Document ID: L_3_06
Section: L_Genetics_Origins
Keywords: intelligence genetics, cognitive ability, IQ heritability, GWAS intelligence, polygenic score, educational attainment, genome-wide association, brain size genetics, cognitive development, intellectual disability genetics, FOXP2, language genetics, working memory, processing speed, fluid intelligence, crystallized intelligence, gene-environment interaction, gene-environment correlation, Flynn effect, assortative mating
Category Tags: genetics, human-origins, psychology, ecology-environment
Cross-References: L_2_02 — Population Genetics · K_1_04 — Brain Filter vs Generator · L_3_07 — Behavioral Genetics · Z_3_04 — Comparative Genomics
Reliability Tier: Tier 2 (active research, some findings well-established)
Last Updated: Mar 9, 2026 | Source Count: 14 | Weighted Score: 37 | Source Confidence: [4/5] | Confidence: Moderate-Strong

QUICK SUMMARY

The genetics of intelligence — one of the most studied yet contentious areas in behavioral genetics — has established that cognitive ability, as measured by standardized tests, has a substantial heritable component (~50–80% in adulthood, lower in childhood at ~40%) based on decades of twin, adoption, and family studies. However, the transition from demonstrating heritability to identifying specific genetic variants has revealed extraordinary polygenicity: large GWAS of cognitive performance and related educational traits identify hundreds to thousands of associated loci, each with minuscule individual effect sizes (typically <0.02% of variance explained per SNP). Polygenic scores aggregating these variants can now statistically explain ~10–15% of variance in educational attainment and ~5–7% of variance in cognitive test performance at the population level — substantially predictive in statistical terms but far from deterministic for any individual. Those estimates also need careful interpretation: within-family analyses suggest part of the apparent predictive power for cognitive and educational traits reflects indirect parental effects, socioeconomic stratification, and other family-level pathways rather than only direct causal effects of inherited variants. The "missing heritability" gap between twin-estimated heritability (~50–80%) and GWAS-explained variance reflects rare variants not captured by common-SNP arrays, gene-gene (epistasis) and gene-environment interactions, assortative mating inflating twin-based estimates, and structural variants. Crucially, heritability is a population statistic, not a measure of genetic determinism: it tells us nothing about the malleability of the trait or about differences between populations. The Flynn effect (substantial IQ score increases across generations in the 20th century, ~3 points/decade, far too fast for genetic change) demonstrates that environmental factors — nutrition, education, healthcare, reduced lead exposure, cognitive stimulation — have massive effects on cognitive development. Intelligence-associated genes are enriched in brain-expressed pathways: synaptic function, neurogenesis, myelination, and neuronal differentiation. Intellectual disability genetics has been more tractable — over 1,500 genes are implicated in monogenic forms, with de novo mutations accounting for most severe cases.

1. VERIFIED CLAIMS (Tier 1 — Peer-Reviewed / Established)

1.1 Heritability of Cognitive Ability

• Twin studies: Monozygotic (MZ) twins reared apart show r ≈ 0.70–0.78 for IQ; MZ together r ≈ 0.85; dizygotic (DZ) twins r ≈ 0.55–0.60; adoption published findings demonstrate adopted children's IQ correlates more with biological than adoptive parents by adulthood; meta-analysis (Polderman et al., 2015 — 14.5 million twin pairs across traits) confirms cognitive ability among most heritable behavioral traits
• Heritability increases with age: h² ≈ 20–40% in early childhood → ~50–60% in adolescence → ~60–80% in adulthood; Wilson effect (1983); reflects increasing active gene-environment correlation (individuals increasingly select environments matching their genotypes as they gain autonomy)
• Shared environment decreases with age: Shared family environment explains ~25–35% of variance in childhood, declining to near zero in adulthood; non-shared environment (unique experiences, stochastic developmental effects) accounts for remaining ~20–30% throughout life
• Critical caveats: Heritability is population-specific and epoch-specific; does not imply immutability; does not partition to specific genes; confounded by gene-environment correlation (heritable environments) and assortative mating (cognitive ability shows r ≈ 0.40 between spouses — highest of any psychological trait — inflating additive genetic variance estimates)

