Journal of Child Psychology and Psychiatry 66:10 (2025), pp 1514–1525 doi:10.1111/jcpp.

14163

Autistic traits in childhood and post-traumatic stress

disorder as young adults: a cohort study
Alice M.G. Quinton,1 Freya Rumball,1,2 Angelica Ronald,3 Helen L. Fisher,1,4
Louise Arseneault,1 Francesca Happe
,1 and Andrea Danese1,5,6
1
Social, Genetic & Developmental Psychiatry Centre, Institute of Psychiatry, Psychology and Neuroscience, King’s
College London, London, UK; 2Adult Autism Assessment Service, Oxleas NHS Foundation Trust, Bexley, UK; 3School
of Psychology, Faculty of Health and Medical Sciences, University of Surrey, Guildford, UK; 4ESRC Centre for Society
and Mental Health, King’s College London, London, UK; 5Department of Child & Adolescent Psychiatry, Institute of
Psychiatry, Psychology and Neuroscience, King’s College London, London, UK; 6National and Specialist CAMHS
Trauma and Anxiety Clinic, South London and Maudsley NHS Foundation Trust, London, UK

Background: Despite the higher prevalence of childhood traumatic experiences and post-traumatic stress disorder
(PTSD) in autistic adults, research on trauma-related psychopathology and autistic traits in young people is lacking.
This study examined if high autistic traits in childhood predispose individuals to traumatic experiences, the
development of PTSD and general psychopathology, and greater functional impairment by age 18, in both the general
population and a subsample of trauma-exposed young people. Methods: Data were utilised from the Environmental
Risk (E-Risk) Longitudinal Twin Study, a nationally representative cohort of 2,232 same-sex twins born in
1994–1995 across England and Wales. Participants were a subset of children whose parents completed the
Childhood Autism Spectrum Test (CAST), during assessments at ages 8, 9 and/or 12 years (N = 1,504). We tested
associations between autistic traits in childhood and age-18 reports of lifetime trauma exposure, lifetime PTSD
diagnosis, general psychopathology (‘p-factor’) and NEET status (‘not in employment, education or training’).
Analyses were conducted controlling for sex, family socioeconomic status (SES), intelligence quotient (IQ) and
accounting for family clustering. Results: Higher autistic traits in childhood were significantly associated with
greater reports of lifetime trauma exposure (Odd Ratio [OR] = 1.26, 95% Confidence Intervals [CI] = 1.03; 1.54),
lifetime PTSD diagnosis (OR = 1.91, 95% CI = 1.29; 2.82), general psychopathology (beta = 3.22, 95% CI = 1.84;
4.60) and NEET status (OR = 1.48, 95% CI = 1.05; 2.09) at age 18. Only the associations of autistic traits with PTSD
and general psychopathology were robust to adjustment for potential confounders. Among trauma-exposed children,
autistic traits were also significantly associated with lifetime PTSD diagnosis (OR = 1.75, 95% CI = 1.15; 2.68) and
psychopathology (beta = 3.36, 95% CI = 0.68; 6.04) at age 18, but only the association with PTSD held when
adjusted for confounders. Conclusions: Our findings suggest a need to develop targeted assessments and
evidence-based treatments for PTSD to meet the needs of children with high autistic traits. However, whether our
findings extend to diagnosed autistic children requires further investigation. Keywords: Autism; autism spectrum;
trauma; PTSD; psychopathology.

non-autistic peers (Kerns, Rast, & Shattuck, 2020)

