CoChat
Sign Up
Cited by 659Genetic Associations and EpidemiologyGenetic and phenotypic traits in livestockGenetic Mapping and Diversity in Plants and Animals

Polygenic prediction of educational attainment within and between families from genome-wide association analyses in 3 million individuals

Open access
Okbay, Aysu, Wu, Yeda, Wang, Nancy, Jayashankar, Hariharan, Bennett, Michael, Nehzati, Seyed Moeen, Sidorenko, Julia, Kweon, Hyeokmoon +294 more·2022·Nature Genetics
DOI 10.1038/s41588-022-01016-zPMC PMC9005349
Open this paper in CoChat — free accountOpen PDF

Abstract

We conduct a genome-wide association study (GWAS) of educational attainment (EA) in a sample of ~3 million individuals and identify 3,952 approximately uncorrelated genome-wide-significant single-nucleotide polymorphisms (SNPs). A genome-wide polygenic predictor, or polygenic index (PGI), explains 12–16% of EA variance and contributes to risk prediction for ten diseases. Direct effects (i.e., controlling for parental PGIs) explain roughly half the PGI’s magnitude of association with EA and other phenotypes. The correlation between mate-pair PGIs is far too large to be consistent with phenotypic assortment alone, implying additional assortment on PGI-associated factors. In an additional GWAS of dominance deviations from the additive model, we identify no genome-wide-significant SNPs, and a separate X-chromosome additive GWAS identifies 57.

AI snapshot

Written from the paper’s full text.

TL;DR

This GWAS meta-analysis of educational attainment in ~3.0 million individuals identified 3,952 approximately uncorrelated autosomal SNPs and a polygenic index explaining 12–16% of EA variance. It shows that the EA PGI’s predictive power partly reflects direct effects, predicts ten diseases, and has mate-pair PGI correlations too large to be explained by phenotypic assortment alone.

Key findings

  • The additive EduYears GWAS (N = 3,037,499) identified 3,952 lead SNPs (pairwise r² < 0.1, P < 5×10⁻⁸); the median SNP effect was 1.4 weeks more schooling per reference allele. The LDpred PGI explained 15.8% of EduYears variance in Add Health and 12.0% in HRS (sample-size-weighted mean 13.3%); an SBayesR PGI explained 17.0% and 12.9%, respectively (mean 14.3%).
  • The X-chromosome GWAS (N = 2,713,033) identified 57 lead SNPs, with X-chromosome SNP heritability of 0.4% and male-female genetic correlation rg = 0.94 (S.E. = 0.03). The dominance GWAS (N = 2,574,253) identified no genome-wide-significant SNPs; dominance SNP heritability was 0.00015 (S.E. = 0.00024, P = 0.54), and the inbreeding-depression estimate implied offspring of first cousins have ~1.0 fewer month of EA (P = 0.04).
  • The EA PGI significantly predicted all ten diseases examined; mean incremental R² was 0.63%, compared with 1.19% for disease-specific PGIs. For EA, the ratio of direct to population effect was 0.556 (S.E. = 0.020), implying 30.9% of the PGI’s R² was due to direct effects; corresponding ratios were 0.910 for height, 0.962 for BMI, and 0.824 for cognitive performance.
  • Mate-pair EA PGI correlation was 0.175 (S.E. = 0.020), versus 0.031 (S.E. = 0.004) predicted by phenotypic assortment; for height, observed 0.106 (S.E. = 0.020) versus predicted 0.087 (S.E. = 0.007). Residualizing EA PGI on EA reduced the mate-pair correlation by only 37%, to 0.110 (S.E. = 0.021); further adjustment for 40 PCs reduced it to 0.091 (S.E. = 0.021).

Methods

  • Meta-analysis of EduYears combined prior public results from 69 cohorts (N = 324,162, excluding UKB and 23andMe), new 23andMe results (N = 2,272,216), and new UKB results (N = 441,121), totaling N = 3,037,499, all in European genetic ancestries, with controls for sex, year of birth, their interaction, and genetic principal components; METAL sample-size-weighted meta-analysis and LD score intercept 1.663.
  • Separate X-chromosome meta-analysis (N = 2,713,033) and dominance-deviation GWAS meta-analysis (N = 2,574,253) were conducted; dominance analyses used adapted LD score regression and ldscdom for inbreeding depression.
  • Polygenic prediction used Add Health, HRS, and WLS holdout samples, with LDpred HapMap3 PGIs as main and SBayesR post hoc; disease prediction used UKB and Nagelkerke incremental R². Within-family analyses used ~53,000 individuals with genotyped siblings and ~3,500 with both parents genotyped from UKB, Generation Scotland, and the Swedish Twin Registry; assortative-mating analyses used 862 UKB and 1,603 Generation Scotland mate pairs.

Limitations

  • Stated: all main GWAS analyses were restricted to European genetic ancestries, and the PGI had much lower predictive power in African-genetic-ancestry samples (incremental R² 1.3% in HRS and 2.3% in Add Health; relative accuracies 11% and 15%). The paper also states the PGI cannot be used to meaningfully predict an individual’s EA, that dominance results do not rule out rare SNPs with substantial dominance effects, and that family samples lacked power for the ten diseases.
  • Evident: the dominance GWAS addresses common SNPs and found no genome-wide-significant dominance effects; family/direct-effect analyses are limited to ~53,000 siblings, ~3,500 parent-genotyped individuals, and 23 phenotypes.
  • Evident: assortative-mating analyses rely on UKB and Generation Scotland mate pairs and leave a substantial mate-pair PGI correlation unexplained after adjustment for EA, PCs, cognitive performance, and vocabulary.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1038/s41588-022-01016-z

Cites

71 of 73 references verified0 citation errors · 2 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

21 of 24 cited sources checked · 12 supported · 1 flagged · accuracy not yet measured

  • Findingown data

    The EduYears PGI shows substantial loss of prediction accuracy in samples of African genetic ancestries.

    “We used a method that was recently developed by Wang et al.25 to investigate the factors contributing to the substantial loss of prediction accuracy of the EduYears PGI in samples of African genetic ancestries.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Wang et al. [25] Wang, Y. et al. Theoretical and empirical quantification of the accuracy of polygenic scores in ancestry divergent populations. Nat. Commun. 11, 3865 (2020). source
  • Premisecitation

    The expected coefficient of determination R2 can be expressed as E(R2)=A/(B+1/N).

