If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Social, Genetic & Developmental Psychiatry Centre, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, UKNational Institute for Health and Care Research (NIHR) Maudsley Biomedical Research Centre at South London and Maudsley NHS Foundation Trust, London, UK
Social, Genetic & Developmental Psychiatry Centre, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, UKNational Institute for Health and Care Research (NIHR) Maudsley Biomedical Research Centre at South London and Maudsley NHS Foundation Trust, London, UKNational Centre for Register-based Research, Aarhus Business and Social Sciences, Aarhus University, Aarhus, Denmark
Social, Genetic & Developmental Psychiatry Centre, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, UKNational Institute for Health and Care Research (NIHR) Maudsley Biomedical Research Centre at South London and Maudsley NHS Foundation Trust, London, UK
Department of Psychiatry and School of Public Health, University of California San Diego, La Jolla, CA, USAVA San Diego Healthcare System, San DIego, CA, USA
Social, Genetic & Developmental Psychiatry Centre, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, UKNational Institute for Health and Care Research (NIHR) Maudsley Biomedical Research Centre at South London and Maudsley NHS Foundation Trust, London, UK
Correspondence: Gerome Breen, Social, Genetic and Developmental Psychiatry Centre, Institute of Psychiatry, Psychology and Neuroscience, King's College London, London, SE5 8AF;
Social, Genetic & Developmental Psychiatry Centre, Institute of Psychiatry, Psychology & Neuroscience, King’s College London, UKNational Institute for Health and Care Research (NIHR) Maudsley Biomedical Research Centre at South London and Maudsley NHS Foundation Trust, London, UK
Decades of research have shown that environmental exposures, including self-reports of trauma, are partly heritable. Heritable characteristics may influence exposure to and interpretations of environmental factors. Identifying heritable factors associated with self-reported trauma could improve our understanding of vulnerability to exposure and the interpretation of life events.
Methods
We used genome-wide association study summary statistics of childhood maltreatment, defined as reporting of abuse (emotional, sexual and physical) and neglect (emotional and physical) (n=185,414). We calculated genetic correlations (rg) between reported childhood maltreatment and 576 traits to identify phenotypes that might explain the heritability of reported childhood maltreatment, retaining those with |rg|>0.25. We specified multiple regression models using genomic structural equation modelling to detect residual genetic variance in childhood maltreatment after accounting for genetically correlated traits.
Results
In two separate models, the shared genetic component of twelve health and behavioural traits and seven psychiatric disorders accounted for 59% and 56% of heritability due to common genetic variants (h2SNP) of childhood maltreatment, respectively. Genetic influences on the h2SNP of childhood maltreatment were generally accounted for by a shared genetic component across traits. The exceptions to this were general risk tolerance, subjective well-being, post-traumatic stress disorder and autism spectrum disorder, identified as independent contributors to its h2SNP. These four traits alone were sufficient to explain 58% of the h2SNP of childhood maltreatment.
Conclusions
We identified putative traits that reflect the h2SNP of childhood maltreatment. Elucidating the mechanisms underlying these associations may improve trauma prevention and posttraumatic intervention strategies.
Traumatic events, namely those perceived as physically or emotionally threatening and violating, are associated with various adverse outcomes, including psychopathology (
Turner RJ, Jay Turner R, Lloyd DA (1995): Lifetime Traumas and Mental Health: The Significance of Cumulative Adversity. Journal of Health and Social Behavior, vol. 36. p 360.
). Decades of behavioural genetics research has shown that reported trauma exposures, like many environmental measures and behavioural traits, are partly heritable (
). Interpersonal assaultive traumas (e.g., physical and sexual assault) have higher heritability than non-interpersonal or non-assaultive traumas (e.g., accidents) (
). In relation to these observations, stressful life events (SLEs) dependent on one’s behaviour (e.g., fights) are more heritable than those that are independent (e.g., natural disasters), with the latter occurring more often due to chance (
). However, being at higher genetic risk for reported trauma does not signify that an individual is genetically predestined to experience trauma. Furthermore, a large proportion of the total phenotypic variability of reported trauma is not attributable to genetics. The environment itself may be harmful, or a perpetrator may exploit those in vulnerable circumstances (
). Exploring traits genetically related to reported trauma in different environmental contexts may provide a framework for social research to help determine trauma risk factors and protect vulnerable individuals (
Heritable behavioural characteristics may contribute to the likelihood of experiencing certain events. Personality traits, such as sensation seeking, openness to experience, and antisocial behaviour, are phenotypically and genetically correlated with reporting interpersonal assaultive trauma (
). Such partially heritable characteristics may contribute to the heritability of reported trauma through gene-environment correlation (rGE), whereby the environment reflects an individual’s genetic propensities via three different processes (
). Passive rGE occurs when a relative’s genotype, such as parental genetic variation contributing to risk-taking behaviours, shapes the child’s environment and potentially creates an unsafe home (
). The environment that the parent creates and the parental genotype are correlated as the child receives both from their biological parents. Thus, parental environmental effects may be captured in genetic analyses of offspring traits (
). Evocative rGE arises when an individual’s genotype shapes how others engage with them. For example, a child’s behavioural difficulties may evoke verbal and physical discipline due to the carer’s expectations of how a child should behave (
), leading to differing risks of exposure to adverse environments.
