CoChat
Sign Up
Cited by 317Genetic Associations and EpidemiologyNeurotransmitter Receptor Influence on BehaviorGenetic Mapping and Diversity in Plants and Animals

Multivariate analysis of 1.5 million people identifies genetic associations with traits related to self-regulation and addiction.

Open access
Karlsson Linnér, Richard, Mallard, Travis T., Barr, Peter B., Sanchez-Roige, Sandra, Madole, James W., Driver, Morgan N., Poore, Holly E., de Vlaming, Ronald +69 more·2021·Nature Neuroscience
DOI 10.1038/s41593-021-00908-3PMC PMC8484054
Open this paper in CoChat — free accountOpen PDF

Abstract

Behaviors and disorders related to self-regulation, such as substance use, antisocial behavior and attention-deficit/hyperactivity disorder, are collectively referred to as externalizing and have shared genetic liability. We applied a multivariate approach that leverages genetic correlations among externalizing traits for genome-wide association analyses. By pooling data from ~1.5 million people, our approach is statistically more powerful than single-trait analyses and identifies more than 500 genetic loci. The loci were enriched for genes expressed in the brain and related to nervous system development. A polygenic score constructed from our results predicts a range of behavioral and medical outcomes that were not part of genome-wide analyses, including traits that until now lacked well-performing polygenic scores, such as opioid use disorder, suicide, HIV infections, criminal convictions and unemployment. Our findings are consistent with the idea that persistent difficulties in self-regulation can be conceptualized as a neurodevelopmental trait with complex and far-reaching social and health correlates.

Integrity signals

Registry

Crossref · 10.1038/s41593-021-00908-3

Cites

99 of 99 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

17 of 27 cited sources checked · 8 supported · 0 flagged · accuracy not yet measured

  • Premisecitation

    For many high-cost, high-risk externalizing behaviors, such as opioid use disorder and suicide attempts, limited genotyped cases are available for gene discovery.

    “For many high-cost, high-risk behaviors with an externalizing component – opioid use disorder and suicide attempts7 being salient examples – there are limited genotyped cases available for gene discovery8,9.”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [8] Zhou, H. et al. Association of OPRM1 functional coding variant with opioid use disorder: a genome-wide association study. JAMA Psychiatry https://doi.org/10.1001/jamapsychiatry.2020.1206 (2020). source
    • cites [9] Mullins, N. et al. GWAS of suicide attempt in psychiatric disorders and association with major depression polygenic risk scores. Am. J. Psychiatry 176, 651–660 (2019). source
  • Premisecitation

    Multivariate analyses boost statistical power by pooling data across genetically correlated traits.

    “A complementary strategy to the single-disease approach is to study the shared genetic architecture across traits in multivariate analyses, which boosts statistical power by pooling data across genetically correlated traits10.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [10] Grotzinger, A. D. et al. Genomic structural equation modelling provides insights into the multivariate genetic architecture of complex traits. Nat. Hum. Behav. 3, 513–525 (2019). source
  • Premisecitation

    Externalizing traits share genetic liability.

    “Collectively, this group of correlated traits are classified as externalizing3. Twin studies have demonstrated that externalizing liability is highly heritable (~80%)4,5.”
    • Also stated in: § Introduction · plain
      “This approach was grounded in the literature showing shared genetic liability across numerous externalizing disorders and with non-psychiatric variation in externalizing behavior5,11.”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [5] Krueger, R. F. et al. Etiologic connections among substance dependence, antisocial behavior and personality: modeling the externalizing spectrum. J. Abnorm. Psychol. 111, 411–424 (2002). source
    • cites [11] Kendler, K. S. & Myers, J. The boundaries of the internalizing and externalizing genetic spectra in men and women. Psychol. Med. 44, 647–655 (2013). source
  • Premisecitation

    Before this study, no large-scale molecular genetic studies had used the genetic overlap among externalizing traits to aid gene discovery, as most studies focused on individual disorders.

    “To date, however, no large-scale molecular genetic studies have utilized the extensive degree of genetic overlap among externalizing traits to aid gene discovery, as most studies have focused on individual disorders6.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [6] Buniello, A. et al. The NHGRI-EBI GWAS Catalog of published genome-wide association studies, targeted arrays and summary statistics 2019. Nucleic Acids Res. 47, D1005–D1012 (2018). source

      The Introduction (page not supplied) discusses GWAS of individual traits and the use of trait pleiotropy to discover loci, but does not address studies combining externalizing traits. Searches for externalizing-trait overlap and large-scale studies focused on individual disorders found no passage establishing the claimed gap in prior research. (LLM, conf=0.76)

      Experimental — accuracy not yet measured
  • Premisecitation

    Genetic variation is probabilistically related to psychiatric, medical, and social outcomes, partly through environmental pathways that can vary by historical, political, and economic setting.

    “Genetic differences are probabilistically associated with psychiatric, medical, and social outcomes, in part via environmental mechanisms that might differ across historical, political, and economic contexts61.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [61] Tucker-Drob, E. M., Briley, D. A. & Harden, K. P. Genetic and environmental influences on cognition across development and context. Curr. Dir. Psychol. Sci. 22, 349–355 (2013). source

      The Transactional Models of Cognitive Development and Socioeconomic Differences in Heritability sections (pages unavailable) describe genetic influences on cognition operating through environmental experiences that vary with socioeconomic opportunity. The educational-attainment example also shows variation across historical settings. The paper does not establish the claim for psychiatric or medical outcomes, so the cited sentence extends beyond its scope. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
  • Classified as probably not a substantial claim. 19
  • May combine two claims. 47
  • The quote only partly states this claim. 3

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

    Premisecitation

    Other genes among the 34, including CADM2, ZIC4, GABRA2, NEGR1, and FURIN, have previously been identified in GWAS of behavioral or psychiatric traits.