Background
and might be at a greater risk of developing PTSD
Post-traumatic stress disorder (PTSD) is a debilitat-
and other types of psychopathology after trauma
ing psychiatric condition that can develop after
exposure (Kerns, Newschaffer, & Berkowitz, 2015).
experiencing a traumatic event (Danese, McLaugh-
Research with adults suggests that autistic traits
lin, Samara, & Stover, 2020). PTSD is characterised
may facilitate the development of trauma-related
by persistent re-experiencing of traumatic events
psychopathology. In a large study, 251 middle-aged
through intrusive and distressing memories, as well
and older adults (aged 50–81) with high autistic
as negative appraisals, avoidance and hypervigilance
traits showed a 12-fold increase in passing the cut-
to threat. In this article, we examine if higher autistic
off for probable PTSD when compared with 9,179
traits in childhood are associated with trauma
age- and gender-matched low-trait controls (Stewart
exposure, PTSD and broader psychopathology and
et al., 2020). Likewise, a study in 103 typically
functional impairment in youths. Clinical, cognitive
developing adults aged 18–34 showed positive asso-
and genetic findings in the literature suggest that
ciations between autistic traits and PTSD symptoms,
this might be the case.
particularly hyper-arousal symptoms (Haruvi-
Compared with non-autistic peers, autistic chil-
Lamdan, Lebendiger, Golan, & Horesh, 2019). How-
dren have a greater risk of adversities, such as
ever, empirical research on trauma-related psycho-
bullying and victimisation (Hoover & Kauf-
pathology in children and young people on the
man, 2018). Autistic children also show a higher
autistic spectrum is lacking.
prevalence of mental health conditions than their
Children with higher autistic traits likely share
similar cognitive styles to autistic children (Best,
Conflict of interest statement: No conflicts declared. Moffat, Power, Owens, & Johnstone, 2008).
Ó 2025 The Author(s). Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for Child and
Adolescent Mental Health.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any
medium, provided the original work is properly cited.
doi:10.1111/jcpp.14163 Childhood autistic traits & PTSD in young adults 1515

Rumball, Happ
e, & Grey (2020)have suggested that UK-based study found that young people in the
specific cognitive characteristics of autism may general population with a history of trauma were
impact how a negative or traumatic event is pro- more likely to be NEET than unexposed peers, and
cessed and perceived and may predispose individ- those with PTSD were more likely to be NEET than
uals to the development of PTSD symptoms. Popular those without PTSD (Lewis et al., 2019). However, it
models of PTSD highlight the role of cognitive factors is unclear how the presence of both trauma exposure
in promoting the development and maintenance of and high autistic traits could affect this measure of
PTSD symptoms (Brewin, Dalgleish, & Joseph, 1996; functional impairment.
Ehlers & Clark, 2000). Research has shown that Building on the evidence above, in this study, we
several of these cognitive risk factors are commonly investigated if higher autistic traits in childhood
found in autistic people or those with more autistic predispose individuals to trauma exposure, PTSD
traits, including detail-focused processing (Happ
e& and worse general psychopathology, as well as
Frith, 2006), sensory sensitivities (Weiland, Polder- greater functional impairment by age 18 in a large
man, Hoekstra, Smit, & Begeer, 2020), rumination UK birth cohort. We considered that any associa-
(Golan, Haruvi-Lamdan, Laor, & Horesh, 2022), tions between autistic traits and trauma-related
emotional dysregulation (Mazefsky, Borue, Day, & psychopathology and functional impairment could
Minshew, 2014), social withdrawal (Brosnan & either stem from a greater likelihood of developing
Gavin, 2023), and poor verbal working memory negative outcomes after trauma exposure or reflect a
(Wang et al., 2017). Studies with autistic adults greater likelihood of being exposed to trauma. To