Correlations between genetic factors and retrospective reports of trauma may also, in part, be driven by heritable characteristics influencing the subjective interpretation, willingness to disclose and recollection of events (
). Genetic research has largely relied on retrospective self-reports of trauma exposure which may be more susceptible to genetically influenced perceptions and recollection of events, as opposed to more objective measures prospectively recorded closer to the time of exposure (e.g., court records, caregiver reports) (
). Subjective appraisal of trauma is important for posttraumatic psychopathology, which is more strongly associated with retrospective self-reports of trauma than objective court records (
). Individual, partially heritable differences in personality traits such as neuroticism and agreeableness may explain the discrepancy between retrospective and prospective measures of trauma (
Baldwin JR, Reuben A, Newbury JB, Danese A (2019): Agreement Between Prospective and Retrospective Measures of Childhood Maltreatment. JAMA Psychiatry, vol. 76. p 584.
). Furthermore, the consistency and frequency of self-reports are impacted by individual factors involved in the willingness to disclose a traumatic event, such as perception of stigma, fear of negative consequences, or pre-existing relationships with the perpetrator (
). Thus, a better understanding of the heritable factors that impact the retrospective report of trauma experiences could help improve posttraumatic support.
In sum, the influences on retrospectively reported trauma are complex and difficult to disentangle. A range of heritable traits may be involved. Heritability and genetic correlations between traits can be estimated using genome-wide association study (GWAS) summary statistics (
van Rheenen W, Peyrot WJ, Schork AJ, Lee SH, Wray NR (2019): Genetic correlations of polygenic disease traits: from theory to practice. Nat Rev Genet.. https://doi.org/10.1038/s41576-019-0137-z
Coleman JRI, Peyrot WJ, Purves KL, Davis KAS, Rayner C, Choi SW, et al. (2020): Genome-wide gene-environment analyses of major depressive disorder and reported lifetime traumatic experiences in UK Biobank. Mol Psychiatry. https://doi.org/10.1038/s41380-019-0546-6
Coleman JRI, Peyrot WJ, Purves KL, Davis KAS, Rayner C, Choi SW, et al. (2020): Genome-wide gene-environment analyses of major depressive disorder and reported lifetime traumatic experiences in UK Biobank. Mol Psychiatry. https://doi.org/10.1038/s41380-019-0546-6
). However, these studies did not analytically explain the extent to which the h2SNP of reported traumas reflects genetic correlations with these complex traits. Identifying specific traits that explain a large proportion of h2SNP can guide follow-up analyses in assessing certain characteristics involved in rGE and/or the subjective experience of trauma.
We hypothesised that genetic variants associated with relevant behavioural and cognitive traits would overlap with those associated with reported trauma, such that no residual genetic variance of reported trauma would remain after accounting for genetically correlated traits. Twin studies have used multivariate structural equation modelling (SEM) to examine the residual genetic variance of life event measures after accounting for genetically correlated traits (
). To our knowledge, multivariate SEM has not been used to explore the extent to which specific heritable characteristics capture the heritability of reported trauma. Twin studies are limited in assessing only a moderate number of traits and environmental measures in the same individuals, which may be particularly challenging in the case of more severe environmental exposures such as trauma (
) allows the inclusion of many more traits measured in different individuals. Here, we decompose the h2SNP of reported trauma using genomic multiple regression with the Genomic SEM R package (
). Our primary aim was to measure the amount of residual genetic variance of reported trauma that remains after accounting for genetically correlated traits. Our secondary aim was to identify the traits contributing to the h2SNP of reported trauma from hundreds of complex traits that were systematically assessed.
Methods and Materials
Samples and measures
We used summary statistics from the largest published GWAS of reported trauma as of 2021, on childhood maltreatment (
). This GWAS built on our previous work (Coleman et al., 2020), extending it to assess childhood maltreatment specifically, and included 185,414 participants predominantly of European ancestry from five datasets: the UK Biobank (
Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, et al. (2015): UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med 12: e1001779.