    “The expected coefficient of determination, R2, can be expressed as the following function of the discovery sample size, N: E(R2)=AB+1∕N.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites de Vlaming et al. [71] de Vlaming, R. et al. Meta-GWAS accuracy and power (MetaGAP) calculator shows that hiding heritability is partially due to imperfect genetic correlations across studies. PLoS Genet. 13, e1006495 (2017). source
  • Findingown data

    At a stricter threshold of P < 1×10−8, the number of lead SNPs declines to 3,277.

    “At the stricter threshold9 of P < 1×10−8, the number declines to 3,277”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [9] Wu, Y., Zheng, Z., Visscher, P. M. & Yang, J. Quantifying the mapping precision of genome-wide association studies using whole-genome sequencing data. Genome Biol. 18, 86 (2017). source
  • Findingown data

    The EA PGI has by far the lowest ratio of direct to population effects among EA, height, BMI, and cognitive performance.

    “The EA PGI has by far the lowest ratio.”
    Quote in the paperSubstantial claim
  • Findingown data

    For height, BMI, and cognitive performance, controlling for parental PGIs has far less impact on their associations with their corresponding phenotype than for EA.

    “In contrast, when we examine PGIs for height, BMI, and cognitive performance, controlling for parental PGIs has far less impact on their associations with their corresponding phenotype.”
    Quote in the paperSubstantial claim
  • May combine two claims. 32
  • Classified as probably not a substantial claim. 19
  • The quote only partly states this claim. 1

  • May combine two claims.

    Findingown data

    The LD score regression intercept for the EA GWAS is 1.66, and confounding accounts for 7% of the inflation.

    “According to the LD score regression8 intercept (1.66), confounding accounts for 7% of the inflation, similar to previous GWAS of EA2”
    Quote in the paperNot a substantial claim2 of 2 sources checked
    • cites [8] Bulik-Sullivan, B. K. et al. LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat. Genet. 47, 291–295 (2015). source
    • cites [2] Lee, J. J. et al. Gene discovery and polygenic prediction from a genome-wide association study of educational attainment in 1.1 million individuals. Nat. Genet. 50, 1112–1121 (2018). source

      In Results > Primary GWAS of EduYears (page not provided), the paper reports an LD Score regression intercept of 1.11 and estimates that bias accounts for about 5% of the inflation. Those figures contradict the attributed intercept of 1.66 and 7% estimate. (LLM, conf=0.91)

      Checked by a reasoning model — may be wrong.

      Experimental — accuracy not yet measured

  • May combine two claims.

    Interpretationcitation

    The remaining reduction in predictive power is due to factors including epistasis, gene-environment interactions, and differences between populations in gene-environment correlations, assortative mating, and environmental variance.

    “The remaining reduction in predictive power is due to factors including epistasis (although epistatic variance is likely small13,15), gene-environment interactions, and differences between populations in gene-environment correlations, assortative mating, and environmental variance.”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [13] Hivert, V. et al. Estimation of non-additive genetic variance in human complex traits from a large sample of unrelated individuals. Am. J. Hum. Genet. 108, 786–798 (2021). source
    • cites [15] Hill, W. G., Goddard, M. E. & Visscher, P. M. Data and theory point to mainly additive genetic variance for complex traits. PLoS Genet. 4, e1000008 (2008). source

  • May combine two claims.

    Premisecitation

    PGIs constructed from GWAS in European genetic ancestries generally have much lower predictive power in samples with other genetic ancestries; on average across phenotypes, relative accuracy in African-ancestry to European-ancestry samples has been 22% and 36%.

    “PGIs like ours that are constructed from GWAS in samples of European genetic ancestries are generally found to have much lower predictive power in samples with other genetic ancestries; for example, on average across phenotypes, estimates of relative accuracy (ratio of R2) in African-genetic-ancestry to European-genetic-ancestry samples have been 22%23 and 36%24.”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [23] Martin, A. R. et al. Clinical use of current polygenic risk scores may exacerbate health disparities. Nat. Genet. 51, 584–591 (2019). source

      Split into 2 atoms — 1 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites [24] Duncan, L. et al. Analysis of polygenic risk score usage and performance in diverse human populations. Nat. Commun. 10, 3328 (2019). source

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingown data

    Using UKB, the relative accuracy is smaller than would be predicted based on population differences in allele frequencies and LD alone, and this discrepancy is greater for EA than has been found in prior work for height, BMI, and six other phenotypes.

    “Using the UKB, we find that the relative accuracy is smaller than would be predicted based on population differences in allele frequencies and LD alone (Online Methods), and this discrepancy is greater for EA than has been found in prior work25 for height, BMI, and six other phenotypes”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [25] Wang, Y. et al. Theoretical and empirical quantification of the accuracy of polygenic scores in ancestry divergent populations. Nat. Commun. 11, 3865 (2020). source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    For UKB analyses, covariates included the top 40 genetic PCs and genotyping array dummies.

    “In addition, for UKB, we included as covariates the top 40 genetic PCs provided by UKB and the genotyping array dummies3.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [3] Bycroft, C. et al. The UK Biobank resource with deep phenotyping and genomic data. Nature 562, 203–209 (2018). source

  • May combine two claims.

    Interpretationown data

    The unexplained remainder of the mate-pair PGI correlation is due to assortment on phenotypes correlated with the EA PGI other than EA, cognitive performance, and vocabulary-possibly including personality traits-and sources of social homogamy other than genetic ancestry captured by the top 40 PCs-possibly including geographic location at courtship age, socioeconomic status, and social class.