    “Other genes among the 34 have previously been identified in GWAS of behavioral or psychiatric traits: Cell Adhesion Molecule 2 (CADM2, previously identified in GWAS related to self-regulation, including drug use and risk tolerance17,33), Zic Family Member 4 (ZIC4, associated with brain volume34), Gamma-Aminobutyric Acid Type A Receptor Subunit Alpha 2 (GABRA2; the site of action for alcohol and”
    Quote in the paperNot a substantial claim3 of 5 sources checked
    • cites [17] Karlsson Linnér, R. et al. Genome-wide association analyses of risk tolerance and risky behaviors in over 1 million individuals identify hundreds of loci and shared genetic influences. Nat. Genet. 51, 245–257 (2019). source

      The Introduction and Results describe GWAS of risk tolerance and risky behaviors, but the passages in our copy do not identify CADM2 as an associated gene. Searches for CADM2, its full name, and the other named genes found no passage establishing the gene-specific claim; page numbers were unavailable. (LLM, conf=0.73)

      Experimental — accuracy not yet measured
    • cites [33] Sanchez-Roige, S. et al. Genome-wide association studies of impulsive personality traits (BIS-11 and UPPS-P) and drug experimentation in up to 22,861 adult research participants identify loci in the CACNA1I and CADM2 genes. J. Neurosci. 39, 2562–2572 (2019). source
    • cites [34] Zhao, B. et al. Genome-wide association analysis of 19,629 individuals identifies variants influencing regional brain volumes and refines their genetic co-architecture with cognitive and mental health traits. Nat. Genet. 51, 1637–1644 (2019). source
    • cites [35] Edenberg, H. J. et al. Variations in GABRA2, encoding the α2 subunit of the GABAA receptor, are associated with alcohol dependence and with brain oscillations. Am. J. Hum. Genet. 74, 705–714 (2004). source
    • cites [36] Dick, D. M. et al. The role of GABRA2 in risk for conduct disorder and alcohol and drug dependence across developmental stages. Behav. Genet. 36, 577–590 (2006). source

  • May combine two claims.

    Premisecitation

    Genetic associations detected in GWAS can be due to direct genetic effects but can also be confounded by population stratification, indirect genetic effects, and assortative mating.

    “Genetic associations detected in GWAS can be due to direct genetic effects, but can also be confounded by population stratification, indirect genetic effects from e.g., parental environment, and assortative mating47,48.”
    Quote in the paperSubstantial claim1 of 2 sources checked
    • cites [47] Young, A. I., Benonisdottir, S., Przeworski, M. & Kong, A. Deconstructing the sources of genotype–phenotype associations in humans. Science 365, 1396–1400 (2019). source
    • cites [48] Kong, A. et al. The nature of nurture: effects of parental genotypes. Science 359, 424–428 (2018). source

  • May combine two claims.

    Premisecitation

    Genomic SEM is not biased by sample overlap or imbalanced sample size, and using summary statistics allows genetic analyses of latent factors with more observations than individual-level data typically permit.

    “Genomic SEM is unbiased by sample overlap and imbalanced sample size, and by applying to summary statistics allows for genetic analyses of latent factors with more observations than is typically possible with individual-level data10.”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites [10] Grotzinger, A. D. et al. Genomic structural equation modelling provides insights into the multivariate genetic architecture of complex traits. Nat. Hum. Behav. 3, 513–525 (2019). source

      The Results report that overlapping samples did not bias Genomic SEM parameter estimates or standard errors in simulations, and the SNP Effects section shows that summary-statistic analyses can use information from more participants than analyses requiring complete individual-level data. However, the paper does not establish that estimates are unbiased by imbalanced sample sizes: the Method (Stage 2 Estimation) cautions that when contributing GWASs differ substantially in power, one estimation approach may be dominated by the better-powered studies. (LLM, conf=0.85)

      Experimental — accuracy not yet measured

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

    Premisecitation

    Modern genetics research is often appropriated by white supremacist movements to argue that disparities by race in health, jobs, and criminal justice stem from genetic inferiority of people of color instead of environmental and historical disadvantages.

    “Modern genetics research is routinely appropriated by white supremacist movements to argue that racialized disparities in health, employment, and criminal justice system involvement are due to the genetic inferiority of people of color rather than environmental and historical disadvantages59.”
    Quote in the paperNot a substantial claim1 of 1 sources checked
    • cites [59] American Society of Human Genetics. ASHG denounces attempts to link genetics and racial supremacy. Am. J. Hum. Genet. 103, 636 (2018). 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.

    Study design and conductcitation

    Sibling comparisons overcome confounding from population stratification, indirect genetic effects, and assortative mating because meiosis randomizes genotypes to siblings.

    “While reducing statistical power, sibling comparisons overcome these methodological challenges, because meiosis randomizes genotypes to siblings47,49.”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [47] Young, A. I., Benonisdottir, S., Przeworski, M. & Kong, A. Deconstructing the sources of genotype–phenotype associations in humans. Science 365, 1396–1400 (2019). source

      In “Using family genotype data,” the paper says random Mendelian segregation makes sibling genetic differences unconfounded by population stratification, assortative mating, and indirect genetic effects from parents. It also says sibling comparisons estimate the direct effect minus any indirect effect from the sibling, so the unqualified claim that they overcome indirect genetic effects goes beyond the source. The same section supports the citing sentence’s reduced-power qualification. (LLM, conf=0.88)

      Experimental — accuracy not yet measured
    • cites [49] Selzam, S. et al. Comparing within- and between-family polygenic score prediction. Am. J. Hum. Genet. 105, 351–363 (2019). source

  • May combine two claims.