have also reported that specific cognitive features disentangle the relative contributions of the two
common to ASD relate to increased PTSD symptoms, mechanisms, we therefore repeated the analyses in
such as everyday and working memory deficits a subset of the overall sample, including only
(Rumball, Brook, Happ
e, & Karl, 2021), brooding trauma-exposed young people.
rumination (Golan et al., 2022) and thought sup-
pression (Rumball, Antal, et al., 2021).
Genetic research also suggests possible links Method
between high autistic traits and risk for traumatic This study was pre-registered with OSF ([Link]
experiences. Traumatic experiences are partly influ-
enced by heritable factors (Dahoun et al., 2024).
Previous research showed that higher polygenic Sample
scores for autism may be associated with
Participants were members of the Environmental Risk (E-Risk)
self-reported childhood trauma (Peel et al., 2022; Longitudinal Twin Study, which tracks the development of
Ratanatharathorn et al., 2021; Warrier & Baron- 2,232 British children. The sample was drawn from a larger
Cohen, 2021; but see Sallis et al., 2021 for contrary birth cohort of twins born in England and Wales in 1994–1995,
results). However, it is unclear if polygenic scores for the Twins Early Development Study (TEDS) (Trouton, Spinath,
autism also increase the risk of trauma-related & Plomin, 2002). Full details about the sample are reported in
Appendix S1 and described elsewhere (Moffitt & the E-Risk
psychopathology (Huckins et al., 2021). Genetic Study Team, 2002). Briefly, E-Risk was constructed in
research has also highlighted that similar genetic 1999–2000, when 1,116 families (93% of those eligible) with
factors influence diagnosed autism and autistic same-sex 5-year-old twins participated in home-visit assess-
traits in the general population (Robinson ments. This sample comprised 56% monozygotic and 44%
et al., 2011). Along with consistent behavioural and dizygotic twin pairs; sex was evenly distributed within zygosity
(49% male); 90% of participants were of White ethnicity.
genetic evidence (Happ
e and Frith, 2020), these Although sampled from England and Wales alone, the sample
findings support a dimensional approach to exam- represents the full socioeconomic spectrum of the UK popula-
ining the role of autistic traits in trauma-related tion, as reflected in the families’ distribution on
psychopathology and related functional impairment. neighbourhood-level socioeconomic indices (Odgers, Caspi,
Although high autistic traits clearly do not equate to Bates, Sampson, & Moffitt, 2012; Reuben et al., 2020) (see
Appendix S1, Figure S1). Follow-up home visits were con-
an autism diagnosis, research on autistic traits can ducted when the children were aged 7 (98% participation), 10
provide novel insights into liability to trauma expo- (96%), 12 (96%) and 18 years (93%). Visits at ages 5–12
sure and trauma-related psychopathology in the included assessments with participants and their mother
general population and prompt more focused work (primary caretaker) and at age 18 included interviews with
in clinical samples. participants. The Joint South London and Maudsley and the
Institute of Psychiatry Research Ethics Committee approved
To examine the wider impact of trauma beyond the each phase of the study. Parents gave informed consent, and
risk for PTSD and general psychopathology, it is also participants gave assent between 5 and 12 years, then
important to test whether children with high autistic informed consent at age 18.
traits have greater functional impairment in daily life In this study, we sought to capitalise on the nested structure
compared with peers with less autistic traits. Being of the E-Risk study within TEDS, as they have complementary
strengths in the assessment of trauma-related disorders and
‘not in education or employment’ (NEET) is a autism, respectively. The sample in these two studies became
common metric of functional impairment in young increasingly non-overlapping over time because of the
adults. Previous work in a large, longitudinal follow-up methods used (in-person assessment for the E-Risk