Fraser A, Macdonald-Wallis C, Tilling K, Boyd A, Golding J, Davey Smith G, et al. (2013): Cohort Profile: the Avon Longitudinal Study of Parents and Children: ALSPAC mothers cohort. Int J Epidemiol 42: 97–110.
Boyd A, Golding J, Macleod J, Lawlor DA, Fraser A, Henderson J, et al. (2013): Cohort profile: the “children of the 90s”—the index offspring of the Avon Longitudinal Study of Parents and Children. Int J Epidemiol 42: 111–127.
Kooijman MN, Kruithof CJ, van Duijn CM, Duijts L, Franco OH, van IJzendoorn MH, et al. (2016): The Generation R Study: design and cohort update 2017. Eur J Epidemiol 31: 1243–1264.
). Childhood maltreatment was defined as reports of emotional, sexual, and physical abuse, and emotional and physical neglect. Most traumas (91.5%) were retrospectively self-reported (n = 169,766); however, a small proportion (8.5%) were reported prospectively by a parent or caregiver (n = 15,651). The genetic correlation between retrospective and prospective childhood maltreatment was previously reported as 0.72 (s.e = 0.36; P= 0.05) (
). In the original publication, the h2SNP of the continuous meta-analysed phenotype of childhood maltreatment was 0.08 (s.e = 0.01) using linkage disequilibrium score (LDSC) regression (
). We also analysed GWAS summary statistics from Coleman et al., 2020 of a retrospectively reported lifetime trauma phenotype that more broadly captures trauma occurring in both childhood and adulthood in the UK Biobank (
Coleman JRI, Peyrot WJ, Purves KL, Davis KAS, Rayner C, Choi SW, et al. (2020): Genome-wide gene-environment analyses of major depressive disorder and reported lifetime traumatic experiences in UK Biobank. Mol Psychiatry. https://doi.org/10.1038/s41380-019-0546-6
Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, et al. (2015): UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med 12: e1001779.
Bulik-Sullivan BK, Loh P-R, Finucane HK, Ripke S, Yang J, Schizophrenia Working Group of the Psychiatric Genomics Consortium, et al. (2015): LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet 47: 291–295.
) to measure the genetic correlations (rg) between reported trauma and a wide range of complex traits. We tested 576 traits from GWAS summary statistics for an rg with reported trauma. We excluded the major histocompatibility complex (MHC) region from our analyses (
LD Hub: a centralized database and web interface to perform LD score regression that maximizes the potential of summary level GWAS data for SNP heritability and genetic correlation analysis.
). We considered traits for downstream analyses in Genomic SEM if they met the following criteria: a |rg| with reported trauma of > 0.25 and |Z| statistic ≥ 5, a GWAS mean 2 value >1.02 and an h2SNP Z statistic ≥ 5. These thresholds were based on recommendations by the software developers (
Nivard MG (2019, February 5): Minimum sample size, SNP-h2 and rG? Genomic SEM users google group. Retrieved January 14, 2022, from https://groups.google.com/g/genomic-sem-users/c/Wn1jptU2VcY/m/9CWxfQbvAAAJ
). All traits that met these criteria were also statistically significant after Bonferroni correction for multiple testing ( = 0.05/ the number of traits; P ≤ 8.68 × 10−5), which was less stringent than our selected threshold rg |Z| statistic ≥ 5 (equivalent to P ≤ 5.73 × 10−7). The criteria were stringent to restrict the number of traits included in downstream analyses to well-powered GWAS with potentially larger genetic contributions to the h2SNP of reported trauma.
). Genomic SEM is a multivariable extension of LDSC that constructs covariance matrices from h2SNP and rg calculated by LDSC. GWAS samples can overlap for Genomic SEM as the sampling covariance matrices adjust for potential sample overlap. All GWAS summary statistics in our analyses were based on individuals drawn from European ancestries.
We fitted fully saturated genomic multiple regression models. This approach simultaneously regressed the outcome (i.e., reported childhood maltreatment) on various explanatory variables, which modelled genetic correlations between each explanatory variable. This was informative in two ways. First, we estimated the residual genetic variance of reported trauma not explained by the genetics of the explanatory variables. Second, we estimated the unique contribution of each explanatory variable to the genetic component of reported trauma independent of other explanatory variables (i.e. conditional genetic association, termed bg). We selected explanatory variables based on results from bivariate LDSC regression. We introduced one explanatory variable at a time, iteratively, from the most highly correlated trait with reported trauma to the least correlated trait. For all models, we used the default diagonally weighted least-squares estimator, in which the precision of genetic covariances (e.g. due to GWAS sample size) is considered.