    “This remainder is due to assortment on phenotypes correlated with the EA PGI other than EA, cognitive performance, and vocabulary—possibly including various personality traits42-44—and sources of social homogamy other than genetic ancestry captured by the top 40 PCs—possibly including geographic location at courtship age45,46, socioeconomic status, and social class47.”
    Quote in the paperSubstantial claim6 of 6 sources checked
    • cites [42] Belsky, D. W. et al. The genetics of success: how single-nucleotide polymorphisms associated with educational attainment relate to life-course development. Psychol. Sci. 27, 957–972 (2016). source

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites [43] Mõttus, R., Realo, A., Vainik, U., Allik, J. & Esko, T. Educational attainment and personality are genetically intertwined. Psychol. Sci. 28, 1631–1639 (2017). source

      Split into 2 atoms — 1 confirmed, 0 partially supported, 0 disputed, 0 misattributed, 1 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites [44] Smith-Woolley, E., Selzam, S. & Plomin, R. Polygenic score for educational attainment captures DNA variants shared between personality traits and educational achievement. J. Pers. Soc. Psychol. 117, 1145–1163 (2019). source

      Split into 2 atoms — 1 confirmed, 0 partially supported, 0 disputed, 0 misattributed, 1 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites [45] Laidley, T., Vinneau, J. & Boardman, J. D. Individual and social genomic contributions to educational and neighborhood attainments: geography, selection, and stratification in the United States. Sociol. Sci. 6, 580–608 (2019). source

      Split into 2 atoms — 0 confirmed, 2 partially supported, 0 disputed, 0 misattributed, 0 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites [46] Abdellaoui, A. et al. Genetic correlates of social stratification in Great Britain. Nat. Hum. Behav. 3, 1332–1342 (2019). source

      Split into 2 atoms — 0 confirmed, 1 partially supported, 0 disputed, 0 misattributed, 1 unverifiable. Expand the claim to see per-atom verdicts.

      Experimental — accuracy not yet measured
    • cites [47] Belsky, D. W. et al. Genetic analysis of social-class mobility in five longitudinal studies. Proc. Natl Acad. Sci. USA 115, E7275–E7284 (2018). source
    • Rests on an assumption: Cognitive performance and vocabulary measures in Generation Scotland are proxies for the composite phenotype on which mate pairs assort.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductcitation

    The LD score regression for dominance was restricted to HapMap3 SNPs, and dominance LD scores were estimated using the 1000 Genomes Phase 1 reference sample.

    “The LD score regression was restricted to the set of HapMap3 SNPs, and the dominance LD scores were estimated using the 1000 Genomes Phase 1 reference sample67.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [67] The 1000 Genomes Project Consortium et al. An integrated map of genetic variation from 1,092 human genomes. Nature 491, 56–65 (2012). source

  • May combine two claims.

    Findingown data

    A PGI generated from ~2.5 million pruned common SNPs using SBayesR is more predictive than the LDpred PGI, explaining 17.0% of EduYears variance in Add Health and 12.9% in HRS, with a sample-size-weighted mean of 14.3%.

    “In post hoc analyses, we found that a PGI generated from ~2.5 million pruned common SNPs using the software SBayesR22 is more predictive than our LDpred PGI. It explains 17.0% of the variance in EduYears in Add Health and 12.9% in HRS, with a sample-size-weighted mean of 14.3% (Supplementary Table 3).”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [22] Lloyd-Jones, L. R. et al. Improved polygenic prediction by Bayesian multiple regression on summary statistics. Nat. Commun. 10, 5086 (2019). source

  • May combine two claims.

    Findingown data

    The EA PGI predicts a range of cognitive phenotypes and adds non-trivial predictive power for 10 diseases even after controlling for disease-specific PGIs.

    “We first document that the EA PGI not only predicts a range of cognitive phenotypes as has been found in previous work2,4, but it also adds non-trivial predictive power for 10 diseases we examine, even after controlling for disease-specific PGIs.”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [2] Lee, J. J. et al. Gene discovery and polygenic prediction from a genome-wide association study of educational attainment in 1.1 million individuals. Nat. Genet. 50, 1112–1121 (2018). source
    • cites [4] Harden, K. P. et al. Genetic associations with mathematics tracking and persistence in secondary school. NPJ Sci. Learn. 5, 1 (2020). source

      The Results sections on mathematics tracking and persistence report that the education polygenic score predicts advanced math placement and continued math coursetaking. These support prediction of specific educational outcomes, but not the broader claim about a range of cognitive phenotypes. The sentence presents the finding about 10 diseases as the citing authors’ own result, not as a finding attributed to this source (page numbers unavailable). (LLM, conf=0.82)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingown data

    A genome-wide polygenic index (PGI) explains 12-16% of EA variance, depending on the validation sample; specifically, the LDpred PGI explains 15.8% in Add Health and 12.0% in HRS, with a sample-size-weighted mean of 13.3%.

    “A genome-wide polygenic predictor, or polygenic index (PGI), explains 12-16% of EA variance and contributes to risk prediction for ten diseases.”
    • Also stated in: § Polygenic Prediction · plain
      “This PGI explains 15.8% of the variance in EduYears in Add Health and 12.0% in HRS (Extended Data Figure 6). The sample-size-weighted mean is 13.3%.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    For height, the mate-pair phenotypic correlation is 0.290 (S.E. = 0.018) and the mate-pair PGI correlation is 0.106 (S.E. = 0.020); for EA, the mate-pair phenotypic correlation is 0.430 (S.E. = 0.017) and the mate-pair PGI correlation is 0.175 (S.E. = 0.020).

    “Height provides a useful comparison because its mate-pair phenotypic correlation (0.290, S.E. = 0.018) and mate-pair PGI correlation (0.106, S.E. = 0.020) are somewhat similar to EA’s mate-pair phenotypic correlation (0.430, S.E. = 0.017) and mate-pair PGI correlation (0.175, S.E. = 0.020).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In HRS African-ancestry samples (N = 2,507) and Add Health African-ancestry samples (N = 1,716), the EA PGI's incremental R2 for EduYears is 1.3% (95% CI 0.6% to 2.2%) and 2.3% (95% CI 1.1% to 3.7%), implying relative accuracies of 11% and 15%.

    “When we used our PGI to predict EduYears in samples with African genetic ancestries from the HRS (N = 2,507) and Add Health (N = 1,716), the incremental R2 was 1.3% (95% CI: 0.6% to 2.2%) and 2.3% (95% CI: 1.1% to 3.7%), implying that the relative accuracies for EA in the HRS and Add Health are only 11% and 15%, respectively.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In Generation Scotland, after residualizing on cognitive performance and vocabulary proxies as well as EA and 40 PCs, the mate-pair PGI correlation is 0.083 (S.E. = 0.027), compared to 0.113 (S.E. = 0.026) when residualizing on EA and PCs alone, leaving a substantial remainder unexplained.