Ó 2025 The Author(s). Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for
Child and Adolescent Mental Health.
1516 Alice M.G. Quinton et al. J Child Psychol Psychiatr 2025; 66(10): 1514–25

Study; telephone/online assessment for TEDS), which reduced

the size of the sample with data available from both studies. We Not in education, employment or training (NEET)
identified participants with complete data for PTSD from the status. At the time of their age-18 interview, participants
E-Risk study (N = 2,061) who had parent-reported Childhood were classified as NEET if they reported that they were neither
Autism Spectrum Test (CAST) (Scott, Baron-Cohen, Bolton, & studying nor working in paid employment nor pursuing a
Brayne, 2002) from TEDS (N = 1,510). The majority had CAST vocational qualification or apprenticeship training. Partici-
data at age 8 (n = 1,213). Because CAST scores showed high pants were asked to ensure that NEET status was not simply a
correlations (r ~ .6; Appendix S2) and were stable function of being on summer holiday or of being a parent. The
(Appendix S2, Figure S2) between ages 8, 9 and 12 years, in employment- and education-related questions were those used
order to maximise the analytical sample, we used available in the Longitudinal Study of Young People in England (LSYPE).
parent-reported CAST scores at ages 9 or 12 where scores at This operationalisation of NEET status follows that used by the
age 8 were not available. This resulted in a sample of 1,510 UK Office of National Statistics and the International Labour
E-Risk Study members with complete data for both autistic Organisation (Chandler & Barrett, 2013).
traits and PTSD. Six participants (0.4% of the sample) had
missing IQ data at age 5 and were removed from the analyses, Potentially confounding variables. Covariates for the
resulting in the final analytical sample of 1,504 participants. adjusted models were selected because previous literature
This sample and the overall E-Risk participants at age 18 reported a relationship with both autistic traits and PTSD,
showed similar distributions for variables included in the suggesting their potential role as confounders. In particular,
analyses (see Appendix S3, Table S1). previous research found that the prevalence of both autistic
traits and PTSD varies according to sex (Lewis et al., 2019;
Napolitano et al., 2022), IQ (Koenen, Moffitt, Poulton, Martin,
Measures & Caspi, 2007; Marinopoulou et al., 2025) and socio-economic
status (Lewis et al., 2019; Skylark & Baron-Cohen, 2017).
Trauma exposure and PTSD measure. Trauma Biological sex and childhood IQ scores were obtained at first
exposure and PTSD to date were assessed at age 18 years,
contact with the twins at age 5. IQ was assessed using
during face-to-face interviews with the twin participants
Vocabulary and Block Design subtests on a short form of the
conducted by the E-Risk Study team. Participants were asked
Wechsler Preschool and Primary Scale of Intelligence-Revised
whether they had been exposed to trauma during their lifetime,
and prorated according to Sattler (1988). Family SES was
according to DSM-5 PTSD Criterion A (American Psychiatric
defined through a standardised composite of parental income,
Association, 2013). Participants who reported trauma expo-
education and occupation obtained from mothers at the age-5
sure were then asked to describe the trauma they had
assessment (Trzesniewski, Moffitt, Caspi, Taylor, &
experienced. Participants who reported trauma exposure were
Maughan, 2006). In the full E-Risk dataset, the
also further evaluated through an adapted version of the
population-wide distribution of this composite variable was
Diagnostic Interview Schedule (Robins, Cottler, Bucholz, &
then split into tertiles. In the present study, the variable was
Compton, 1995) to assess for current (past-year) and lifetime
coded so that 1 indicates the highest socioeconomic status,
PTSD according to the DSM-5 criteria. Our analyses focus on
and 3 indicates the lowest.
E-Risk Study members meeting criteria for lifetime PTSD, but
results for current PTSD are also reported in sensitivity
analyses detailed in the Appendix S6.
Data analysis
Autistic traits measure. The Childhood Autism Spec- CAST scores were log-transformed to approximate a normal
trum Test (CAST) (originally named ‘Childhood Asperger’s distribution. Bivariate logistic regression models were used to
Syndrome Test’) (Scott et al., 2002) was designed as a investigate the association of autistic traits with (1) report of
parent-report screening instrument for autism in non-clinical trauma exposure (PTSD Criterion A), (2) PTSD diagnosis and
samples. The CAST is a quantitative scale; all questions are (3) NEET status (from age 18 years assessment). Bivariate
answered ‘yes’ or ‘no’ by the child’s mother (or primary linear regression models were used to test the association of
caregiver) and are scored additively. Where items had missing autistic traits with general psychopathology. These bivariate
data, if participants’ parents had completed more than half of models were then expanded to account for the influence of
the questions, the total score was calculated by summing the possible confounding factors (SES, IQ and sex) in multivariate
completed items and finding the average based on the total analyses, first accounting for each single potential confounder
possible scores. CAST parent-report total score at age 8 was and then in a saturated model including all potential
selected as the primary measure of autistic traits, and to confounders at once. The analyses were first carried out in
increase the coverage, missing scores were replaced with CAST the overall sample and subsequently in the subsample of
scores taken when participants were 9 or 12 years old (see trauma-exposed participants in order to test if any associa-
Appendix S2). tions of autistic traits and outcomes could be explained by
greater risk for trauma exposure in participants with higher
autistic traits. Analysis code is available at the study’s OSF
General psychopathology measure. The ‘p’ factor is a pre-registration ([Link]
composite index of general psychopathology. The measure of We also undertook sensitivity analyses to test the possible
‘p’ was computed via a bi-factor model that was fitted to 11 biases introduced by our approach to missingness in CAST
distinct symptom scales encompassing various mental health data and examining past-year rather than lifetime PTSD
conditions (Schaefer et al., 2018). The symptoms were diagnosis. Standard errors were adjusted to account for the
assessed in the E-Risk Study during the age-18 follow-up with non-independence of twin observations (Williams, 2000)
a past-year reporting period including major depressive applying heteroskedasticity-robust standard errors while clus-
disorder, generalised anxiety disorder, PTSD, disordered tering for membership in the same family. Because our
eating, attention-deficit hyperactivity disorder (ADHD), con- analytical sample and the overall E-Risk sample at age 18
duct disorder, alcohol dependence, cannabis dependence, showed similar distributions for variables included in the
nicotine dependence, psychotic symptoms and prodromal analyses (Appendix S3, Table S1), we carried out analyses on
symptoms. A description of how each of these was assessed study members with complete data. None of the included
is provided in Appendix S4. After scaling the scores, the mean variables breached multicollinearity thresholds (Appendix S5,
value was set to 100, and the standard deviation was fixed Figure S4).
at 15.