    Findingown data

    Results from LD Score regression favored polygenicity over stratification bias, with an LD Score intercept of 1.115 (SE=0.019) and attenuation ratio of 0.054 (SE=0.009).

    “LD Score regression suggested polygenicity rather than bias from population stratification10,12, as the (default settings) LD Score intercept and attenuation ratio were estimated to be 1.115 (SE = 0.019) and 0.054 (SE = 0.009), respectively.”
    Quote in the paperSubstantial claim2 of 2 sources checked
    • cites [10] Grotzinger, A. D. et al. Genomic structural equation modelling provides insights into the multivariate genetic architecture of complex traits. Nat. Hum. Behav. 3, 513–525 (2019). source

      The Results > SNP Effects section uses LD Score regression intercepts to conclude that its findings were not due to uncontrolled inflation, supporting the general interpretation of genetic signal rather than bias. It reports intercepts of 0.987 and 0.997 for its analyses, not the claimed intercept of 1.115, and does not report the claimed attenuation ratio of 0.054 or their standard errors. (LLM, conf=0.86)

      Experimental — accuracy not yet measured
    • cites [12] 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

      The Results (Real Data; page not provided) say that LD Score regression found polygenicity contributed substantially more than bias to GWAS test-statistic inflation. Our copy does not report the claimed intercept of 1.115 (SE = 0.019) or attenuation ratio of 0.054 (SE = 0.009); its schizophrenia example instead gives an intercept of 1.07. (LLM, conf=0.87)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingown data

    EXT had a positive genetic correlation with extraversion (rg = 0.32, SE = 0.03) and negative correlations with conscientiousness (rg = −0.23, SE = 0.04) and agreeableness (rg = −0.09, SE = 0.04).

    “positive correlation with extraversion (rg = 0.32, SE = 0.03), and negative with conscientiousness (rg = −0.23, SE = 0.04) and agreeableness (rg = −0.09, SE = 0.04)11,21.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    EXT was correlated with suicide attempts (rg = 0.68, SE = 0.08) and post-traumatic stress disorder (rg = 0.53, SE = 0.06).

    “Notably, EXT was also correlated with suicide attempts (rg = 0.68, SE = 0.08) and post-traumatic stress disorder (rg = 0.53, SE = 0.06).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    EXT showed strong positive genetic correlations with drug exposure (rg = 0.91, SE = 0.09), antisocial behavior (rg = 0.65, SE = 0.17), motor impulsivity (rg = 0.70, SE = 0.17), and failures to plan (rg = 0.68, SE = 0.13).

    “EXT showed strong positive genetic correlations with drug exposure (rg = 0.91, SE = 0.09), antisocial behavior (rg = 0.65, SE = 0.17), and impulsivity measures, including motor impulsivity (rg = 0.70, SE = 0.17) and failures to plan (rg = 0.68, SE = 0.13).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Unlike prior work, EXT showed a positive genetic correlation with openness (rg = 0.22, SE = 0.04) but not with neuroticism (rg = 0.02, SE = 0.05).

    “while we found a positive correlation with openness (rg = 0.22, SE = 0.04) but not with neuroticism (rg = 0.02, SE = 0.05).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    A revised common factor model with seven externalizing phenotypes provided the best fit across specifications, with χ2(12) = 390.234, AIC = 422.23, CFI = 0.957, and SRMR = 0.079.

    “The revised common factor model with seven externalizing phenotypes provided the best fit across all specifications, and it closely approximated the observed genetic covariance matrix (i.e., χ2(12) = 390.234, AIC = 422.23, CFI = 0.957, SRMR = 0.079).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    EXT had inverse genetic correlations with educational attainment (rg = −0.32, SE = 0.02) and intelligence (rg = −0.23, SE = 0.02).

    “EXT showed more modest inverse correlations with educational attainment (rg = −0.32, SE = 0.02) and intelligence (rg = −0.23, SE = 0.02)”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    A polygenic score using only the 579 EXT SNPs was associated with the phenotypic externalizing factor, explaining about 3-4% of variance (βAdd Health = 0.20, ΔR2 = 4.1%; βCOGA = 0.17, ΔR2 = 3.0%).

    “we created a polygenic score using only the 579 EXT SNPs (this score was only used for this quasi-replication exercise), and also this polygenic score was found associated with the phenotypic externalizing factor, explaining ~3–4% of the variance (βAdd Health = 0.20, 95% CI: 0.17–0.23, ΔR2 = 4.1%; βCOGA = 0.17, 95% CI: 0.13–0.20, ΔR2 = 3.0%).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The study identified 579 genome-wide significant loci associated with liability toward externalizing outcomes, 121 of which are entirely novel for any of the seven phenotypes analyzed.

    “We used multivariate genomic analyses to accelerate genetic discovery, identifying 579 genome-wide significant loci associated with a liability toward externalizing outcomes, 121 of which are entirely novel discoveries for any of the seven phenotypes analyzed.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The fourth factor explained only 12.5% of variance, so the three-factor solution was considered the best-fitting exploratory model.