We identified a practical limit of ≤11 explanatory variables in a fully saturated multiple regression model. Standard errors increased as more explanatory variables were added (>11 explanatory variables resulted in standard errors >1). Therefore, we had to fit two models for categorically distinct traits identified in LDSC regression analysis: model 1) health and behavioural traits and model 2) psychiatric disorders.
Results
Results from analysing reported childhood maltreatment were highly similar to that of reported lifetime trauma. We therefore present results for the more highly powered GWAS of childhood maltreatment (power quantified by h2SNP Z statistic = 18.7; mean 2 value = 1.27) and report our findings for reported lifetime trauma (h2SNP Z statistic = 15.6; mean 2 value = 1.22) in the Supplementary Results and Supplementary Tables 7-11.
Traits genetically correlated with reported childhood maltreatment
Figure 1 shows the 18 bivariate genetic correlations with childhood maltreatment (|rg| > 0.25 and rg |Z| statistic ≥ 5). Genetic correlations with all 576 traits are summarised in Supplementary Table 1. After filtering for sufficiently powered GWAS, 18 traits were genetically correlated with childhood maltreatment with |rg| > 0.25 and rg |Z| statistic ≥ 5 (Figure 1; Supplementary Table 2). The pairwise genetic correlations amongst the health and behavioural traits and psychiatric disorders modelled in Genomic SEM are shown in Supplementary Tables 3L and 4H, respectively. The most well-powered GWAS was retained in cases where pairwise |rg| in each category was not significantly different from one (calculated using the chi-squared distribution function and [(|rg|−1)/se]2 in R v. 4.1.1) (
Coleman JRI, Peyrot WJ, Purves KL, Davis KAS, Rayner C, Choi SW, et al. (2020): Genome-wide gene-environment analyses of major depressive disorder and reported lifetime traumatic experiences in UK Biobank. Mol Psychiatry. https://doi.org/10.1038/s41380-019-0546-6
Team RC (n.d.): R Core Team (2019) RA Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. References-Scientific Research Publishing.
Figure 1Top bivariate genetic correlations (rg) between reported childhood maltreatment and various heritable traits. We calculated the correlations using linkage disequilibrium score (LDSC) regression. We tested correlations with 576 traits. Only traits with an |rg| > 0.25 and rg |Z| statistic ≥ 5, sufficiently powered with a mean 2 value > 1.02 and a common genetic variant-based heritability Z statistic ≥ 5 are shown. Bars represent standard errors. ADHD = attention deficit hyperactivity disorder; PTSD = post-traumatic stress disorder.
As noted above, due to being able to include a maximum of 11 explanatory variables at once, we specified separate models for health and behavioural traits and psychiatric disorders. The path diagram in Figure 2A shows the results for eleven health and behavioural traits simultaneously specified as explanatory variables of the genetic component of childhood maltreatment. Heritability can be defined as the proportion of variance between individuals for a given trait or disorder that is accounted for by genetic factors. The residual genetic variance is the amount of heritable variation of childhood maltreatment that is unexplained by the genetic factors of the explanatory variables in the model. The genetic influences on these health and behavioural traits explained 59% of the h2SNP of childhood maltreatment (one minus the residual genetic variance of 0.41± 0.07; P = 7.76 × 10−9).
Figure 2Path diagrams representing results from genomic multiple regression analyses. Models were specified in Genomic SEM. The genetic component of A) eleven health and behavioural traits B) nine psychiatric disorders are the simultaneously fitted explanatory variables of the genetic component of reported childhood maltreatment using a weighted least-squares estimator. Single-headed arrows are conditional genetic associations (bg ± SE) between the explanatory variables and childhood maltreatment independent of the genetic influences of the other explanatory variables. A solid line indicates that the conditional genetic association is significant, and a dashed line indicates the conditional genetic association is non-significant. Double-headed arrows connecting explanatory variables represent genetic correlations; for simplicity, these values are not shown here but are in Supplementary Tables 3L and 4H. Double-headed arrows connecting the genetic component of childhood maltreatment to itself is the residual genetic variance of childhood maltreatment (umaltreatment ± SE) that is unexplained by the genetic influence of either the psychiatric disorders or the health and behavioural traits. ADHD = attention deficit hyperactivity disorder; ALCH = alcohol dependence; MDD = major depressive disorder; PTSD = post-traumatic stress disorder; SCZ = schizophrenia.