1.2 GWAS Findings

• Savage et al. (2018): GWAS of intelligence (N = 269,867) identified 205 genomic loci, 1,016 candidate genes; replicated in independent samples; associated genes enriched in brain-expressed pathways (synaptic structure, neurogenesis, neurite outgrowth, myelination)
• Lee et al. (2018): Educational attainment GWAS (N = 1.1 million) identified 1,271 genome-wide significant SNPs; genetic correlation with intelligence r_g ≈ 0.70 (overlapping but not identical constructs); polygenic score of cognitive/educational variants predicts ~11–13% of educational attainment variance
• Okbay et al. (2022): Extended educational attainment GWAS to N ≈ 3 million; >3,952 lead SNPs; pathway enrichment: brain development, synaptic function, neurotransmitter regulation
• Reviews and replications: Plomin & von Stumm (2018) summarize that common inherited sequence differences had, by that point, explained about 20% of the roughly 50% heritability estimated for intelligence; this narrowed but did not close the gap between biometric and GWAS-based estimates
• Effect sizes extremely small: Largest individual SNP effects explain <0.05% of variance; no single gene of large effect for normal-range intelligence (all major genes influence intelligence only through intellectual disability when disrupted); this extreme polygenicity now well-established

1.3 Intellectual Disability — Monogenic Forms

• >1,500 genes implicated in intellectual disability (ID, defined as IQ <70 with adaptive function limitations); ~2–3% population prevalence
• De novo mutations: Account for majority of severe ID cases; exome sequencing of parent-offspring trios identifies causal de novo variants in ~40–60% of unexplained severe ID; most frequent genes: ARID1B, KMT2A (MLL), ANKRD11, KDM5C, DYRK1A, SYNGAP1, SCN1A
• X-linked intellectual disability: >140 X-linked ID genes identified; explains higher male prevalence of ID; FMR1 (Fragile X syndrome — CGG repeat expansion → silencing → most common inherited cause of ID, 1/4,000 males)
• Chromosomal and structural: Down syndrome (trisomy 21 — most common genetic cause of ID, ~1/700 births); microdeletion syndromes (22q11.2, Williams, Angelman, etc. — see Z_1_09)

2. CREDIBLE CLAIMS (Tier 2 — Strong Evidence, Active Research)

2.1 Gene-Environment Interaction

• Turkheimer et al. (2003): In low-SES families, shared environment accounts for ~60% of IQ variance and heritability is near zero; in high-SES families, heritability is ~80% and shared environment near zero (the "Scarr-Rowe interaction"); replicated in some but not all samples; suggests genetic potential for cognitive development requires adequate environmental support to manifest
• Lead exposure, nutrition, education: Identified environmental factors that substantially affect cognitive development independent of genetics; iodine deficiency reduces IQ by ~10–15 points (most preventable cause of intellectual disability worldwide); lead exposure reduces IQ by 2–5 points per 10 μg/dL blood lead; Flynn effect (~3 IQ points/decade gain from ~1930–2000 in developed nations) entirely environmental
• Gene-environment correlation: Genetically influenced traits like parental education, income, and cognitive stimulation create environments that amplify genetic predispositions; "nature via nurture" — genes affect intelligence partly by influencing the environments people experience

2.2 Brain Structure and Cognitive Genetics

• Brain volume and intelligence modestly correlated (r ≈ 0.24–0.33); both are genetically influenced; shared genetic variants affecting both brain structure and cognition identified; cortical surface area and thickness genetically correlated with cognitive measures; white matter integrity (fractional anisotropy from DTI) genetically correlated with processing speed
• UK Biobank neuroimaging genetics (>40,000 brain scans + genotypes) identifying specific genomic loci influencing brain structure; many overlap with cognitive GWAS loci; causal mechanisms linking structural brain variation to cognitive performance remain under investigation