Ó 2025 The Author(s). Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for
Child and Adolescent Mental Health.
doi:10.1111/jcpp.14163 Childhood autistic traits & PTSD in young adults 1517

unadjusted and fully adjusted results are displayed

Results
in Figure 2A,B.
As shown in Figure 1, the overall sample consisted of
1,504 people, 460 of whom reported trauma exposure
Trauma exposure. Children with higher autistic
by age 18 and were therefore further assessed for PTSD.
traits in childhood had a significantly increased
Of those who reported trauma exposure, 24% (n = 110)
likelihood of reporting exposure to traumatic events
participants met PTSD diagnostic criteria. The overall
by 18 years (OR = 1.26, 95% CI = 1.03; 1.54;
analytical sample was 53.8% female (n = 809), had a
Table 2, Panel A). This association was attenuated
mean CAST score of 5.47, a mean IQ of 102, and was
and became statistically non-significant when
similarly distributed in SES tertiles. Table 1 shows the
accounting for SES (OR = 1.17, 95% CI = 0.96;
demographic characteristics of the overall sample and
1.44) and in the fully adjusted model (OR = 1.20,
trauma-exposed subsample, as well as details on
95% CI = 0.97; 1.48).
trauma-unexposed participants, all participants with
no PTSD, trauma-exposed participants with no PTSD,
and those with PTSD. PTSD diagnosis. Children with higher autistic traits
had a significantly increased likelihood of meeting the
diagnostic criteria for PTSD by the age of 18 (OR = 1.91,
Potential confounders 95% CI = 1.29; 2.82; Table 2, Panel B). This association
Associations of key potential confounders (i.e. sex, remained significant and was enhanced when sex
IQ and SES) with the predictor variable (autistic (OR = 2.10, 95% CI = 1.40; 3.14) was accounted for,
traits) and each outcome variable are reported in but was slightly attenuated when IQ (OR = 1.78, 95%
Appendix S5. In brief, children with autistic traits CI = 1.20; 2.64) and SES (OR = 1.59, 95% CI = 1.08;
were less likely to be female in the overall sample. In 2.33) were accounted for individually and in the fully
both the overall and trauma-exposed samples, adjusted model (OR = 1.77, 95% CI = 1.17; 2.66).
autistic traits were associated with lower IQ. Asso-
ciations between the potential confounders and General psychopathology. Children with higher
outcome variables are shown in the first panels of autistic traits had a significantly increased likeli-
Tables 2 and 3. In both the overall sample and the hood of having more severe psychopathology at age
trauma-exposed subsample, univariate analyses 18 (Beta = 3.22, 95% CI = 1.84; 4.60; Table 2, Panel
with potential confounders showed significant asso- D). This statistically significant relationship was
ciations between low SES and all outcomes, and maintained when the analysis was individually
female sex with PTSD and general psychopathology. adjusted for potential confounders; the association
was enhanced when sex was accounted for
(Beta = 3.53, 95% CI = 2.16; 4.91) and attenuated
Associations of autistic traits with trauma exposure,
when adjusted for IQ (Beta = 2.79, 95% CI = 1.41;
PTSD, general psychopathology and functional
4.17) or SES (Beta = 2.46, 95% CI = 1.05; 3.86). The
impairment in the overall sample
association also remained significant, albeit slightly
Full results of analyses in the overall sample attenuated, in the fully adjusted analysis
(N = 1,504) are presented in Table 2, and the main (Beta = 2.65, 95% CI = 1.25; 4.04).