    “The fourth factor explained only 12.5% variance, and thus, the three-factor solution was considered the best-fitting exploratory model”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Within-family estimates for the genome-wide EXT polygenic score predicting the phenotypic externalizing factor were smaller than OLS estimates but remained significant in Add Health (βWF = 0.12, P = 4.89×10−3) and COGA (βWF = 0.14, P = 1.87×10−6).

    “Parameter estimates from the within-family model (βWF Add Health = 0.12, 95% CI: 0.04–0.20; βWF COGA = 0.14, 95% CI: 0.08–0.20) were smaller compared to OLS models without family-specific intercepts (βAdd Health = 0.20, 95% CI, 0.16 to 0.24; βCOGA = 0.16, 95% CI, 0.12 to 0.20), but remained statistically significant (Add Health two-sided P = 4.89×10−3; COGA two-sided test P = 1.87×10−6).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The formal test of attenuation found the standardized difference between βWF and β was −1.988 (P = 0.047) in Add Health and −0.704 (P = 0.481) in COGA.

    “we evaluated the standardized difference between βWF and β (i.e., a Z-statistic assumed to be normally distributed, see Supplementary Information section 5.2.6) and found that it was −1.988 (two-sided P = 0.047) and −0.704 (two-sided P = 0.481) for the PRS-CS polygenic score in Add Health and COGA, respectively.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The second and third factors captured complex residual variation and divergent residual cross-trait correlations among phenotypes with the weakest loadings on the single common factor.

    “the second and third factor accounted for complex residual variation and divergent residual cross-trait correlations among the subset of phenotypes that had the weakest loadings on the single common factor.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Quasi-replication tests strongly rejected the null hypotheses, suggesting the externalizing GWAS is not spurious overall and is more enriched for association with second-stage phenotypes than their polygenic background GWAS signal.

    “We strongly rejected the null hypotheses in all quasi-replication tests, suggesting that the externalizing GWAS is not spurious overall and that it was more enriched for association with the second-stage phenotypes than their respective polygenic background GWAS signal.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Only model 4 closely fit the observed genetic covariance matrix (χ2(12)=390.234, AIC=422.234, CFI=0.957, SRMR=0.079), satisfied preregistered fit criteria, and aligned with theoretical expectations for a shared genetic liability underlying externalizing.

    “We found that model (4) was the only model that closely approximated the observed genetic covariance matrix (χ2(12) = 390.234, AIC = 422.234, CFI = 0.957, SRMR = 0.079), fulfilled our preregistered model fit criteria, and coalesced with theoretical expectations of a common genetic liability to externalizing.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The loci were enriched for genes expressed in the brain and related to nervous system development.

    “The loci were enriched for genes expressed in the brain and related to nervous system development.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In the UKB sibling hold-out, within-family estimates were on average the same for risky behavior (mean attenuation = 0.08; 95% CI −1.67 to 1.83) and only modestly attenuated for personality (mean attenuation = −0.35; 95% CI −1.06 to 0.36).

    “within-family estimates were, on average, the same for the risky behavior category (mean attenuation = 0.08; 95% CI: −1.67 to 1.83), and only attenuated modestly for personality (mean attenuation = −0.35; 95% CI: −1.06 to 0.36).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In the UKB sibling hold-out, within-family estimates attenuated more for cognitive ability (mean attenuation = −6.55; 95% CI −9.93 to −3.17), socioeconomic status (−2.43; 95% CI −4.39 to −0.48), and overall and reproductive health (−2.20; 95% CI −4.18 to −0.21).

    “the within-family estimates attenuated more for cognitive ability (mean attenuation −6.55; 95% CI: −9.93 to −3.17), socioeconomic status (mean attenuation −2.43; 95% CI: −4.39 to −0.48), and overall and reproductive health (mean attenuation −2.20; 95% CI: −4.18 to −0.21).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Higher EXT polygenic scores were associated with worse health in nearly every bodily system, including ischemic heart disease, viral hepatitis C, HIV infection, type 2 diabetes, obesity, cirrhosis, sepsis, and lung cancer.

    “Individuals with higher EXT polygenic scores also showed worse health in nearly every bodily system. They were more likely to suffer, for example, from ischemic heart disease, viral hepatitis C and HIV infection, type 2 diabetes and obesity, cirrhosis of liver, sepsis, and lung cancer.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Behaviors related to self-regulation, such as smoking, drinking, drug use, condomless sex, and overeating, causally contribute to many of the medical outcomes associated with the EXT polygenic score.

  • May combine two claims.

    Findingown data

    The genome-wide EXT polygenic score was associated with the phenotypic externalizing factor in Add Health (β = 0.33, 95% CI 0.30-0.36, ΔR2 = 10.5%) and COGA (β = 0.30, 95% CI 0.27-0.34, ΔR2 = 8.9%).

    “the genome-wide polygenic score was associated with the phenotypic factor in both data sets (βAdd Health = 0.33, 95% CI: 0.30–0.36, ΔR2 = 10.5%; βCOGA = 0.30, 95% CI: 0.27–0.34, ΔR2 = 8.9%; Figure 3A and Supplementary Table 28).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Thirty-four genes were consistently identified by all four gene-based methods, and 741 overlapped across two or more methods.

    “We found 34 genes that were consistently identified by all four methods, while 741 overlapped across two or more methods (Supplementary Table 22; Extended Data Fig. 7).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Polygenic scores from the GWAS results were associated with social outcomes such as lower employment and greater criminal justice system involvement.