Once controlling for the genetic influences of the other explanatory variables, the majority of conditional genetic associations between reported childhood maltreatment and health and behavioural traits were non-significant (P > 0.05; dashed lines in Figure 2A & Supplementary Table 3K). Therefore, most of the genetic correlations between childhood maltreatment and health and behavioural traits were shared with the other health and behavioural traits in the model. The shared genetic component across these traits may drive the direction of some bivariate genetic correlations, as several traits showed conditional associations with childhood maltreatment in the opposite direction (e.g. reported health rg = -0.48; bg = 0.16). However, two traits were uniquely genetically associated with childhood maltreatment, over and above their genetic correlations with the other traits: subjective well-being (bg = -0.47 ± 0.18; P= 0.01) and general risk tolerance (the reporting of how comfortable you are with taking risks; bg = 0.31 ± 0.06; P = 9.96 × 10-7). Models where each health trait was introduced iteratively are shown in Supplementary Tables 3A-K.
Psychiatric disorders
When taking into account the genetic influences of all nine psychiatric disorders in a genomic multiple regression model (Figure 2B), the residual genetic variance of reported childhood maltreatment was 0.44 ± 0.08 (P= 9.23 × 10−9). Thus, 56% (calculated as 1 – 0.44) of the h2SNP of childhood maltreatment was explained by genetic components of these psychiatric disorders. In this model, most psychiatric disorders shared their genetic overlap with childhood maltreatment, as indicated by non-significant conditional associations after accounting for shared genetics (P > 0.05; Figure 2B and Supplementary Table 4G). Autism spectrum disorder (ASD) (bg = 0.25 ± 0.11, P= 0.02) and post-traumatic stress disorder (PTSD) symptoms (bg = 0.29 ± 0.09, P= 1.40 × 10−3) had significant independent associations with childhood maltreatment independent of other disorders. Supplementary Tables 4A-G show genomic multiple regression results when psychiatric disorders were included in iterative stages.
Model of independently contributing health and psychiatric traits
To explore the contribution of the independently genetically associated traits (i.e., ASD, PTSD, general risk tolerance, and subjective well-being), as a sensitivity analysis, we specified the genetic component of these four traits as explanatory variables of the genetic component of reported childhood maltreatment simultaneously (Supplementary Figure 1 & Supplementary Table 5). The residual genetic variance of reported childhood maltreatment was 0.42 ± 0.06 (P= 1.24 × 10−12). This means 58% of the h2SNP of childhood maltreatment was explained by the genetic component of ASD, PTSD, general risk tolerance, and subjective well-being. In a model with two additional traits (ADHD and self-reported tiredness), the residual genetic variance of childhood maltreatment only decreased by ∼2% (0.40 ± 0.06; P= 2.11 × 10−12; Supplementary Results & Supplementary Table 6).
Discussion
Using multivariate genomic SEM, we identified traits that accounted for ∼60% of the h2SNP of reported trauma. Health and behavioural traits together accounted for 59% of h2SNP variance, whilst the model exploring psychiatric disorders explained 56%. In both models, a shared genetic component was observed across traits except subjective well-being, general risk tolerance, PTSD symptoms and ASD, which were independently associated with childhood maltreatment. Together, these latter four traits alone were sufficient to explain 58% of the h2SNP of reported childhood maltreatment.
We found similar results for retrospective lifetime trauma, which included adult trauma. We could not directly compare adulthood and childhood trauma as the lifetime trauma phenotype included both measures. Our findings suggest that the exact timing of trauma does not strongly affect the proportion of genetic variance accounted for by health and psychiatric traits. However, replication in non-overlapping datasets and appropriate trauma measures are required to make strong conclusions about differences between adulthood and childhood trauma.
Independently associated traits, in addition to the genetic components shared across health and psychiatric traits, may be involved in rGE, and/or the reporting of such environments as traumatic. This raises the question of which processes explain these associations and could inform strategies to minimise risk and consequences of trauma.
General risk tolerance is measured by endorsing a willingness to take risks, broadly capturing risk-taking behaviours (
Genome-wide association analyses of risk tolerance and risky behaviors in over 1 million individuals identify hundreds of loci and shared genetic influences.
). rGE may explain the genetic contribution of risk tolerance to the genetic component of reported trauma, consistent with a previous study that found environmental adversities mediate the association between genetic propensity for risk-taking and reported childhood maltreatment (
). A child may passively inherit a parent’s genetic propensity for risk-taking behaviours, such as substance use, and be exposed to an environment where the child may be neglected or abused (
). Alternatively, risk tolerance may capture behaviours associated with disclosure of trauma. A barrier to disclosing trauma includes perceiving it as a risk and fearing negative consequences (
). As such, individuals who take fewer risks may also be less inclined to disclose traumas. Further research is required to elucidate the mechanisms that explain the genetic association between general risk tolerance and reported trauma.