    “In this cohort, after residualizing on these proxies as well as on EA and 40 PCs, the mate-pair PGI correlation is 0.083 (S.E. = 0.027) compared to 0.113 (S.E. = 0.026) when residualizing on EA and PCs alone—which leaves a substantial remainder of the mate-pair PGI correlation unexplained.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Cognitive performance and vocabulary measures in Generation Scotland are proxies for the composite phenotype on which mate pairs assort.

  • May combine two claims.

    Findingown data

    A PGI constructed using only genome-wide-significant SNPs has an incremental R2 of 9.1% in Add Health and 7.0% in HRS.

    “A PGI constructed using only genome-wide-significant SNPs has an incremental R2 of 9.1% in Add Health and 7.0% in HRS (Extended Data Figure 5).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The median lead SNP's effect is 1.4 weeks more schooling per additional copy of the reference allele; effects at the 5th and 95th percentiles are 0.9 and 3.5 weeks.

    “On average, an additional copy of the reference allele of the median SNP is associated with 1.4 weeks more schooling: the effects at the 5th and 95th percentiles (in absolute value) are 0.9 and 3.5 weeks, respectively”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingreasoning

    Biological annotation results are very similar across the current and previous meta-analyses, but standard errors are smaller in the current meta-analysis.

    “The results are very similar across the two meta-analyses, but standard errors are smaller when using the current meta-analysis results, as expected given the larger sample size”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For height, BMI, and cognitive performance, the ratios of direct to population effect estimates are 0.910 (S.E. = 0.009), 0.962 (S.E. = 0.017), and 0.824 (S.E. = 0.033), implying direct-effect shares of 82.8%, 92.5%, and 67.9%, respectively.

    “The ratio of direct to population effect estimates is 0.910 (S.E. = 0.009) for height, 0.962 (S.E. = 0.017) for BMI, and 0.824 (S.E. = 0.033) for cognitive performance, implying that 82.8%, 92.5%, and 67.9%, respectively, of the PGIs’ R2 are due to their direct effects”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In Add Health, the PGI explains 8.7% of the variation in Peabody verbal test scores and 12.3% in overall GPA.

    “For example, in Add Health, we found that the PGI explains 8.7% of the variation in Peabody verbal test scores and 12.3% in overall GPA.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The EA and disease-specific PGIs contribute roughly independently to predicting disease risk; the incremental R2 from adding both PGIs and their interaction is typically roughly equal to the sum of their separate incremental R2s.

    “Moreover, the EA and disease-specific PGIs contribute roughly independently to predicting disease risk: the incremental R2 from adding both PGIs and their interaction to the regression model is typically roughly equal to the sum of the incremental R2’s of each of the two PGIs considered separately.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingreasoning

    The larger sample size yields more accurate effect-size estimates that allow construction of a PGI with greater prediction accuracy, increasing the percentage of EA variance explained from 11-13% to 12-16%, an increase of approximately 20%.

    “The larger sample size yields more accurate effect-size estimates that allow us to construct a genome-wide polygenic index (PGI) (also called a polygenic score) that has greater prediction accuracy, increasing the percentage of variance in EA explained from 11-13% to 12-16%, depending on the validation sample, an increase of approximately 20%.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    A conditional-joint (COJO) analysis identified 2,925 SNPs, 41 of which are in LD (r2 > 0.1) with other COJO lead SNPs.

    “This analysis identified 2,925 SNPs (Supplementary Table 2); 41 of these are in LD (r2 > 0.1) with other COJO lead SNPs”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For height, phenotypic assortment predicts a mate-pair PGI correlation of 0.087 (S.E. = 0.007), which is only somewhat smaller than the observed estimate of 0.106 and is contained within the 95% confidence interval.

    “For height, phenotypic assortment predicts a mate-pair PGI correlation of 0.087 (S.E. = 0.007) (the gray point in the figure), which is only somewhat smaller than the observed estimate of 0.106 and is contained within the 95% confidence interval.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For EA, phenotypic assortment predicts a mate-pair PGI correlation of 0.031 (S.E. = 0.004), much smaller than and statistically distinguishable from the observed correlation of 0.175.

    “In contrast, for EA, the predicted value of 0.031 (S.E. = 0.004) is much smaller than, and statistically distinguishable from, the mate-pair PGI correlation of 0.175.”
    Quote in the paperSubstantial claim

  • The quote only partly states this claim.

    Findingown data

    Across the other 22 phenotypes, the inverse-variance-weighted average ratio of direct to population effects of the EA PGI is 0.588 (S.E. = 0.013).

    “Across the phenotypes, the inverse-variance-weighted average ratio of direct to population effects is 0.588 (S.E. = 0.013).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In WLS, the PGI explains 6.1% of the variation in Henmon-Nelson test scores and 7.7% in high school grade percentile rank.

    “In WLS, the PGI explains 6.1% of the variation in Henmon-Nelson test scores and 7.7% in high school grade percentile rank.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingreasoning

    For predicting EA, the ratio of direct to population effect estimates is 0.556 (S.E. = 0.020), implying that 30.9% of the PGI's R2 is due to its direct effect.

    “For predicting EA, the ratio of direct to population effect estimates is 0.556 (S.E. = 0.020), implying that 100% * 0.5562 = 30.9% of the PGI’s R2 is due to its direct effect.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The dominance SNP heritability estimate is 0.00015 (S.E. = 0.00024), statistically indistinguishable from zero (P = 0.54).

    “we estimate a SNP heritability of 0.00015 (S. E. = 0.00024), which is statistically indistinguishable from zero (P = 0.54).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    The correlation between mate pairs' PGIs for EA is far too large to be consistent with phenotypic assortment alone, implying additional assortment on PGI-associated factors.

    “The correlation between mate pairs’ PGIs is far too large to be consistent with phenotypic assortment alone, implying additional assortment on PGI-associated factors.”
    • Also stated in: § Assortative Mating · plain
      “In contrast, for EA, the predicted value of 0.031 (S.E. = 0.004) is much smaller than, and statistically distinguishable from, the mate-pair PGI correlation of 0.175.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    A transformation was applied to the meta-analysis that accounts for assortative mating to estimate the population effect of the PGI and the difference between direct and population effects.