    “Polygenic scores created using our GWAS results were associated not just with psychiatric and substance use disorders, but also with correlated social outcomes, such as lower employment and greater criminal justice system involvement, as well as with biomedical conditions affecting nearly every system in the body.”
    Quote in the paperSubstantial claim

  • May combine two claims. The quote only partly states this claim.

    Findingown data

    In Add Health, COGA, and PNC, genome-wide EXT polygenic scores explained significant variance in ADHD (mean ΔR2 = 1.65%), conduct disorder (3.1%), oppositional defiant disorder (1.96%), substance use initiation (1.3-6.5%), substance use disorders (0.8-1.7%), disinhibited behaviors (1.5-2.5%), criminal justice involvement (1.0-3.0%), reproductive health (0.3-3.7%), and socioeconomic attainment (0.1

    “Polygenic scores for EXT explained significant variance (ΔR2) in criteria counts of ADHD (mean ΔR2 = 1.65%), conduct disorder (CD; mean ΔR2 = 3.1%), and oppositional defiant disorder (ODD; ΔR2 = 1.96%), as well as in the categories substance use initiation (mean ΔR2 = 1.3–6.5%), substance use disorders (mean ΔR2 = 0.8–1.7%), disinhibited behaviors (mean ΔR2 = 1.5–2.5%), criminal justice system”
    Quote in the paperSubstantial claim

  • May combine two claims. The quote only partly states this claim.

    Findingcitation

    Among the 34 consistently identified genes, ALMS1 and ERAP2 are novel for the psychiatric or behavioral literature and were previously identified only in relation to non-psychiatric biomedical diseases.

    “Several of the 34 implicated genes are novel discoveries for the psychiatric/behavioral literature and have previously been identified only in relation to non-psychiatric biomedical diseases. Such discoveries include ALMS1 (previously associated with kidney function and urinary metabolites30), and ERAP2 (blood protein levels and autoimmune disease31,32).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For 317 of the 579 EXT SNPs (54.7%), effects were concordant across all seven phenotypes; 203 (35.1%), 47 (8.1%), and 12 (2.1%) had six, five, or four concordant effects, respectively.

    “For 317 of the 579 EXT SNPs (54.7%), the concordance was perfect (i.e., the same direction of effect on all seven phenotypes), and for 203 (35.1%), 47 (8.1%), and 12 (2.1%) we observed either six, five, or four concordant effects, respectively.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The quasi-replication polygenic score using the 579 EXT SNPs remained significant and basically did not attenuate in within-family models (standardized difference −0.338, P = 0.735 in Add Health; 0.07, P = 0.944 in COGA).

    “Also, the association of the quasi-replication polygenic score constructed with the 579 EXT SNPs remained significant and basically did not attenuate in within-family models (for this score, the standardized difference between βWF and β was −0.338 (two-sided P = 0.735) and 0.07 (two-sided P = 0.944) in Add Health and COGA, respectively).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In quasi-replication, 75.4% of the 579 SNPs showed sign concordance with alcohol use disorder (P = 6.84×10−36) and 66.9% with antisocial behavior (P = 1.39×10−15).

    “75.4% of SNPs showed sign concordance with alcohol use disorder (two-sided test P = 6.84×10−36) and 66.9% with antisocial behavior (two-sided test P = 1.39×10−15)”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    QSNP analysis detected considerable heterogeneity (160 near-independent genome-wide significant QSNP loci) but none among 99% (571/579) of the EXT SNPs.

    “The QSNP analysis identified substantial heterogeneity (160 near-independent genome-wide significant QSNP loci), but reassuringly, did not identify heterogeneity among 99% (571/579) of the EXT SNPs”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    In the BioVU PheWAS, 255 disease phenotypes were associated with the EXT polygenic score at false discovery rate < 0.05, with odds ratios from 0.8 to 1.4 per standard deviation increase in the score.

    “255 disease phenotypes were associated with the EXT polygenic score at false discovery rate <0.05, with odds ratios ranging from 0.8 to 1.4 per standard deviation increase in the score (Figure 4 and Supplementary Table 32).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The externalizing GWAS had an estimated effective sample size of 1,492,085, mean χ2 of 3.114, and genomic inflation factor λGC of 2.337.

    “The estimated effective sample size of the “externalizing GWAS” is Neff = 1,492,085, and the mean χ2 and genomic inflation factor (λGC) are 3.114 and 2.337, respectively.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The 11 phenotypes showed appreciable genetic overlap with at least one other phenotype, with maximum absolute genetic correlations from 0.245 to 0.773.

    “The 11 phenotypes displayed appreciable genetic overlap with at least one other phenotype (max |rg| = 0.245–0.773)”
    Quote in the paperSubstantial claim

  • The quote only partly states this claim.

    Interpretationown data

    The small minority of heterogeneous EXT loci and the majority of QSNP loci outside EXT loci provide evidence that EXT loci primarily index a unitary dimension of genetic liability rather than an amalgamation of variants with divergent associations.

    “provide evidence that the EXT loci primarily index a unitary dimension of genetic liability rather than representing an amalgamation of variants with divergent associations across the discovery phenotypes.”
    Quote in the paperSubstantial claim

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

    Study design and conductcitation

    The study used a preregistered analysis plan.

    “The study followed a preregistered analysis plan (https://doi.org/10.17605/OSF.IO/XKV36), which specified that we would generate new, or collect existing, single-phenotype genome-wide association study (GWAS) summary statistics on externalizing phenotypes”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    Additional exclusions were to be based on negligible SNP-based heritability or GWAS signal.