Subjective well-being captures an individual’s cognitive evaluation of life satisfaction and positive affectivity (
Okbay A, Baselmans BML, De Neve J-E, Turley PA, Nivard MG, Fontana MA, et al. (2016): Genetic associations with subjective well-being also implicate depression and neuroticism. Nat Genet. Retrieved from https://dash.harvard.edu/handle/1/32303189
). Our findings may reflect the role of such cognitions in the perception, recall, and thus reporting of trauma exposure. As found with life events, those with a genetic propensity to positive subjective well-being may be less likely to report trauma retrospectively, while the converse may occur with negative subjective well-being (
Baldwin JR, Reuben A, Newbury JB, Danese A (2019): Agreement Between Prospective and Retrospective Measures of Childhood Maltreatment. JAMA Psychiatry, vol. 76. p 584.
). This could explain why some individuals with objective records of experiencing childhood maltreatment do not retrospectively self-report maltreatment (
). Positive emotional affectivity may contribute to resiliency by countering the adverse effects of stressful experiences due to a lack of subjective valence (
). Conversely, if negative affectivity indicates greater sensitivity to trauma and vulnerability to psychopathology, this could have implications for screening for psychopathology risk following trauma (
Polygenic risk for autism, attention-deficit hyperactivity disorder, schizophrenia, major depressive disorder, and neuroticism is associated with the experience of childhood abuse.
). Family-based polygenic score analyses exploring genetic differences between siblings suggest a greater risk of childhood maltreatment in those with ASD is partly attributable to evocative and active rGE (
). Difficulty processing social cues may place an individual at greater risk of harmful environmental situations such as exploitation by a perpetrator (
Rumball F, Happé F, Grey N (2020): Experience of Trauma and PTSD Symptoms in Autistic Adults: Risk of PTSD Development Following DSM-5 and Non-DSM-5 Traumatic Life Events. Autism Res.. https://doi.org/10.1002/aur.2306
). An association between the polygenic score for ASD and trauma has been consistently found with retrospectively but not prospectively reported trauma (
Polygenic risk for autism, attention-deficit hyperactivity disorder, schizophrenia, major depressive disorder, and neuroticism is associated with the experience of childhood abuse.
). Together, these findings suggest the importance of subjective trauma interpretation in ASD. Future research should determine the specific heritable components of ASD related to the subjective experience of and exposure to trauma, and the potential for screening for posttraumatic symptoms in ASD to provide appropriate support (
Rumball F, Happé F, Grey N (2020): Experience of Trauma and PTSD Symptoms in Autistic Adults: Risk of PTSD Development Following DSM-5 and Non-DSM-5 Traumatic Life Events. Autism Res.. https://doi.org/10.1002/aur.2306
As trauma exposure is necessary for a PTSD diagnosis, the unique genetic association between reported trauma and PTSD symptoms could be explained by trauma exposure increasing the risk of PTSD. However, there are plausible reverse or bidirectional mechanisms. This includes passive rGE, whereby parental genetic predisposition to PTSD may act to increase the risk of trauma exposure in the child, as suggested by studies of PTSD and parenting (
). However, evidence suggests the association between a higher genetic risk for PTSD and increased self-report of childhood trauma could be explained by subjective interpretation processes and not rGE (
). Except for ASD and PTSD, genetic associations between reported childhood maltreatment and the other psychiatric disorders were explained by genetic factors shared across all other psychiatric disorders included in the model. However, GWAS often use brief phenotypic measures to achieve sufficient power, which may impact our ability to detect disorder-specific genetic influences (
). Shared genetics may underlie transdiagnostic psychological mechanisms, such as those involved in the subjective experience of trauma, which is more robustly associated with psychiatric disorders than objective measures of trauma (
Baldwin JR, Wang B, Karwatowska L, Schoeler T, Tsaligopoulou A, Munafò MR, Pingault J-B (2023): Childhood Maltreatment and Mental Health Problems: A Systematic Review and Meta-Analysis of Quasi-Experimental Studies. Am J Psychiatry appiajp20220174.
