).

Author information

Affiliations

  1. Institute of Psychiatry, Psychology and Neuroscience, King's College London, 16 De Crespigny Park, London SE5 8AF, UK.

    • Robert Plomin

  2. Department of Psychological and Behavioural Science, London School of Economics and Political Science, Queens House, 55–56 Lincoln's Inn Fields, London WC2A 3LJ, UK.

    • Sophie von Stumm

Contributions

The authors contributed equally to all aspects of the manuscript.

Competing interests

The authors declare no competing financial interests.

Corresponding authors

Correspondence to Robert Plomin or Sophie von Stumm.

Glossary

Twin studies

Studies comparing the resemblance of identical and fraternal twins to estimate genetic and environmental components of variance.

Variance

An index of how spread out scores are in a study population, which is calculated as the average of the squared deviations from the mean.

Genome-wide association studies

(GWAS). Studies that aim to identify loci throughout the genome associated with an observed trait or disorder.

Heritability

The proportion of observed differences among individuals that can be attributed to inherited differences in genome sequence.

Genome-wide polygenic scores

(GPSs). Genetic indices of a trait for each individual that are the sum across the genome of thousands of single-nucleotide polymorphisms (SNPs) of the individual's increasing alleles associated with the trait, usually weighted by the effect size of each SNP's association with the trait in genome-wide association studies.

Candidate gene studies

Studies that focus on genes for which the function suggests that they are associated with a trait, in contrast to genome-wide association studies.

Effect sizes

Proportions of variance of traits in the study population accounted for by a particular factor such as a genome-wide polygenic score.

Single-nucleotide polymorphisms

(SNPs). Single base pair differences in inherited DNA sequence between individuals.

Linkage disequilibrium (LD) score regression analysis

Analysis that, for each single-nucleotide polymorphism in a genome-wide association study (GWAS), regresses χ2 statistics from GWAS summary statistics against LD scores.

Rights and permissions

To obtain permission to re-use content from this article visit RightsLink.