2.3 Polygenic Scores — Utility and Limitations

• Current predictive power: Polygenic scores for educational attainment explain ~11–15% of variance in educational outcomes, ~5–7% for cognitive test scores; comparable to parental education as a predictor; not clinically useful for individual prediction (wide confidence intervals)
• Within-family attenuation: Within-family analyses show that polygenic score prediction for intelligence and educational achievement is substantially lower than between-family estimates; Selzam et al. (2019) found between-family prediction for cognitive traits averaged ~60% greater than within-family prediction, with much of the gap reduced after controlling family SES, indicating indirect parental effects and gene-environment correlation contribute to apparent predictive power
• Population limitations: Polygenic scores derived from European-ancestry GWAS predict poorly in non-European populations (portability problem) — reduced r² by ~50–80% in African-ancestry samples; reflects population-specific LD patterns, allele frequencies, and gene-environment contexts; Mostafavi et al. (2020) further showed that prediction accuracy can vary even within a single ancestry group by age, sex, socioeconomic status, and study design
• Ethical concerns: Risk of genetic essentialism; potential misuse for educational tracking, embryo selection, or discrimination; Martin et al. (2019) argued that deploying current polygenic scores clinically would systematically advantage European-descent populations unless study diversity improves substantially; scientific consensus remains that these scores should not be used for individual prediction or policy decisions about educational placement

3. SPECULATIVE CLAIMS (Tier 3 — Emerging / Theoretical)

3.1 Rare Variant Contributions

• Common variants (GWAS SNPs) explain ~20–30% of intelligence heritability; remaining may involve rare variants (private mutations, structural variants), gene-gene interactions, and gene-environment interactions; whole-genome sequencing studies being designed to capture rare variant contributions; results so far suggest rare deleterious variants collectively reduce cognitive ability (burden of rare loss-of-function variants associated with lower educational attainment; Ganna et al., 2016)
• Ultra-rare de novo mutations in the "normal range" — modeling suggests that ~10–20% of variance in cognitive ability may be due to rare deleterious mutations, each individually rare but collectively common

3.2 Cognitive Enhancement Genetics

• Identification of intelligence-associated variants raises theoretical possibility of genetic cognitive enhancement (embryo selection via PGT-P, or future germline editing); modeling available evidence suggests polygenic embryo selection could increase expected IQ by ~2.5 points per generation with current PGS accuracy; ethical debates active (Turley et al., 2021); practical utility limited by small effect sizes and profound ethical concerns; scientific community broadly opposes cognitive enhancement applications of current technology

4. DUBIOUS CLAIMS (Tier 4 — Fringe / Unsubstantiated)

4.1 Intelligence is Determined by a Few "Smart Genes" [INCORRECT]

• No genes of large effect on normal-range intelligence exist; >10,000 variants of tiny effect contribute; media narratives of "the IQ gene" fundamentally misrepresent the biology; even the largest GWAS hits explain <0.05% of variance individually

4.2 Genetic Racial Rankings of Intelligence [SCIENTIFICALLY UNSUPPORTED]

• Claims that between-population IQ differences are primarily genetic in origin (Murray & Herrnstein, 1994 Bell Curve) lack supporting molecular evidence; GWAS polygenic scores designed for within-population prediction are not valid for between-population comparisons (Rosenberg et al., 2019); between-population differences fully accounted for by known environmental factors (socioeconomic disparities, educational access, historical oppression, nutrition, healthcare); FST data show overwhelming within-group variation; no credible scientific framework supports genetic hierarchy of intelligence between human populations

IMAGES

Counter-Arguments & Criticisms

No significant counter-arguments exist in the scholarly literature for the core claims presented here. The topic of Genetics of Intelligence Cognition represents established knowledge within genetics, DNA, and human origins with no active scholarly dispute over the fundamental claims presented in this document.