    “additional exclusions would be based on negligible SNP-based heritability or GWAS signal.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductnone given

    Two partly overlapping UK Biobank hold-out subsamples were excluded from all single-phenotype GWAS summary statistics that included UK Biobank data and retained for polygenic score analyses.

    “Two partly overlapping hold-out subsamples of UKB participants were excluded from all single-phenotype GWAS summary statistics that included UKB data, and the participants were instead retained as a hold-out sample for polygenic score analyses”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    Genetic relatives of the held-out individuals were excluded to ensure independence between discovery and follow-up analyses.

    “Genetic relatives (pairwise KING coefficient ≥ 0.0442, version 2.1.5) of the held-out individuals were excluded from the study altogether to ensure independence between the discovery and follow-up analyses.”
    Quote in the paperSubstantial claim

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

    Study design and conductreasoning

    The preregistered plan set a minimum sample size of N > 15,000.

    “To ensure satisfying statistical power, we preregistered a minimum sample-size of N > 15,000”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    Meta-analysis was sample-size weighted and ensured no sample overlap.

    “We performed sample-size weighted meta-analysis with METAL (versions 2011-03-25 and 2020-05-05)66, while ensuring absence of sample overlap”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    Some of the 11 indicators may not be optimal for identifying a common genetic liability to externalizing, and a simpler model with fewer indicators may perform better in confirmatory analyses.

    “Thus, we learned from exploratory analysis that some of the 11 indicators may not be optimal for identifying a common genetic liability to externalizing, and that a less complex model with fewer indicators may perform better in subsequent confirmatory analyses.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Premisecitation

    Connecting biology to socially valued behavioral outcomes can be politically sensitive.

    “Linking biology with socially-valued behavioral outcomes can be politically sensitive58.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [58] Fox, D. Subversive science. Penn State Law Rev. 124, 153–191 (2019).

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

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    Summary statistics with negligible SNP-based heritability (h2 < 0.05) or GWAS signal (χ ̄2 < 1.05) were excluded.

    “We excluded any summary statistics with negligible SNP-based heritability (h2 < 0.05) or GWAS signal (χ¯2 < 1.05), estimated with LD Score regression (version 1.0.0)12,55.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    The multivariate approach is statistically more powerful than single-trait analyses and identifies more than 500 genetic loci.

    “By pooling data from ~1.5 million people, our approach is statistically more powerful than single-trait analyses and identifies more than 500 genetic loci.”
    Quote in the paperSubstantial claim

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

    Study design and conductreasoning

    The study did not manipulate an experimental condition or collect new individual-level data, so it was neither randomized nor blinded.

    “The study did not manipulate an experimental condition or collect any new individual-level data, and thus, was neither randomized nor blinded.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    The EXT polygenic score remained nominally significant with 24 outcomes across five categories, showing that the externalizing GWAS captures genetic effects not solely due to uncontrolled population stratification, indirect genetic effects, or other environmental confounding.

    “Nonetheless, the EXT polygenic score remained nominally significant (two-sided P < 0.05) with 24 outcomes across the five categories, showing that the externalizing GWAS captures genetic effects that are not solely a consequence of uncontrolled population stratification, indirect genetic effects, or other forms of environmental confounding.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    The genetic correlations indicate convergent and discriminant validity of the common EXT factor.

    “The genetic correlations indicate convergent and discriminant validity of the common EXT factor (Figure 1C)”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    When an existing GWAS was based on UK Biobank, it was re-conducted with the same phenotype definition to generate summary statistics excluding the hold-out sample and their genetic relatives.

    “In summary, whenever an existing GWAS (or GWAS meta-analysis) was based on UKB, we re-conducted it using the same phenotype definition to generate summary statistics that excluded the hold-out sample and their genetic relatives.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    The quasi-replication exercises suggest that the GWAS of EXT is not spurious overall and is enriched for genetic signal with alcohol use disorder and antisocial behavior.

    “Overall, the three exercises consistently suggested that the GWAS of EXT is not spurious overall, and that it is enriched for genetic signal with two phenotypes of central importance to the literature on externalizing.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductown data

    Within-family analyses were performed among full siblings in Add Health, COGA, and the UK Biobank siblings hold-out cohort.

    “We performed within-family analyses among full siblings in Add Health, COGA, and the UKB siblings hold-out cohort”
    Quote in the paperSubstantial claim

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

    Study design and conductnone given

    A detailed definition of externalizing phenotypes was preregistered.

    “A detailed definition of “externalizing phenotypes” was preregistered to delimit the collection of single-phenotype summary statistics”
    Quote in the paperNot a substantial claim

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

    Study design and conductcitation

    The analyses followed a preregistered analysis plan.

    “Following our preregistered analysis plan (https://doi.org/10.17605/OSF.IO/XKV36), we collated summary statistics from GWAS on externalizing-related traits”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Premisereasoning

    Non-disease phenotypes, such as self-reported age at first sex, are often easier to measure in the general population than diagnostic status, which helps achieve large samples.

    “Pragmatically, non-disease phenotypes such as the ones we use here (e.g., self-reported age at first sex) are often easier to measure in the general population than diagnostic status, making it easier to achieve large sample sizes.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductcitation

    Quality control removed rare SNPs (minor allele frequency < 0.005), SNPs with imputation quality (INFO) below 0.9, SNPs that could not be mapped or had discrepant alleles, and otherwise low-quality variants.

