Dimensional and transdiagnostic phenotypes in psychiatric genome-wide association studies.

Genome-wide association studies (GWAS) provide biological insights into disease onset and progression and have potential to produce clinically useful biomarkers. A growing body of GWAS focuses on quantitative and transdiagnostic phenotypic targets, such as symptom severity or biological markers, to enhance gene discovery and the translational utility of genetic findings. The current review discusses such phenotypic approaches in GWAS across major psychiatric disorders. We identify themes and recommendations that emerge from the literature to date, including issues of sample size, reliability, convergent validity, sources of phenotypic information, phenotypes based on biological and behavioral markers such as neuroimaging and chronotype, and longitudinal phenotypes. We also discuss insights from multi-trait methods such as genomic structural equation modelling. These provide insight into how hierarchical 'splitting' and 'lumping' approaches can be applied to both diagnostic and dimensional phenotypes to model clinical heterogeneity and comorbidity. Overall, dimensional and transdiagnostic phenotypes have enhanced gene discovery in many psychiatric conditions and promises to yield fruitful GWAS targets in the years to come.

Guidelines for Evaluating the Comparability of Down-Sampled GWAS Summary Statistics.

Proprietary genetic datasets are valuable for boosting the statistical power of genome-wide association studies (GWASs), but their use can restrict investigators from publicly sharing the resulting summary statistics. Although researchers can resort to sharing down-sampled versions that exclude restricted data, down-sampling reduces power and might change the genetic etiology of the phenotype being studied. These problems are further complicated when using multivariate GWAS methods, such as genomic structural equation modeling (Genomic SEM), that model genetic correlations across multiple traits. Here, we propose a systematic approach to assess the comparability of GWAS summary statistics that include versus exclude restricted data. Illustrating this approach with a multivariate GWAS of an externalizing factor, we assessed the impact of down-sampling on (1) the strength of the genetic signal in univariate GWASs, (2) the factor loadings and model fit in multivariate Genomic SEM, (3) the strength of the genetic signal at the factor level, (4) insights from gene-property analyses, (5) the pattern of genetic correlations with other traits, and (6) polygenic score analyses in independent samples. For the externalizing GWAS, although down-sampling resulted in a loss of genetic signal and fewer genome-wide significant loci; the factor loadings and model fit, gene-property analyses, genetic correlations, and polygenic score analyses were found robust. Given the importance of data sharing for the advancement of open science, we recommend that investigators who generate and share down-sampled summary statistics report these analyses as accompanying documentation to support other researchers' use of the summary statistics.

A multivariate approach to understanding the genetic overlap between externalizing phenotypes and substance use disorders.

Substance use disorders (SUDs) are phenotypically and genetically correlated with each other and with other psychological traits characterized by behavioural under-control, termed externalizing phenotypes. In this study, we used genomic structural equation modelling to explore the shared genetic architecture among six externalizing phenotypes and four SUDs used in two previous multivariate genome-wide association studies of an externalizing and an addiction risk factor, respectively. We first evaluated five confirmatory factor analytic models, including a common factor model, alternative parameterizations of two-factor structures and a bifactor model. We next explored the genetic correlations between factors identified in these models and other relevant psychological traits. Finally, we quantified the degree of polygenic overlap between externalizing and addiction risk using MiXeR. We found that the common and two-factor structures provided the best fit to the data, evidenced by high factor loadings, good factor reliability and no evidence of concerning model characteristics. The two-factor models yielded high genetic correlations between factors (rg s ≥ 0.87), and between the effect sizes of genetic correlations with external traits (rg  ≥ 0.95). Nevertheless, 21 of the 84 correlations with external criteria showed small, significant differences between externalizing and addiction risk factors. MiXer results showed that approximately 81% of influential externalizing variants were shared with addiction risk, whereas addiction risk shared 56% of its influential variants with externalizing. These results suggest that externalizing and addiction genetic risk are largely shared, though both constructs also retain meaningful unshared genetic variance. These results can inform future efforts to identify specific genetic influences on externalizing and SUDs.

The Hierarchical Taxonomy of Psychopathology (HiTOP) in psychiatric practice and research.