    “We applied a transformation to the meta-analysis that accounts for assortative mating to estimate the population effect of the PGI and the difference between the direct and population effects.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    Any model of EA that requires substantial dominance to fit the data, restricts gene-environment correlations to zero, or assumes assortative mating is purely based on phenotype is likely to be misspecified.

    “Overall, these findings suggest that any model of EA that requires substantial dominance to fit the data, restricts gene-environment correlations to zero, or assumes assortative mating is purely based on phenotype is likely to be misspecified.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    Any factor that contributes to explaining the mate-pair PGI correlation must be correlated with the EA PGI; therefore, these factors likely contribute to the EA PGI's predictive power for EA and other phenotypes.

    “Any factor that contributes to explaining the mate-pair PGI correlation must be correlated with the EA PGI. Therefore, these factors likely contribute to the EA PGI’s predictive power for EA and other phenotypes.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Both for predicting EA and other phenotypes, a substantial part of the EA PGI's predictive power results from direct effects, but a substantial part results from factors other than direct effects.

    “Thus, both for predicting EA and other phenotypes, a substantial part of the EA PGI’s predictive power results from direct effects, but a substantial part results from factors other than direct effects.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Further analyses adjusted for genetic PCs, birth coordinates, UKB assessment center, cognitive performance, and vocabulary to test whether assortative mating on factors related to ancestry, geography, and cognition explained the mate-pair PGI correlations.

    “We performed further analyses adjusting for genetic PCs, birth coordinates, UKB assessment center, cognitive performance, and vocabulary to test whether assortative mating on factors related to ancestry, geography, and cognition explained the mate-pair PGI correlations (Supplementary Note section 9).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    Throughout the PGI distribution, there is substantial variation in EA at the individual level, so the PGI cannot be used to meaningfully predict an individual's EA.

    “Thus, while average EA varies substantially across the PGI distribution, the PGI cannot be used to meaningfully predict an individual’s EA.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    A series of pre-registered replication exercises were conducted to assess whether estimates of dominance effects for various subsets of SNPs are consistent across UKB and 23andMe.

    “We also conducted a series of pre-registered replication exercises (https://osf.io/uegqv/) to assess whether the estimates of the dominance effects for various subsets of SNPs are consistent across UKB and 23andMe (Supplementary Note section 4.6 and 8, Supplementary Table 19).”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    The meta-analysis of siblings and trios accounted for covariance between estimates from the sibling and trio samples from the same datasets.

    “We meta-analyzed the results from the siblings and trios, accounting for covariance between the estimates from the sibling and trio samples from the same datasets.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    After the dominance meta-analysis, standard errors were inflated by the square root of the intercept from an LD score regression.

    “Similar to the additive GWAS, after the meta-analysis we inflated the standard errors by the square root of the intercept from an LD score regression.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    There is compelling evidence from prior work that the male-female genetic correlation for EduYears is close to one; the Okbay et al. data yields an estimate of 0.98 (S.E. = 0.029).

    “We did not run sex-stratified analyses for the autosomal meta-analysis because there is compelling evidence from our prior work that the male-female genetic correlation for EduYears is close to one. For example, the Okbay et al.62 data yields an estimate of 0.98 (S.E. = 0.029).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Okbay et al., 62

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    We removed 621 individuals from GS that had been previously identified by GS as being also present in UKB.

    “We removed 621 individuals from GS that had been previously identified by GS as being also present in UKB (Supplementary Note section 7.3).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationnone given

    The results do not rule out the possibility that rare SNPs have substantial dominance effects.

    “Our results do not rule out the possibility that rare SNPs have substantial dominance effects.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductown data

    PGIs for direct/population effect analyses were made using GWAS results that exclude GS, STR and all related individuals of up to third degree from UKB.

    “PGIs were made using GWAS results that exclude GS, STR and all related individuals of up to third degree from UKB (Supplementary Note section 7.3), following the LDpred PGI pipeline described in Supplementary Note section 5.1.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    For GS and STR analyses, covariates included the top 20 genetic PCs.

    “For GS and STR, we included the top 20 genetic PCs (see Supplementary Note section 5.3 for how the PCs were created).”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductown data

    In siblings and trios, a linear mixed model was used to account for relatedness.

    “In both the siblings and trios, we used a linear mixed model to account for relatedness in the samples.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    In analyses of grade point average outcomes in Add Health, high school fixed effects were also controlled.

    “In our analyses of grade point average outcomes in Add Health, we also controlled for high school fixed effects (Supplementary Note section 5.3).”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    All regressions of phenotypes on PGIs included controls for year of birth or age at assessment, sex, their interactions, and the first 10 principal components.

    “All regressions include controls for year of birth or age at assessment, sex, their interactions, and the first 10 principal components (PCs) of the variance-covariance matrix of the genomic relatedness matrix.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation

    LD scores used for dominance effects account for the faster decay of information from tagged SNPs as a function of LD.

    “We used LD scores that account for the faster decay of information from tagged SNPs as a function of LD for dominance effects (see, e.g., ref. 13).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites ref. 13

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Interpretationown data

    The EA PGI's predictive power derives both from direct genetic effects and from gene-environment correlation (likely including indirect genetic effects from relatives), with assortative mating amplifying the predictive power over what would be expected under random mating.

    “Our results show that this predictive power derives both from direct genetic effects and from gene-environment correlation (likely including indirect genetic effects from relatives), with assortative mating amplifying the predictive power over what would be expected under random mating.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationnone given

    The correlation between the residual of the father's PGI after regression onto the father's phenotype and the residual of the mother's PGI after regression onto the mother's phenotype should be zero under phenotypic assortment if the relationship between phenotype and PGI is linear.

    “We also estimated the correlation between the residual of the father’s PGI after regression onto the father’s phenotype and the residual of the mother’s PGI after regression onto the mother’s phenotype, which should be zero under phenotypic assortment if the relationship between phenotype and PGI is linear.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The relationship between phenotype and PGI is linear.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Odds ratios and 95% confidence intervals for diseases were estimated using logistic regression while controlling for covariates.