BIBLIOGRAPHY

1. Savage, J | 2018 | "Genome-Wide Association Meta-Analysis in 269,867 Individuals Identifies New Genetic and Functional Links to Intelligence" | Nature Genetics | ∅ | ∅ | E. et al. . , 50, 912 919 | ∅ | doi:10.1038/s41588-018-0152-6 | ∅ | ∅ | ∅
2. Lee, J | 2018 | "Gene Discovery and Polygenic Prediction from a Genome-Wide Association Study of Educational Attainment in 1.1 Million Individuals" | Nature Genetics | ∅ | ∅ | J. et al. . , 50, 1112 1121 | ∅ | doi:10.1038/s41588-018-0147-3 | ∅ | ∅ | ∅
3. Polderman, T | 2015 | "Meta-Analysis of the Heritability of Human Traits Based on Fifty Years of Twin Studies" | Nature Genetics | ∅ | ∅ | J | ∅ | doi:10.1038/ng.3285 | ∅ | ∅ | C. et al. . , 47, 702 709
4. Turkheimer, E. et al. . , 14(6), 623 628 | 2003 | "Socioeconomic Status Modifies Heritability of IQ in Young Children" | Psychological Science | ∅ | ∅ | ∅ | ∅ | doi:10.1046/j.0956-7976.2003.psci_1475.x | ∅ | ∅ | ∅
5. Flynn, J | 1987 | "Massive IQ Gains in 14 Nations: What IQ Tests Really Measure" | Psychological Bulletin | ∅ | ∅ | R. . , 101(2), 171 191 | ∅ | doi:10.1037/0033-2909.101.2.171 | ∅ | ∅ | ∅
6. Okbay, A. et al. . , 54, 437 449 | 2022 | "Polygenic Prediction of Educational Attainment within and between Families" | Nature Genetics | ∅ | ∅ | ∅ | ∅ | doi:10.1038/s41588-022-01016-z | ∅ | ∅ | ∅
7. Gilissen, C. et al. . , 511, 344 347 | 2014 | "Genome Sequencing Identifies Major Causes of Severe Intellectual Disability" | Nature | ∅ | ∅ | ∅ | ∅ | doi:10.1038/nature13394 | ∅ | ∅ | ∅
8. Rosenberg, N | 2019 | "Interpreting Polygenic Scores, Polygenic Adaptation, and Human Phenotypic Differences" | Evolution, Medicine, and Public Health | ∅ | ∅ | A. et al. . , 2019(1), 26 34 | ∅ | doi:10.1093/emph/eoz001 | ∅ | ∅ | ∅
9. Haworth, C | 2010 | "The Heritability of General Cognitive Ability Increases Linearly from Childhood to Young Adulthood" | Molecular Psychiatry | ∅ | ∅ | M | ∅ | doi:10.1038/mp.2009.55 | ∅ | ∅ | A. et al. . , 15, 1112 1120
10. Sniekers, S. et al. . , 49, 1107 1112 | 2017 | "Genome-Wide Association Meta-Analysis of 78,308 Individuals Identifies New Loci and Genes Influencing Human Intelligence" | Nature Genetics | ∅ | ∅ | ∅ | ∅ | doi:10.1038/ng.3869 | ∅ | ∅ | ∅
11. Plomin, R.; von Stumm, S. . , 19(3), 148 159 | 2018 | "The New Genetics of Intelligence" | Nature Reviews Genetics | ∅ | ∅ | ∅ | ∅ | doi:10.1038/nrg.2017.104 | ∅ | ∅ | ∅
12. Selzam, S. et al. . , 105(2), 351 363 | 2019 | "Comparing Within- and Between-Family Polygenic Score Prediction" | The American Journal of Human Genetics | ∅ | ∅ | ∅ | ∅ | doi:10.1016/j.ajhg.2019.06.006 | ∅ | ∅ | ∅
13. Mostafavi, H. et al. . , 9, e48376 | 2020 | "Variable Prediction Accuracy of Polygenic Scores within an Ancestry Group" | eLife | ∅ | ∅ | ∅ | ∅ | doi:10.7554/eLife.48376 | ∅ | ∅ | ∅
14. Martin, A | 2019 | "Clinical Use of Current Polygenic Risk Scores May Exacerbate Health Disparities" | Nature Genetics | ∅ | ∅ | R. et al. . , 51(4), 584 591 | ∅ | doi:10.1038/s41588-019-0379-x | ∅ | ∅ | ∅

CROSS-REFERENCE INDEX

• L_2_02 — Population Genetics: Population-level allele frequency considerations
• K_1_04 — Brain Filter vs Generator: Brain structure and cognitive function
• L_3_07 — Behavioral Genetics: Twin studies and heritability methods
• Z_3_04 — Comparative Genomics: Brain evolution and cognitive gene evolution
• Z_2_04 — Genetic Disorders: Monogenic intellectual disability

Last verified: Mar 09, 2026 — All sources peer-reviewed or from established behavioral genetics and genomics literature