The Hierarchical Taxonomy of Psychopathology (HiTOP) has emerged out of the quantitative approach to psychiatric nosology. This approach identifies psychopathology constructs based on patterns of co-variation among signs and symptoms. The initial HiTOP model, which was published in 2017, is based on a large literature that spans decades of research. HiTOP is a living model that undergoes revision as new data become available. Here we discuss advantages and practical considerations of using this system in psychiatric practice and research. We especially highlight limitations of HiTOP and ongoing efforts to address them. We describe differences and similarities between HiTOP and existing diagnostic systems. Next, we review the types of evidence that informed development of HiTOP, including populations in which it has been studied and data on its validity. The paper also describes how HiTOP can facilitate research on genetic and environmental causes of psychopathology as well as the search for neurobiologic mechanisms and novel treatments. Furthermore, we consider implications for public health programs and prevention of mental disorders. We also review data on clinical utility and illustrate clinical application of HiTOP. Importantly, the model is based on measures and practices that are already used widely in clinical settings. HiTOP offers a way to organize and formalize these techniques. This model already can contribute to progress in psychiatry and complement traditional nosologies. Moreover, HiTOP seeks to facilitate research on linkages between phenotypes and biological processes, which may enable construction of a system that encompasses both biomarkers and precise clinical description.

External validation of a bifactor model of oppositional defiant disorder.

Dimensions of irritability and defiant behavior, though correlated within the structure of ODD, convey separable developmental risks through adolescence and adulthood. Irritability predicts depression and anxiety, whereas defiant behavior is a precursor to antisocial outcomes. Previously we demonstrated that a bifactor model comprising irritability and defiant behavior dimensions, in addition to a general factor, provided the best-fitting structure of ODD symptoms in five large datasets. Herein we extend our previous work by externally validating the bifactor model of ODD using multiple regression and multivariate behavior genetic analyses. We used parent ratings of DSM IV ODD symptoms, and symptom dimensions for ADHD (i.e., inattention and hyperactivity-impulsivity), conduct disorder (CD), depression/dysthymia, and generalized anxiety disorder (GAD) from 846 6-18-year-old twin pairs. We found that the ODD irritability factor was associated only with depression/dysthymia and GAD and the ODD defiant behavior factor was associated only with inattention, hyperactivity-impulsivity, and CD, whereas the ODD general factor was associated with all five symptom dimensions. Multivariate behavior genetic analyses found all five symptom dimensions shared genetic influences in common with the ODD general, irritability, and defiant behavior factors. In contrast, the defiant behavior factor shared genetic influences uniquely with inattention and hyperactivity-impulsivity, whereas the irritability factor shared genetic influences uniquely with depression/dysthymia and GAD, but not vice versa. This suggests that genes that influence irritability in early childhood also predispose to depression and anxiety in adolescence and adulthood. These multivariate genetic findings also support the external validity of the three ODD dimensions at the etiological level. Our study provides additional support for subtyping ODD based on these symptom dimensions, as in the revisions in the ICD-11, and suggests potential mechanisms underlying the development from ODD to behavioral or affective disorders.

The Hierarchical Taxonomy of Psychopathology (HiTOP): A Quantitative Nosology Based on Consensus of Evidence.

Traditional diagnostic systems went beyond empirical evidence on the structure of mental health. Consequently, these diagnoses do not depict psychopathology accurately, and their validity in research and utility in clinicalpractice are therefore limited. The Hierarchical Taxonomy of Psychopathology (HiTOP) consortium proposed a model based on structural evidence. It addresses problems of diagnostic heterogeneity, comorbidity, and unreliability. We review the HiTOP model, supporting evidence, and conceptualization of psychopathology in this hierarchical dimensional framework. The system is not yet comprehensive, and we describe the processes for improving and expanding it. We summarize data on the ability of HiTOP to predict and explain etiology (genetic, environmental, and neurobiological), risk factors, outcomes, and treatment response. We describe progress in the development of HiTOP-based measures and in clinical implementation of the system. Finally, we review outstanding challenges and the research agenda. HiTOP is of practical utility already, and its ongoing development will produce a transformative map of psychopathology.

Les progrès dans la réalisation de la classification quantitative de la psychopathologie.