Our modelling approach has several limitations. First, the model is fully saturated, and we could not objectively estimate which model best fits the data. Second, we were limited by the number of traits that could be fitted in one model. However, sensitivity analyses did not indicate that all of the residual genetic variance of childhood maltreatment could be explained if all available genetically correlated traits were accounted for in one model. Third, our estimates are based on the proportion of on lower-bounds of the total genetic variance explained by common genetic variants that can be, as captured by h2SNP estimated from summary statistics and may differ from results using more advanced methods or individual-level genetic data (
Coleman JRI, Peyrot WJ, Purves KL, Davis KAS, Rayner C, Choi SW, et al. (2020): Genome-wide gene-environment analyses of major depressive disorder and reported lifetime traumatic experiences in UK Biobank. Mol Psychiatry. https://doi.org/10.1038/s41380-019-0546-6
Further research is needed to establish the role of traits in terms of whether they are associated with the risk of exposure or the interpretation and recollection of events. Preliminary evidence suggests that the h2SNP of prospective trauma is lower than retrospectively reported trauma (
). Thus, genetic factors involved in the retrospective reporting of trauma may have a greater impact on the h2SNP of trauma than traits involved in the exposure of events. The residual h2SNP of trauma may reflect traits involved in memory recall (
) that lack adequately powered GWASs. Alternatively, unaccounted for parental traits involved in passive rGE, such as parental harmful drug use contributing to an unsafe environment, may partly explain the residual genetic variance. Disentangling the genetic associations with vulnerability to exposure and those with the subjective experience of trauma will be important for distinguishing whether factors are relevant to trauma prevention or posttraumatic interventions.
In summary, we systematically examined traits genetically correlated with reported trauma, implicating possible mechanisms that partly explain the h2SNP of trauma. Potentially, indirect genetic effects regulating behaviour and cognition are associated with trauma exposure and/or retrospectively self-reporting trauma. We emphasise that our findings do not suggest that an individual is ever at fault or responsible for their exposure to trauma. Furthermore, an h2SNP of reported trauma does not mean some individuals are genetically determined to experience trauma. Most of the phenotypic variance of reported trauma is explained by the environment, which is malleable and can be modified into a more supportive and protective environment to mitigate vulnerabilities. For example, if a genetic propensity for ASD, PTSD, general risk tolerance, and subjective well-being reflect genetic risk to reported trauma, more social support may protect such individuals and alleviate adverse posttraumatic effects. However, our findings are correlational, not necessarily causal, and better delineation of the processes involved is needed. Future studies could assess the specific role of these traits in large family-based datasets using within-family designs (
). Disentangling passive from evocative and active rGE that may explain our trait-specific associations could have implications for prevention strategies. As GWASs increase in power, the direction of causal relationships or testing the types of pleiotropy could be explored through Mendelian randomisation techniques (
Causal Inference in Psychopathology: A Systematic Review of Mendelian Randomisation Studies Aiming to Identify Environmental Risk Factors for Psychopathology.
). Such approaches are crucial to understanding vulnerability to trauma exposure and the subjective interpretation of trauma.
Acknowledgements
We thank the contribution of participants from all cohorts who have shared their life experiences. We gratefully acknowledge the scientists involved in the construction of all cohorts and those who provided the genetic summary results used in this study. We thank Dr A. Grotzinger for his guidance and contribution to the interpretation of our findings using genomic multiple regression with the Genomic SEM R package. This study represents years of independent research funded by the National Institute for Health and Care Research (NIHR) Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King's College London. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR, the Department of Health and Social Care or King's College London. C.H. is supported by Lundbeck Foundation (Grant no. R276-2018-4581). The authors acknowledge use of the research computing facility at King’s College London, Rosalind (https://rosalind.kcl.ac.uk), which is delivered in partnership with the National Institute for Health Research (NIHR) Biomedical Research Centres at South London & Maudsley and Guy’s & St. Thomas’ NHS Foundation Trusts, and part-funded by capital equipment grants from the Maudsley Charity (award 980) and Guy’s & St. Thomas’ Charity (TR130505). Figure 3 and figure 4 were created with BioRender.com.
Turner RJ, Jay Turner R, Lloyd DA (1995): Lifetime Traumas and Mental Health: The Significance of Cumulative Adversity. Journal of Health and Social Behavior, vol. 36. p 360.
Baldwin JR, Reuben A, Newbury JB, Danese A (2019): Agreement Between Prospective and Retrospective Measures of Childhood Maltreatment. JAMA Psychiatry, vol. 76. p 584.
van Rheenen W, Peyrot WJ, Schork AJ, Lee SH, Wray NR (2019): Genetic correlations of polygenic disease traits: from theory to practice. Nat Rev Genet.. https://doi.org/10.1038/s41576-019-0137-z
Coleman JRI, Peyrot WJ, Purves KL, Davis KAS, Rayner C, Choi SW, et al. (2020): Genome-wide gene-environment analyses of major depressive disorder and reported lifetime traumatic experiences in UK Biobank. Mol Psychiatry. https://doi.org/10.1038/s41380-019-0546-6
Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, et al. (2015): UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med 12: e1001779.