    “Odds ratios and 95% confidence intervals were estimated using logistic regression while controlling for covariates (Supplementary Note section 6.2.1).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationnone given

    For some purposes, such as attaining greater predictive power for the PGI, there are clearly diminishing returns.

    “For some purposes, such as attaining greater predictive power for the PGI, there are clearly diminishing returns.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S4 (based).

    Interpretationnone given

    The correlation between mate pairs' PGIs should equal ryrprm if the correlation is explained by assortative mating on the phenotype alone and the relationship between PGI and phenotype is linear.

    “The correlation between the mate pairs’ PGIs should be equal to ryrprm if the correlation is explained by assortative mating on the phenotype alone, and the relationship between the PGI and the phenotype is linear.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The relationship between phenotype and PGI is linear.
51 of 51

  • The paper relies on this without support.

    Cognitive performance and vocabulary measures in Generation Scotland are proxies for the composite phenotype on which mate pairs assort.

    “The Generation Scotland cohort contains high quality measures of cognitive performance and vocabulary, proxies for plausible candidates of such a composite.”

  • The paper relies on this without support.

    The top 40 principal components capture the genetic ancestry relevant to assortative mating.

    “After residualizing the EA PGI on 40 principal components (PCs) of the genomic relatedness matrix in addition to EA, we find that the mate-pair PGI correlation falls to 0.091 (S.E. = 0.021).”

  • The authors state this as an assumption.

    The SBayesR finite mixture model has 4 components with initial mixture probabilities π = (0.95, 0.02, 0.02, 0.01) and fixed γ = (0.0, 0.01, 0.1, 1).

    “We assume 4 components in the finite mixture model, with initial mixture probabilities π = (0.95, 0.02, 0.02, 0.01) and fixed γ = (0.0, 0.01, 0.1, 1), where γ is a parameter that constrains how the SNP-effect-size variance scales in each of the four distributions.”

  • The paper relies on this without support.

    In the expected prediction accuracy model, A may vary by prediction sample but B does not.

    “While A may vary by prediction sample, B does not.”

  • The paper relies on this without support.

    The relationship between phenotype and PGI is linear.

    “The correlation between the mate pairs’ PGIs should be equal to ryrprm if the correlation is explained by assortative mating on the phenotype alone, and the relationship between the PGI and the phenotype is linear.”
5 of 5

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Genetic determinants of education may partly operate through indirect familial influences and socioeconomic factors.

    “Genetic determinants of education may partly operate through indirect familial influences and socioeconomic factors34.”

    From Locus for severity implicates CNS resilience in progression of multiple sclerosis (2023) · cites this paper as [34]

  • Loci reaching GWAS and nominal significance and SNPs in LD with them at D' ≥ 0.8 show predominant associations with traits such as cognitive decline, educational attainment, mean bone density, intelligence and performance intelligence quotient, and soluble TREM-2 levels.

    “Analysis of loci that reach GWAS and nominal significance and SNPs in LD with them at D’ ≥ 0.8 reveals predominant associations with traits such as cognitive decline [12, 68, 69], educational attainment [70–73], mean bone density [57, 74], intelligence and performance intelligence quotient [75, 76] and soluble TREM-2 levels [77].”

    From Genetic drivers of progression in Alzheimer’s disease are distinct from disease risk (2026) · cites this paper as [73]

  • The largest GWAS meta-analysis for EduYears, including approximately 3 million individuals of European genetic ancestries, identified 3,952 independent genome-wide significant loci.

    “Notably, the largest GWAS meta-analysis for EduYears, which included approximately 3 million individuals of European (EUR) genetic ancestries, has identified 3,952 independent genome-wide significant loci10.”

    From Shared genetic architectures of educational attainment in East Asian and European populations (2024) · cites this paper as [10]

  • All 11 independent SNPs identified in the EAS GWAS meta-analysis were previously reported, located within ±500 kb of the 3,952 lead SNPs reported by Okbay et al.

    “All these 11 independent SNPs were previously reported (SNPs located within ±500 kb of the 3,952 lead SNPs reported by Okbay et al.10).”

    From Shared genetic architectures of educational attainment in East Asian and European populations (2024) · cites this paper as Okbay et al. [10]

  • The two previously unreported SNPs were located beyond ±500 kb of lead SNPs reported in previous EduYears GWAS.

    “they were located beyond ±500 kb of the lead SNPs reported in previous EduYears GWAS (Supplementary Figs. 10 and 11)10,14.”

    From Shared genetic architectures of educational attainment in East Asian and European populations (2024) · cites this paper as [10]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Figure caption misidentifies the PGI training sample size

    Minor

    The figure caption says the PGIs were trained in 6,514 European-genetic-ancestry UKB participants, whereas the methods give 425,231 as the European-genetic-ancestry GWAS sample size. The methods identify 6,514 as the African-genetic-ancestry holdout sample.