Shortcomings of approaches to classifying psychopathology based on expert consensus have given rise to contemporary efforts to classify psychopathology quantitatively. In this paper, we review progress in achieving a quantitative and empirical classification of psychopathology. A substantial empirical literature indicates that psychopathology is generally more dimensional than categorical. When the discreteness versus continuity of psychopathology is treated as a research question, as opposed to being decided as a matter of tradition, the evidence clearly supports the hypothesis of continuity. In addition, a related body of literature shows how psychopathology dimensions can be arranged in a hierarchy, ranging from very broad "spectrum level" dimensions, to specific and narrow clusters of symptoms. In this way, a quantitative approach solves the "problem of comorbidity" by explicitly modeling patterns of co-occurrence among signs and symptoms within a detailed and variegated hierarchy of dimensional concepts with direct clinical utility. Indeed, extensive evidence pertaining to the dimensional and hierarchical structure of psychopathology has led to the formation of the Hierarchical Taxonomy of Psychopathology (HiTOP) Consortium. This is a group of 70 investigators working together to study empirical classification of psychopathology. In this paper, we describe the aims and current foci of the HiTOP Consortium. These aims pertain to continued research on the empirical organization of psychopathology; the connection between personality and psychopathology; the utility of empirically based psychopathology constructs in both research and the clinic; and the development of novel and comprehensive models and corresponding assessment instruments for psychopathology constructs derived from an empirical approach.

Correction to: The Association of Oxytocin Receptor Gene (OXTR) Polymorphisms with Antisocial Behavior: A Meta-analysis.

The original version of this article inadvertently omitted the word "with" between "Polymorphisms" and "Antisocial" from the title. The title "The Association of Oxytocin Receptor Gene (OXTR) Polymorphisms Antisocial Behavior: A Meta-Analysis" should be "The Association of Oxytocin Receptor Gene (OXTR) Polymorphisms with Antisocial Behavior: A Meta-Analysis." as presented above.

The Association of Oxytocin Receptor Gene (OXTR) Polymorphisms with Antisocial Behavior: A Meta-analysis.

Evidence suggests that the Oxytocin Receptor Gene (OXTR) influences human social cognition and behavior. OXTR has been investigated in relation to antisocial behavior, but studies examining this association have produced varying results in terms of the magnitude and significance of the association as well as which SNPs are implicated. This meta-analysis, based on 15 samples in 12 studies with a total sample of 12,236 individuals, examined the overall effects and consistency of associations between eight SNPs in OXTR and antisocial behavior. Random effects models identified a significant association between rs237887 and antisocial behavior (r = 0.06, p = 0.002) based on six studies that included a total of 6278 individuals. Sensitivity analyses suggest that these results were robust to exclusion of any individual study and publication bias. Nevertheless, the high levels of heterogeneity and quality control concerns with the original publications lead us to interpret this one significant finding with caution. We conclude that the available literature does not rule out, nor strongly support, an effect of OXTR on antisocial behavior.

Prospective test of the developmental propensity model of antisocial behavior: from childhood and adolescence into early adulthood.

BACKGROUND: The developmental propensity model of antisocial behavior posits that several dispositional characteristics of children transact with the environment to influence the likelihood of learning antisocial behavior across development. Specifically, greater dispositional negative emotionality, greater daring, and lower prosociality-operationally, the inverse of callousness- and lower cognitive abilities are each predicted to increase risk for developing antisocial behavior. METHODS: Prospective tests of key predictions derived from the model were conducted in a high-risk sample of 499 twins who were assessed on dispositions at 10-17 years of age and assessed for antisocial personality disorder (APD) symptoms at 22-31 years of age. Predictions were tested separately for parent and youth informants on the dispositions using multiple regressions that adjusted for oversampling, nonresponse, and clustering within twin pairs, controlling demographic factors and time since the first assessment. RESULTS: Consistent with predictions, greater numbers of APD symptoms in adulthood were independently predicted over a 10-15 year span by higher youth ratings on negative emotionality and daring and lower youth ratings on prosociality, and by parent ratings of greater negative emotionality and lower prosociality. A measure of working memory did not predict APD symptoms. CONCLUSIONS: These findings support future research on the role of these dispositions in the development of antisocial behavior.

Genetic and Environmental Contributions to Associations between Infant Fussy Temperament and Antisocial Behavior in Childhood and Adolescence.

Previous research suggests that fussy temperament in infancy predicts risk for later antisocial behavior (ASB) in childhood and adolescence. It remains unclear, however, to what extent infant fussiness is related to later ASB through causal processes or if they both reflect the same family risk factors for ASB. The current study used two approaches, the comparison of siblings and bivariate biometric modeling, to reduce familial confounding and examine genetic and environmental influences on associations between fussiness in the first 2 years of life and ASB in childhood and late adolescence. Analyses were conducted on data from a prospective cohort (9237 at 4-9 years and 7034 at 14-17 years) who are the offspring of a nationally representative sample of US women. In the full sample, fussiness predicted both child and adolescent ASB to small but significant extents, controlling for a wide range of measured child and family-level covariates. When siblings who differed in their fussiness were compared, fussiness predicted ASB in childhood, but not ASB during adolescence. Furthermore, results from a bivariate Cholesky model suggested that even the association of fussiness with childhood ASB found when comparing siblings is attributable to familial factors. That is, although families with infants who are higher in fussiness also tend to have children and adolescents who engage in greater ASB, the hypothesis that infant fussiness has an environmentally mediated impact on the development of future ASB was not strongly supported.