Fraser A, Macdonald-Wallis C, Tilling K, Boyd A, Golding J, Davey Smith G, et al. (2013): Cohort Profile: the Avon Longitudinal Study of Parents and Children: ALSPAC mothers cohort. Int J Epidemiol 42: 97–110.
Boyd A, Golding J, Macleod J, Lawlor DA, Fraser A, Henderson J, et al. (2013): Cohort profile: the “children of the 90s”—the index offspring of the Avon Longitudinal Study of Parents and Children. Int J Epidemiol 42: 111–127.
Kooijman MN, Kruithof CJ, van Duijn CM, Duijts L, Franco OH, van IJzendoorn MH, et al. (2016): The Generation R Study: design and cohort update 2017. Eur J Epidemiol 31: 1243–1264.
Bulik-Sullivan BK, Loh P-R, Finucane HK, Ripke S, Yang J, Schizophrenia Working Group of the Psychiatric Genomics Consortium, et al. (2015): LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nat Genet 47: 291–295.
LD Hub: a centralized database and web interface to perform LD score regression that maximizes the potential of summary level GWAS data for SNP heritability and genetic correlation analysis.
Nivard MG (2019, February 5): Minimum sample size, SNP-h2 and rG? Genomic SEM users google group. Retrieved January 14, 2022, from https://groups.google.com/g/genomic-sem-users/c/Wn1jptU2VcY/m/9CWxfQbvAAAJ
Team RC (n.d.): R Core Team (2019) RA Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. References-Scientific Research Publishing.
Genome-wide association analyses of risk tolerance and risky behaviors in over 1 million individuals identify hundreds of loci and shared genetic influences.
Okbay A, Baselmans BML, De Neve J-E, Turley PA, Nivard MG, Fontana MA, et al. (2016): Genetic associations with subjective well-being also implicate depression and neuroticism. Nat Genet. Retrieved from https://dash.harvard.edu/handle/1/32303189
Polygenic risk for autism, attention-deficit hyperactivity disorder, schizophrenia, major depressive disorder, and neuroticism is associated with the experience of childhood abuse.
Rumball F, Happé F, Grey N (2020): Experience of Trauma and PTSD Symptoms in Autistic Adults: Risk of PTSD Development Following DSM-5 and Non-DSM-5 Traumatic Life Events. Autism Res.. https://doi.org/10.1002/aur.2306
Baldwin JR, Wang B, Karwatowska L, Schoeler T, Tsaligopoulou A, Munafò MR, Pingault J-B (2023): Childhood Maltreatment and Mental Health Problems: A Systematic Review and Meta-Analysis of Quasi-Experimental Studies. Am J Psychiatry appiajp20220174.
Causal Inference in Psychopathology: A Systematic Review of Mendelian Randomisation Studies Aiming to Identify Environmental Risk Factors for Psychopathology.
8 PROMENTA Research Center, Department of Psychology, University of Oslo, Oslo, Norway
Conflict of interest
G.B. has received honoraria, research or conference grants and consulting fees from Illumina, Otsuka, and COMPASS Pathways Ltd. All other authors report no biomedical financial interests or potential conflicts of interest.
Author contributions: CRediT author statement
Abigail R. ter Kuile: Conceptualization, Formal analysis, Writing - Original Draft, Writing - Review & Editing, Visualisation. Christopher Hübel: Conceptualization, Formal analysis, Data Curation, Writing - Review & Editing, Visualisation. Rosa Cheesman: Conceptualization, Writing - Review & Editing. Jonathan R. I. Coleman: Conceptualization, Data Curation, Writing - Review & Editing. Alicia Peel: Conceptualization, Writing - Review & Editing. Daniel F. Levey: Data Curation, Writing - Review & Editing. Murray B. Stein: Data Curation, Writing - Review & Editing. Joel Gelernter: Data Curation, Writing - Review & Editing. Christopher Rayner: Formal analysis, Writing - Review & Editing. Thalia C. Eley: Conceptualization, Writing - Original Draft, Writing - Review & Editing, Supervision. Gerome Breen: Conceptualization, Writing - Original Draft, Writing - Review & Editing, Supervision.
Data availability
Data availability of all GWAS summary statistics used for analyses can be found in Supplementary Table 12. Access to Million Veteran Programme PTSD GWAS summary statistics can be requested through bdGaP (phs001672.v7.p1).
00037-X&id=gr1.jpg){kind=link}
00037-X&id=gr2.jpg){kind=link}