Figure 1 Flow chart illustrating sample selection

Ó 2025 The Author(s). Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for
Child and Adolescent Mental Health.
1518 Alice M.G. Quinton et al. J Child Psychol Psychiatr 2025; 66(10): 1514–25

Table 1 Sample description

Sample

Trauma exposed Trauma unexposed Overall sample Trauma exposed Trauma exposed
Overall participants participants with no PTSD with no PTSD with PTSD
(N = 1,504) (n = 460) (n = 1,044) (n = 1,394) (n = 350) (n = 110)

Sex, N (%)
Male 695 (46.2) 201 (43.7) 494 (47.3) 661 (47.4) 167 (47.7) 34 (30.9)
Female 809 (53.8) 259 (56.3) 550 (52.7) 733 (52.6) 183 (52.3) 76 (69.1)
SES, N (%)
High SES 576 (38.3) 153 (33.3) 423 (40.5) 553 (39.7) 130 (37.1) 23 (20.9)
Medium SES 500 (33.2) 152 (33.0) 348 (33.3) 462 (33.1) 114 (32.6) 38 (34.5)
Low SES 428 (28.5) 155 (33.7) 273 (26.1) 379 (27.2) 106 (30.3) 49 (44.5)
NEET N (%) 151 (10.0) 71 (15.4) 80 (7.7) 125 (9.0) 45 (12.9) 84 (76.4)
Autistic traits 5.47 (3.5) 5.79 (3.6) 5.33 (3.5) 5.38 (3.5) 5.52 (3.5) 6.62 (3.9)
(max = 31)
mean (SD)
IQ mean (SD) 102 (14.9) 101 (14.2) 102 (15.2) 102 (14.9) 102 (14.1) 99.0 (14.4)
‘p’ factor mean 99.2 (14.7) 107.0 (15.8) 95.5 (12.5) 97.4 (13.3) 103.0 (13.9) 121.0 (13.7)
(SD)

Data are n/N (%) unless stated otherwise. IQ, intelligence quotient; NEET, not in education, employment, or training; PTSD,
post-traumatic stress disorder; SES, socio-economic status; ‘p’-factor, measure of general psychopathology.

Table 2 Results from the overall sample (N = 1,504); univariate models of autistic traits and confounding variables’ individual
relationships with outcomes at age 18, and multivariate models showing the association between autistic traits and outcomes of
interest, adjusting for confounding variables

Multivariate models adjusted for:

Univariate models Sex IQ SES Sex, IQ and SES

Panel A: Associations with trauma exposure in the overall sample-OR [95% CI]
Autistic traits 1.26 [1.03–1.54] 1.29 [1.05–1.58] 1.23 [1.00–1.50] 1.17 [0.96–1.44] 1.20 [0.97–1.48]
Female sex 1.16 [0.93–1.44] 1.20 [0.96–1.50] – – 1.18 [0.94–1.48]
IQ 0.99 [0.99–1.00] – 1.00 [0.99–1.00] – 1.00 [0.99–1.01]
Medium SES 1.21 [0.93–1.58] – – 1.18 [0.90–1.54] 1.17 [0.89–1.55]
Low SES 1.57 [1.20–2.06] – – 1.49 [1.13–1.97] 1.46 [1.08–1.98]
Panel B: Associations with PTSD diagnosis in the overall sample-OR [95% CI]
Autistic traits 1.91 [1.29–2.82] 2.10 [1.40–3.14] 1.78 [1.20–2.64] 1.59 [1.08–2.33] 1.77 [1.17–2.66]
Female sex 2.02 [1.33–3.06] 2.24 [1.46–3.44] – – 2.21 [1.43–3.41]
IQ 0.99 [0.97–1.00] – 0.99 [0.98–1.00] – 1.00 [0.99–1.02]
Medium SES 1.98 [1.16–3.37] – – 1.85 [1.08–3.18] 1.89 [1.09–3.29]
Low SES 3.11 [1.86–5.19] – – 2.67 [1.58–4.53] 2.62 [1.49–4.59]
Panel C: Associations with NEET status in the overall sample – OR [95% CI]
Autistic traits 1.48 [1.05–2.09] 1.53 [1.08–2.17] 1.18 [0.85–1.64] 1.05 [0.76–1.47] 0.98 [0.70–1.37]
Female sex 1.22 [0.87–1.72] 1.30 [0.92–1.84] – – 1.18 [0.82–1.70]
IQ 0.96 [0.95–0.98] – 0.97 [0.95–0.98] – 0.98 [0.96–0.99]
Medium SES 1.51 [0.88–2.58] – – 1.49 [0.87–2.58] 1.26 [0.73–2.19]
Low SES 6.20 [3.91–9.85] – – 6.10 [3.78–9.83] 4.68 [2.86–7.65]
Panel D: Associations with the ‘p’-factor in the overall sample – Beta [95% CI]
Autistic traits 3.22 [1.84–4.60] 3.53 [2.16–4.91] 2.79 [1.41–4.17] 2.46 [1.05–3.86] 2.65 [1.25–4.04]
Female sex 2.33 [0.85–3.80] 2.83 [1.36–4.29] – – 2.64 [1.17–4.11]
IQ 0.10 [ 0.15 to 0.05] – 0.08 [ 0.13 to 0.03] – 0.04 [ 0.09–0.02]
Medium SES 2.22 [0.53–3.91] – – 1.84 [0.13–3.55] 1.57 [ 0.19–3.32]
Low SES 5.28 [3.42–7.14] – – 4.46 [2.55–6.37] 3.85 [1.79–5.90]