1 of 1

Cites

Flagged first, then by citations

Research works printed but not linked: 1

  1. Bulmer, M. G. The Mathematical Theory of Quantitative Genetics (Clarendon Press, 1980).

Cited by

Sorted by citations

Related

Authors
Okbay, A., Wu, Y., Wang, N., Jayashankar, H., Bennett, M., Nehzati, S. M., Sidorenko, J., Kweon, H., Goldman, G., Gjorgjieva, T., Jiang, Y., Hicks, B., Tian, C., Hinds, D. A., Ahlskog, R., Magnusson, P. K. E., Oskarsson, S., Hayward, C., Campbell, A., Porteous, D. J., Freese, J., Herd, P., 23andMe Research Team, Agee, M., Alipanahi, B., Auton, A., Bell, R. K., Bryc, K., Elson, S. L., Fontanillas, P., Furlotte, N. A., Hinds, D. A., Huber, K. E., Kleinman, A., Litterman, N. K., McCreight, J. C., McIntyre, M. H., Mountain, J. L., Northover, C. A. M., Pitts, S. J., Sathirapongsasuti, J. F., Sazonova, O. V., Shelton, J. F., Shringarpure, S., Tung, J. Y., Vacic, V., Wilson, C. H., Social Science Genetic Association Consortium, Fontana, M. A., Pers, T. H., Rietveld, C. A., Chen, G.-B., Emilsson, V., Meddens, S. F. W., Pickrell, J. K., Thom, K., Timshel, P., de Vlaming, R., Abdellaoui, A., Ahluwalia, T. S., Bacelis, J., Baumbach, C., Bjornsdottir, G., Brandsma, J. H., Concas, M. P., Derringer, J., Galesloot, T. E., Girotto, G., Gupta, R., Hall, L. M., Harris, S. E., Hofer, E., Horikoshi, M., Huffman, J. E., Kaasik, K., Kalafati, I. P., Karlsson, R., Lahti, J., van der Lee, S. J., de Leeuw, C., Lind, P. A., Lindgren, K.-O., Liu, T., Mangino, M., Marten, J., Mihailov, E., Miller, M. B., van der Most, P. J., Oldmeadow, C., Payton, A., Pervjakova, N., Peyrot, W. J., Qian, Y., Raitakari, O., Rueedi, R., Salvi, E., Schmidt, B., Schraut, K. E., Shi, J., Smith, A. V., Poot, R. A., Pourcain, B. S., Teumer, A., Thorleifsson, G., Verweij, N., Vuckovic, D., Wellmann, J., Westra, H.-J., Yang, J., Zhao, W., Zhu, Z., Alizadeh, B. Z., Amin, N., Bakshi, A., Baumeister, S. E., Biino, G., Bønnelykke, K., Boyle, P. A., Campbell, H., Cappuccio, F. P., Davies, G., De Neve, J.-E., Deloukas, P., Demuth, I., Ding, J., Eibich, P., Eisele, L., Eklund, N., Evans, D. M., Faul, J. D., Feitosa, M. F., Forstner, A. J., Gandin, I., Gunnarsson, B., Halldórsson, B. V., Harris, T. B., Heath, A. C., Hocking, L. J., Holliday, E. G., Homuth, G., Horan, M. A., Hottenga, J.-J., de Jager, P. L., Joshi, P. K., Jugessur, A., Kaakinen, M. A., Kähönen, M., Kanoni, S., Keltigangas-Järvinen, L., Kiemeney, L. A. L. M., Kolcic, I., Koskinen, S., Kraja, A. T., Kroh, M., Kutalik, Z., Latvala, A., Launer, L. J., Lebreton, M. P., Levinson, D. F., Lichtenstein, P., Lichtner, P., Liewald, D. C. M., Study, L. C., Loukola, A., Madden, P. A., Mägi, R., Mäki-Opas, T., Marioni, R. E., Marques-Vidal, P., Meddens, G. A., McMahon, G., Meisinger, C., Meitinger, T., Milaneschi, Y., Milani, L., Montgomery, G. W., Myhre, R., Nelson, C. P., Nyholt, D. R., Ollier, W. E. R., Palotie, A., Paternoster, L., Pedersen, N. L., Petrovic, K. E., Räikkönen, K., Ring, S. M., Robino, A., Rostapshova, O., Rudan, I., Rustichini, A., Salomaa, V., Sanders, A. R., Sarin, A.-P., Schmidt, H., Scott, R. J., Smith, B. H., Smith, J. A., Staessen, J. A., Steinhagen-Thiessen, E., Strauch, K., Terracciano, A., Tobin, M. D., Ulivi, S., Vaccargiu, S., Quaye, L., van Rooij, F. J. A., Venturini, C., Vinkhuyzen, A. A. E., Völker, U., Völzke, H., Vonk, J. M., Vozzi, D., Waage, J., Ware, E. B., Willemsen, G., Attia, J. R., Bennett, D. A., Berger, K., Bertram, L., Bisgaard, H., Boomsma, D. I., Borecki, I. B., Bültmann, U., Chabris, C. F., Cucca, F., Cusi, D., Deary, I. J., Dedoussis, G. V., van Duijn, C. M., Eriksson, J. G., Franke, B., Franke, L., Gasparini, P., Gejman, P. V., Gieger, C., Grabe, H.-J., Gratten, J., Groenen, P. J. F., Gudnason, V., van der Harst, P., Hoffmann, W., Hyppönen, E., Iacono, W. G., Jacobsson, B., Järvelin, M.-R., Jöckel, K.-H., Kaprio, J., Kardia, S. L. R., Lehtimäki, T., Lehrer, S. F., Martin, N. G., McGue, M., Metspalu, A., Pendleton, N., Penninx, B. W. J. H., Perola, M., Pirastu, N., Pirastu, M., Polasek, O., Posthuma, D., Power, C., Province, M. A., Samani, N. J., Schlessinger, D., Schmidt, R., Sørensen, T. I. A., Spector, T. D., Stefansson, K., Thorsteinsdottir, U., Thurik, A. R., Timpson, N. J., Tiemeier, H., Uitterlinden, A. G., Vitart, V., Vollenweider, P., Weir, D. R., Wilson, J. F., Wright, A. F., Conley, D. C., Krueger, R. F., Smith, G. D., Hofman, A., Laibson, D. I., Medland, S. E., Yang, J., Esko, T., Watson, C., Jala, J., Conley, D., Koellinger, P. D., Johannesson, M., Laibson, D., Meyer, M. N., Lee, J. J., Kong, A., Yengo, L., Cesarini, D., Turley, P., Visscher, P. M., Beauchamp, J. P., Benjamin, D. J., & Young, A. I.
Affiliations
  • Vrije Universiteit Amsterdam (NL)
  • The University of Queensland (AU)
  • National Bureau of Economic Research (US)
  • University of California, Los Angeles (US)
  • 23andMe (United States) (US)
  • Uppsala University (SE)
  • Karolinska Institutet (SE)
  • Western General Hospital (GB)
  • Georgetown University (US)
  • University of Southern California (US)
  • Statens Serum Institut (DK)
  • Erasmus MC (NL)
  • University of Iceland (IS)
  • New York Genome Center (US)
  • New York University (US)
  • University of Copenhagen (DK)
  • Helmholtz Munich (DE)
  • Institute of Genetic and Biomedical Research (IT)
  • University of Illinois Urbana-Champaign (US)
  • Radboud University Nijmegen (NL)
  • University of Trieste (IT)
  • University of Helsinki (FI)
  • University of Leicester (GB)
  • University of Edinburgh (GB)
  • University of Graz (AT)
  • Centre for Human Genetics (GB)
  • Harokopio University of Athens (GR)
  • QIMR Berghofer Medical Research Institute (AU)
  • Max Planck Institute for Human Development (DE)
  • King's College London (GB)
  • University of Tartu (EE)