    “Our quality-control procedure applied recommended65 SNP-filtering to (i) remove rare SNPs (minor allele frequency < 0.005), (ii) SNPs with an IMPUTE imputation quality (INFO) score less than 0.9, (iii) SNPs that could not be mapped to or had discrepant alleles with the reference panel, and (iv) otherwise low-quality variants”
    Quote in the paperSubstantial claim

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

    Study design and conductown data

    The quasi-replication analyses were preregistered.

    “we performed a series of preregistered quasi-replication analyses (Supplementary Tables 11–12).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    Follow-up bioinformatic analyses suggest that implicated genes have early neurodevelopmental effects, which are then associated with behavioral patterns with repercussions across the lifespan.

    “Follow-up bioinformatic analyses suggest the implicated genes have early neurodevelopmental effects, which are then associated with behavioral patterns that have repercussions across the lifespan.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning

    The results do not show that people are genetically determined to experience particular life outcomes or are innately antisocial.

    “Our results are not evidence that some people are genetically determined to experience certain life outcomes or are “innately” antisocial.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data

    All GWAS included individuals of European ancestry, without missing relevant covariates, successfully genotyped and passing standardized sample-level quality control, and unrelated unless linear mixed models were used.

    “All GWAS were conducted among individuals that (a) were of European ancestry, (b) were not missing any relevant covariates, (c) were successfully genotyped and passed standardized sample-level quality control (according to study-specific protocols13–16,18), and (d) were unrelated (unless a particular GWAS was conducted with linear mixed models).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Premisecitation

    A lack of knowledge about how individual genetic differences affect susceptibility to externalizing can heighten stigma and blame for these behaviors.

    “At the same time, failing to understand how individual genetic differences contribute to vulnerability to externalizing can increase stigma and blame for these behaviors60.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [60] Kvaale, E. P., Gottdiener, W. H. & Haslam, N. Biogenetic explanations and stigma: a meta-analytic review of associations among laypeople. Soc. Sci. Med. 96, 95–103 (2013). source

  • Classified as probably not a substantial claim.

    Study design and conductreasoning

    Antisocial behavior was preregistered to be held out from the externalizing GWAS to enable quasi-replication with a central externalizing trait not included in the model.

    “We had preregistered to hold out antisocial behavior from the externalizing GWAS to enable quasi-replication with a central externalizing trait that was not included in the model.”
    Quote in the paperSubstantial claim
63 of 63

  • The paper relies on this without support.

    Behaviors related to self-regulation, such as smoking, drinking, drug use, condomless sex, and overeating, causally contribute to many of the medical outcomes associated with the EXT polygenic score.

    “Behaviors related to self-regulation, e.g., smoking, drinking, drug use, condomless sex, and overeating, contribute to many of these medical outcomes.”
1 of 1

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.

  • S-PrediXcan significance thresholds differ

    Minor

    The results give a significance threshold of 2.73×10−7 for S-PrediXcan, while the methods give 2.77×10−7 for the same analysis. The methods' threshold is consistent with the stated Bonferroni calculation.

  • Two different S-PrediXcan significance thresholds are reported

    Minor

    The results and methods give different Bonferroni significance thresholds for the S-PrediXcan analysis: 2.73×10−7 and 2.77×10−7.

2 of 2

Cites

Flagged first, then by citations

10 of the 109 references couldn’t be matched to a known paper.

Research works printed but not linked: 2

  1. Fox, D. Subversive science. Penn State Law Rev. 124, 153–191 (2019).
  2. Ge, T., Chen, C.-Y., Ni, Y., Feng, Y.-C. A. & Smoller, J. W. Polygenic prediction via Bayesian regression and continuous shrinkage priors. Nat. Commun. 10, 1776 (2019). · DOI · Paper page