Extending the 'cross-disorder' relevance of executive functions to dimensional neuropsychiatric traits in youth.

BACKGROUND: Evidence that different neuropsychiatric conditions share genetic liability has increased interest in phenotypes with 'cross-disorder' relevance, as they may contribute to revised models of psychopathology. Cognition is a promising construct for study; yet, evidence that the same cognitive functions are impaired across different forms of psychopathology comes primarily from separate studies of individual categorical diagnoses versus controls. Given growing support for dimensional models that cut across traditional diagnostic boundaries, we aimed to determine, within a single cohort, whether performance on measures of executive functions (EFs) predicted dimensions of different psychopathological conditions known to share genetic liability. METHODS: Data are from 393 participants, ages 8-17, consecutively enrolled in the Longitudinal Study of Genetic Influences on Cognition (LOGIC). This project is conducting deep phenotyping and genomic analyses in youth referred for neuropsychiatric evaluation. Using structural equation modeling, we examined whether EFs predicted variation in core dimensions of the autism spectrum disorder, bipolar illness, and schizophrenia (including social responsiveness, mania/emotion regulation, and positive symptoms of psychosis, respectively). RESULTS: We modeled three cognitive factors (working memory, shifting, and executive processing speed) that loaded on a second-order EF factor. The EF factor predicted variation in our three target traits, but not in a negative control (somatization). Moreover, this EF factor was primarily associated with the overlapping (rather than unique) variance across the three outcome measures, suggesting that it related to a general increase in psychopathology symptoms across those dimensions. CONCLUSIONS: Findings extend support for the relevance of cognition to neuropsychiatric conditions that share underlying genetic risk. They suggest that higher-order cognition, including EFs, relates to the dimensional spectrum of each of these disorders and not just the clinical diagnoses. Moreover, results have implications for bottom-up models linking genes, cognition, and a general psychopathology liability.

Candidate genes for aggression and antisocial behavior: a meta-analysis of association studies of the 5HTTLPR and MAOA-uVNTR.

Variation in central serotonin levels due to genetic mutations or experimental modifications has been associated with the manifestation of aggression in humans and animals. Many studies have examined whether common variants in serotonergic genes are implicated in aggressive or antisocial behaviors (ASB) in human samples. The two most commonly studied polymorphisms have been the serotonin transporter linked polymorphic region of the serotonin transporter gene (5HTTLPR) and the 30 base pair variable number of tandem repeats of the monoamine oxidase A gene (MAOA-uVNTR). Despite the aforementioned theoretical justification for these polymorphisms, findings across studies have been mixed and are thus difficult to interpret. A meta-analysis of associations of the 5HTTLPR and MAOA-uVNTR with ASB was conducted to determine: (1) the overall magnitude of effects for each polymorphism, (2) the extent of heterogeneity in effect sizes across studies and the likelihood of publication bias, and (3) whether sample-level or study-level characteristics could explain observed heterogeneity across studies. Both the 5HTTLPR and the MAOA-uVNTR were significantly associated with ASB across studies. There was also significant and substantial heterogeneity in the effect sizes for both markers, but this heterogeneity was not explained by any sample-level or study-level characteristics examined. We did not find any evidence for publication bias across studies for the MAOA-uVNTR, but there was evidence for an oversampling of statistically significant effect sizes for the 5HTTLPR. These findings provide support for the modest role of common serotonergic variants in ASB. Implications regarding the role of serotonin in antisocial behavior and the conceptualization of antisocial and aggressive phenotypes are discussed.

A sibling-comparison study of smoking during pregnancy and childhood psychological traits.