  • University of Minnesota (US)
  • Hunter Medical Research Institute (AU)
  • University of Manchester (GB)
  • Finnish Institute for Health and Welfare (FI)
  • GGZ inGeest (NL)
  • National Institute on Aging (US)
  • University of Turku (FI)
  • SIB Swiss Institute of Bioinformatics (CH)
  • University of Milan (IT)
  • Essen University Hospital (DE)
  • National Cancer Institute (US)
  • Universitätsmedizin Greifswald (DE)
  • University Medical Center Groningen (NL)
  • University of Münster (DE)
  • Broad Institute (US)
  • Rush University Medical Center (US)
  • University of Michigan (US)
  • Istituto di Genetica Molecolare (IT)
  • University of Warwick (GB)
  • Queen Mary University of London (GB)
  • Charité – Universitätsmedizin Berlin (DE)
  • German Institute for Economic Research (DE)
  • Translational Research Institute (AU)
  • Washington University in St. Louis (US)
  • University of Bonn (DE)
  • deCODE Genetics (Iceland) (IS)
  • Reykjavík University (IS)
  • National Institutes of Health (US)
  • University of Aberdeen (GB)
  • Centre Hospitalier Universitaire Vaudois (CH)
  • Norwegian Institute of Public Health (NO)
  • Imperial College London (GB)
  • Tampere University (FI)
  • University of Amsterdam (NL)
  • Stanford Medicine (US)
  • University of Toronto (CA)
  • University of Bristol (GB)
  • Queensland University of Technology (AU)
  • Medical University of Graz (AT)
  • IRCCS Materno Infantile Burlo Garofolo (IT)
  • Harvard University (US)
  • Endeavor Health (US)
  • University of Dundee (GB)
  • Maastricht University (NL)
  • Florida State University (US)
  • Autism & Developmental Medicine Institute (US)
  • University of South Australia (AU)
  • Oulu University Hospital (FI)
  • Princeton University (US)
  • University of Wisconsin–Madison (US)
  • Geisinger Medical Center (US)
  • George Mason University (US)
Corresponding author
  • Okbay, Aysu
  • Visscher, Peter M.
  • Benjamin, Daniel J.
  • Young, Alexander I.
ORCID
Venue
Nature Genetics
Publisher
Springer Science and Business Media LLC
Published
2022-03-31
Volume
54
Issue
4
Pages
437-449
ISSN
  • 1061-4036 (ISSN-L)
  • 1546-1718
DOI
10.1038/s41588-022-01016-z
PMC
PMC9005349
OpenAlex
W4220711675
Cited by
654
References
69
Citations by year
2026: 103 · 2025: 183 · 2024: 179 · 2023: 144 · 2022: 43 · 2021: 2
FWCI
83.18
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Genetics (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Discovery Research Portal (University of Dundee) (publishedVersion, cc-by)
  • Amsterdam UMC (VU Amsterdam) - Institutional Repository (publishedVersion, other-oa)
  • Research at the University of Copenhagen (University of Copenhagen) (publishedVersion, cc-by)
  • EUR Research Repository (Erasmus University Rotterdam) (publishedVersion, cc-by)
  • VU Research Portal (publishedVersion, other-oa)
  • e-space (Manchester Metropolitan University) (submittedVersion)
  • Edinburgh Research Explorer (University of Edinburgh) (submittedVersion, cc-by)
  • Radboud Repository (Radboud University) (submittedVersion)
  • ArTS Archivio della ricerca di Trieste (University of Trieste https://www.units.it/) (submittedVersion, cc-by)
  • Data Archiving and Networked Services (DANS) (submittedVersion, other-oa)
  • EUR Research Repository (Erasmus University Rotterdam) (submittedVersion, cc-by)
  • KTH Publication Database DiVA (KTH Royal Institute of Technology) (submittedVersion, other-oa)
  • eScholarship (California Digital Library) (submittedVersion, cc-by)
  • NORA - Norwegian Open Research Archives (submittedVersion, other-oa)
  • Figshare (submittedVersion, cc-by)
  • Työväentutkimus Vuosikirja (submittedVersion)
  • UnissResearch (Università degli Studi di Sassari) (submittedVersion)
  • University of Groningen research database (University of Groningen / Centre for Information Technology) (submittedVersion, cc-by)
  • PuSH - Publication Server of Helmholtz Zentrum München (submittedVersion, cc-by)
  • OPUS (Augsburg University) (submittedVersion, cc-by)
  • Radboud Repository (Radboud University) (submittedVersion, other-oa)
  • NOVA (University of Newcastle Australia) (submittedVersion)
Funders
  • U.S. Department of Health & Human Services | NIH | National Institute on Aging (R56-AG058726, R24-AG065184, R01-AG042568, K99-AG062787-01)
  • Nederlandse Organisatie voor Wetenschappelijk Onderzoek (016.Veni.198.058)
  • F.G. Meade Scholarship and UQ Research Training Scholarship from the University of Queensland Senate
  • Riksbankens Jubileumsfond (P18-0782:18)
  • Vetenskapsrådet (2019-00244, 421-2013-1061)
  • RCUK | Medical Research Council (MC_UU_00007/10)
  • ERC Consolidator Grant
  • Pershing Square Foundation
  • Li Ka Shing Foundation
  • Department of Education and Training | Australian Research Council (DE200100425, FL180100072)
  • Ragnar Söderbergs stiftelse (E42/15)
  • Department of Health | National Health and Medical Research Council (GNT113400)
  • Genome Canada (OGI-152)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Associations and Epidemiology
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic and phenotypic traits in livestock
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Mapping and Diversity in Plants and Animals
Keywords
educational attainment, polygenic index, polygenic prediction, single nucleotide polymorphisms, direct genetic effects, parental effects, assortative mating, dominance effects, X chromosome, genetic risk prediction, disease risk prediction
MeSH terms
Humans, Multifactorial Inheritance, Polymorphism, Single Nucleotide, Genome-Wide Association Study

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-29.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-29.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review