Cited by

Sorted by citations

Related

Authors
Karlsson Linnér, R., Mallard, T. T., Barr, P. B., Sanchez-Roige, S., Madole, J. W., Driver, M. N., Poore, H. E., de Vlaming, R., Grotzinger, A. D., Tielbeek, J. J., Johnson, E. C., Liu, M., Rosenthal, S. B., Ideker, T., Zhou, H., Kember, R. L., Pasman, J. A., Verweij, K. J. H., Liu, D. J., Vrieze, S., Collaborators, C., Porjesz, B., Hesselbrock, V., Foroud, T. M., Agrawal, A., Edenberg, H. J., Nurnberger Jr, J. I., Liu, Y., Kuperman, S., Kramer, J., Meyer, J. L., Kamarajan, C., Pandey, A. K., Bierut, L., Rice, J., Bucholz, K. K., Schuckit, M. A., Tischfield, J., Brooks, A., Hart, R. P., Almasy, L., Dick, D. M., Salvatore, J. E., Goate, A., Kapoor, M., Slesinger, P., Scott, D. M., Bauer, L., Wetherill, L., Xuei, X., Lai, D., O’Connor, S. J., Plawecki, M. H., Lourens, S., Acion, L., Chan, G., Chorlian, D. B., Zhang, J., Kinreich, S., Pandey, G., Chao, M. J., Anokhin, A. P., McCutcheon, V. V., Saccone, S., Aliev, F., Barr, P. B., Chin, H., Parsian, A., Kranzler, H. R., Gelernter, J., Harris, K. M., Tucker-Drob, E. M., Waldman, I. D., Palmer, A. A., Harden, K. P., Koellinger, P. D., & Dick, D. M.
Affiliations
  • The University of Texas at Austin (US)
  • Virginia Commonwealth University (US)
  • Emory University (US)
  • Vrije Universiteit Amsterdam (NL)
  • Washington University in St. Louis (US)
  • University of Minnesota (US)
  • University of California San Diego (US)
  • Yale University (US)
  • Philadelphia VA Medical Center (US)
  • Radboud University Nijmegen (NL)
  • Amsterdam University Medical Centers (NL)
  • Pennsylvania State University (US)
  • University of Connecticut (US)
  • Indiana University Indianapolis (US)
  • University of Iowa (US)
  • SUNY Downstate Health Sciences University (US)
  • Rutgers, The State University of New Jersey (US)
  • Children's Hospital of Philadelphia (US)
  • Icahn School of Medicine at Mount Sinai (US)
  • Howard University (US)
  • National Institute on Alcohol Abuse and Alcoholism (US)
  • University of North Carolina at Chapel Hill (US)
  • University of Wisconsin–Madison (US)
Corresponding author
Koellinger, Philipp D.
ORCID
Venue
Nature Neuroscience
Publisher
Springer Science and Business Media LLC
Published
2021-08-26
Volume
24
Issue
10
Pages
1367-1376
ISSN
  • 1097-6256 (ISSN-L)
  • 1546-1726
DOI
10.1038/s41593-021-00908-3
PMC
PMC8484054
OpenAlex
W3193815771
Cited by
317
References
109
Citations by year
2026: 52 · 2025: 73 · 2024: 70 · 2023: 71 · 2022: 38 · 2021: 13
FWCI
27.09
Type
journal article
Language
en
Update notices
correction (2024-12-06)
Open access
Licence
Locations
  • Nature Neuroscience (publishedVersion)
  • PubMed (publishedVersion)
  • VU Research Portal (publishedVersion, other-oa)
  • Pure Amsterdam UMC (acceptedVersion, cc-by)
  • PubMed Central (acceptedVersion)
  • ScholarWorks @ UTRGV (The University of Texas Rio Grande Valley) (submittedVersion)
  • eScholarship (California Digital Library) (submittedVersion)
  • Data Archiving and Networked Services (DANS) (submittedVersion, other-oa)
  • SUNY Digital Repository Support (State University of New York System) (submittedVersion, cc-by-nc-nd)
  • Radboud Repository (Radboud University) (submittedVersion)
Funders
  • Tobacco-Related Disease Research Program (#28IR-0070, T29KT0526, K02AA018755)
  • Children's Hospital of Philadelphia
  • Vanderbilt University (UL1 RR024975, UL1‐TR000445, U01HG006378)
  • University of Pennsylvania
  • State University of New York
  • Washington University in St. Louis
  • University of Connecticut
  • Virginia Commonwealth University (K02AA018755)
  • National Alliance for Research on Schizophrenia and Depression (27676)
  • Vanderbilt University Medical Center (U19HL065962, U01HG006378, RC2GM092618, UL1TR000445, S10RR025141, UL1 RR024975, P50GM115305, UL1 TR002243, R01NS032830, R01HD074711, U01HG004798)
  • University of Texas Rio Grande Valley
  • Georgia Clinical and Translational Science Alliance (UL1 RR024975, U19HL065962, UL1TR002243, UL1TR000445, RC2GM092618, U01HG006378, P50GM115305, R01NS032830, U01HG004798, R01HD074711)
  • Jacobs Foundation
  • National Institutes of Health (T29KT0526, UL1TR000445, UL1TR002243)
  • University of North Carolina at Chapel Hill
  • University of California, San Diego
  • Perelman School of Medicine, University of Pennsylvania
  • SUNY Downstate Medical Center
  • National Institute on Alcohol Abuse and Alcoholism (K02AA018755, R01AA026281, R01AA015416, P50AA022537, U10AA008401)
  • National Institute on Aging (U24AG021886)
  • National Heart, Lung, and Blood Institute (U19HL065962)
  • National Institute of Mental Health (RC2MH089983, R01MH120219, RC2MH089924)
  • National Institute on Drug Abuse (28IR-0070, R01DA044283, P50DA037844, R01DA050721)
  • National Human Genome Research Institute (U01HG006378, U01HG004798)
  • National Institute of General Medical Sciences (P50GM115305, RC2GM092618)
  • National Institute of Neurological Disorders and Stroke (R01NS032830)
  • National Institute of Child Health and Human Development
  • National Center for Research Resources (S10RR025141, UL1RR024975)
  • National Center for Advancing Translational Sciences
  • Eunice Kennedy Shriver National Institute of Child Health and Human Development (R01HD074711, R01HD092548, R01HD083613, P01HD031921, R01HD073342, R01HD060726, P2CHD042849)
  • H2020 European Research Council (647648)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Associations and Epidemiology
  • Life Sciences › Neuroscience › Cellular and Molecular Neuroscience › Neurotransmitter Receptor Influence on Behavior
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic Mapping and Diversity in Plants and Animals
Keywords
self-regulation, externalizing traits, addiction, substance use, antisocial behavior, attention deficit hyperactivity disorder, genetic correlation, polygenic risk score, genetic loci, brain development, opioid use disorder, suicide, HIV infection, criminal conviction, unemployment, child behavior problems, shared genetic risk
MeSH terms
Self-Control, Attention Deficit Disorder with Hyperactivity, Behavioral Symptoms, Crime, Humans, Opioid-Related Disorders, Suicide, Unemployment, Meta-Analysis as Topic, Reproducibility of Results, HIV Infections, Multivariate Analysis, Behavior, Addictive, Computational Biology, Multifactorial Inheritance, Genome-Wide Association Study, Genetic Association Studies

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 under definition version 3 (extract-v3) on 2026-09-28.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-09-28.

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