Prenatal exposure to substances of abuse is associated with numerous psychological problems in offspring, but quasi-experimental studies controlling for co-occurring risk factors suggest that familial factors (e.g., genetic and environmental effects shared among siblings) confound many associations with maternal smoking during pregnancy (SDP). Few of the quasi-experimental studies in this area have explored normative psychological traits in early childhood or developmental changes across the lifespan, however. The current study used multilevel growth curve models with a large, nationally-representative sample in the United States to investigate for potential effects of SDP on the developmental trajectories of cognitive functioning, temperament/personality, and disruptive behavior across childhood, while accounting for shared familial confounds by comparing differentially exposed siblings and statistically controlling for offspring-specific covariates. Maternal SDP predicted the intercept (but not change over time) for all cognitive and externalizing outcomes. Accounting for familial confounds, however, attenuated the association between SDP exposure and all outcomes, except the intercept (age 5) for reading recognition. These findings, which are commensurate with previous quasi-experimental research on more severe indices of adolescent and adult problems, suggest that the associations between SDP and developmental traits in childhood are due primarily to confounding factors and not a causal association.

Assessing structural models of neighborhood and family sociodemographic characteristics and their relations with externalizing psychopathology.

Externalizing psychopathology has been found to have small to moderate associations with neighborhood and family sociodemographic characteristics. However, prior studies may have used suboptimal operationalizations of neighborhood sociodemographic characteristics and externalizing psychopathology, potentially misestimating relations between these constructs. To address these limitations, in the current study we test different measurement models of these constructs and assess the structural relations between them. Using a population-representative sample of 2,195 twins and siblings from the Georgia Twin Study and data from the National Neighborhood Data Archive and 2000 U.S. Census, we assessed the fit of competing measurement models for family sociodemographic, neighborhood sociodemographic, and neighborhood environment characteristics. In structural models, we regressed a general externalizing dimension on different operationalizations of these variables separately and then simultaneously in a final model. Latent variable operationalizations of family sociodemographic, neighborhood sociodemographic, and neighborhood environment characteristics explained no more variance in broad externalizing psychopathology than other operationalizations. In an omnibus model, family sociodemographic characteristics showed a small association with externalizing psychopathology, while neighborhood sociodemographic and environmental characteristics did not. Family sociodemographic characteristics showed small associations with neighborhood sociodemographic and environmental characteristics, and neighborhood sociodemographic characteristics were moderately associated with neighborhood environment. These findings suggest that family sociodemographic characteristics are more associated with the development of broad externalizing psychopathology in youth than neighborhood sociodemographic characteristics and neighborhood environment. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Principles and procedures for revising the hierarchical taxonomy of psychopathology.

Quantitative, empirical approaches to establishing the structure of psychopathology hold promise to improve on traditional psychiatric classification systems. The Hierarchical Taxonomy of Psychopathology (HiTOP) is a framework that summarizes the substantial and growing body of quantitative evidence on the structure of psychopathology. To achieve its aims, HiTOP must incorporate emerging research in a systematic, ongoing fashion. In this article, we describe the historical context and grounding of the principles and procedures for revising the HiTOP framework. Informed by strengths and shortcomings of previous classification systems, the proposed revisions protocol is a formalized system focused around three pillars: (a) prioritizing systematic evaluation of quantitative evidence by a set of transparent criteria and processes, (b) balancing stability with flexibility, and (c) promoting inclusion over gatekeeping in all aspects of the process. We detail how the revisions protocol will be applied in practice, including the scientific and administrative aspects of the process. Additionally, we describe areas of the HiTOP structure that will be a focus of early revisions and outline challenges for the revisions protocol moving forward. The proposed revisions protocol is designed to ensure that the HiTOP framework reflects the current state of scientific knowledge on the structure of psychopathology and fulfils its potential to advance clinical research and practice. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

The double-edged sword of grandiose narcissism: implications for successful and unsuccessful leadership among U.S. Presidents.

Recent research and theorizing suggest that narcissism may predict both positive and negative leadership behaviors. We tested this hypothesis with data on the 42 U.S. presidents up to and including George W. Bush, using (a) expert-derived narcissism estimates, (b) independent historical surveys of presidential performance, and (c) largely or entirely objective indicators of presidential performance. Grandiose, but not vulnerable, narcissism was associated with superior overall greatness in an aggregate poll; it was also positively associated with public persuasiveness, crisis management, agenda setting, and allied behaviors, and with several objective indicators of performance, such as winning the popular vote and initiating legislation. Nevertheless, grandiose narcissism was also associated with several negative outcomes, including congressional impeachment resolutions and unethical behaviors. We found that presidents exhibit elevated levels of grandiose narcissism compared with the general population, and that presidents' grandiose narcissism has been rising over time. Our findings suggest that grandiose narcissism may be a double-edged sword in the leadership domain.

Early concern and disregard for others as predictors of antisocial behavior.

BACKGROUND: Prediction of antisocial behavior is important, given its adverse impact on both the individuals engaging in antisocial behavior and society. Additional research identifying early predictors of future antisocial behavior, or antisocial propensity, is needed. The present study tested the hypothesis that both concern for others and active disregard for others in distress in toddlers and young children predict antisocial behavior during middle childhood and adolescence. METHODS: A representative sample of same-sex twins (N=956) recruited in Colorado was examined. Mother-rated and researcher-observed concern and disregard for others assessed at age 14-36 months were examined as predictors of parent- (age 4-12), teacher- (age 7-12), and self-reported (age 17) antisocial behavior. RESULTS: Observed disregard for others predicted antisocial behavior assessed by three different informants (parents, teachers, and self), including antisocial behavior assessed 14 years later. It also predicted a higher order antisocial behavior factor (ß=.58, p<.01) after controlling for observed concern for others. Mother-rated disregard for others predicted parent-reported antisocial behavior. Contrary to predictions, neither mother-rated nor observed concern for others inversely predicted antisocial behavior. RESULTS of twin analyses suggested that the covariation between observed disregard for others and antisocial behavior was due to shared environmental influences. CONCLUSIONS: Disregard for others in toddlerhood/early childhood is a strong predictor of antisocial behavior in middle childhood and adolescence. The results suggest the potential need for early assessment of disregard for others and the development of potential interventions.

Evaluating alternative models of youth externalizing using quantitative genetic analyses.

Interest has increased in the recent literature on characterizing psychopathology dimensionally in hierarchical models. One dimension of psychopathology that has received considerable attention is externalizing. Although extensively studied and well-characterized in late adolescents and adults, delineation of the externalizing spectrum in youth has lagged behind. As a complement to structural analyses of externalizing, in this study, we use quantitative genetic analyses of twin data to adjudicate among alternative models of youth externalizing that differ in granularity. Specifically, we compared model fit, estimates of genetic and environmental influences on the externalizing dimension, and the average, variability, and precision of genetic and environmental influences on individual symptoms due to the externalizing dimension, specific symptom dimensions, and unique etiological influences. Given that none of these criteria are definitive on their own, we looked to the confluence of these criteria to exclude particular models while highlighting others as leading contenders. We analyzed parent-report data on 38 externalizing symptoms from a population-representative, ethnically diverse sample of 883 youth twin pairs (51% female), who were on average 8.5 years old. Although models including an externalizing composite and attention deficit hyperactivity disorder, oppositional defiant disorder, and conduct disorder diagnoses and symptom dimensions showed similar heritability to latent variable models of externalizing, models that included latent dimensions of externalizing and more fine-grained symptom dimensions fit better and were more balanced in the magnitude of genetic and environmental influences on individual symptoms due to the externalizing dimension and specific symptom dimensions. Pending replication, these more granular and elaborated model(s) can be useful for advancing research on causes and outcomes of youth externalizing and its fine-grained specific components. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Guidelines for Evaluating the Comparability of Down-Sampled GWAS Summary Statistics.

Proprietary genetic datasets are valuable for boosting the statistical power of genome-wide association studies (GWASs), but their use can restrict investigators from publicly sharing the resulting summary statistics. Although researchers can resort to sharing down-sampled versions that exclude restricted data, down-sampling reduces power and might change the genetic etiology of the phenotype being studied. These problems are further complicated when using multivariate GWAS methods, such as genomic structural equation modeling (Genomic SEM), that model genetic correlations across multiple traits. Here, we propose a systematic approach to assess the comparability of GWAS summary statistics that include versus exclude restricted data. Illustrating this approach with a multivariate GWAS of an externalizing factor, we assessed the impact of down-sampling on (1) the strength of the genetic signal in univariate GWASs, (2) the factor loadings and model fit in multivariate Genomic SEM, (3) the strength of the genetic signal at the factor level, (4) insights from gene-property analyses, (5) the pattern of genetic correlations with other traits, and (6) polygenic score analyses in independent samples. For the externalizing GWAS, down-sampling resulted in a loss of genetic signal and fewer genome-wide significant loci, while the factor loadings and model fit, gene-property analyses, genetic correlations, and polygenic score analyses are robust. Given the importance of data sharing for the advancement of open science, we recommend that investigators who share down-sampled summary statistics report these analyses as accompanying documentation to support other researchers' use of the summary statistics.