Values in bold text indicate statistically significant results (p < .05). Values in Italic text are contributions of confounding variables
to the association between autistic traits and outcomes of interest. All models are adjusted for the non-independence of twin
observations. 95% CI, 95% confidence intervals; Beta, beta coefficient; IQ, intelligence quotient; NEET, not in education,
employment, or training; OR, odds ratio; ‘p’-factor, measure of general psychopathology; PTSD, post-traumatic stress disorder; SES,
socio-economic status.

Functional impairment. Children with higher remained significant and slightly enhanced when
autistic traits had a significantly increased likeli- accounting for sex (OR = 1.53, 95% CI = 1.08; 2.17).
hood of being NEET at age 18 (OR = 1.48, 95% However, this association was attenuated and not
CI = 1.05; 2.09; Table 2, Panel C). This association statistically significant when IQ (OR = 1.18, 95%

Ó 2025 The Author(s). Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for
Child and Adolescent Mental Health.
doi:10.1111/jcpp.14163 Childhood autistic traits & PTSD in young adults 1519

Table 3 Results from the trauma-exposed subsample (n = 460); univariate models of autistic traits and confounding variables’
individual relationships with outcomes at age 18, and multivariate models showing associations between autistic traits and
outcomes of interest adjusting for confounding variables

Multivariate model adjusted for

Univariate models Sex IQ SES Sex, IQ and SES

Panel A: Associations with PTSD diagnosis in the trauma-exposed subsample – OR [95% CI]
Autistic traits 1.75 [1.15–2.68] 1.86 [1.21–2.88] 1.66 [1.08–2.54] 1.52 [1.00–2.32] 1.62 [1.04–2.53]
Female sex 2.04 [1.29–3.22] 2.17 [1.36–3.46] – – 2.18 [1.35–3.53]
IQ 0.99 [0.97–1.00] – 0.99 [0.98–1.01] – 1.00 [0.98–1.02]
Medium SES 1.88 [1.06–3.36] – – 1.80 [1.00–3.24] 1.84 [1.00–3.37]
Low SES 2.61 [1.49–4.58] – – 2.29 [1.29–4.08] 2.29 [1.24–4.21]
Panel B: Associations with NEET status in the trauma-exposed subsample – OR [95% CI]
Autistic traits 1.07 [0.65–1.77] 1.07 [0.64–1.77] 0.88 [0.56–1.39] 0.77 [0.46–1.29] 0.66 [0.40–1.08]
Female sex 0.94 [0.56–1.56] 0.94 [0.56–1.58] – – 0.85 [0.49–1.47]
IQ 0.96 [0.94–0.98] – 0.96 [0.94–0.98] – 0.97 [0.95–0.99]
Medium SES 1.19 [0.53–2.68] – – 1.23 [0.54–2.80] 0.99 [0.42–2.35]
Low SES 4.81 [2.41–9.57] – – 5.26 [2.55–10.85] 4.14 [2.00–8.58]
Panel C: Associations with the ‘p’-factor in the trauma-exposed sub sample – Beta [95% CI]
Autistic traits 3.36 [0.68–6.04] 3.58 [0.95–6.21] 2.68 [0.00–5.36] 2.47 [ 0.21–5.16] 2.35 [ 0.28–4.99]
Female sex 3.23 [0.30–6.16] 3.51 [0.60–6.41] – – 3.26 [0.34–6.18]
IQ 0.16 [ 0.26 - -0.06] – 0.14 [ 0.24 - -0.03] – 0.10 [ 0.21–0.01]
Medium SES 2.43 [ 1.06–5.92] – – 2.14 [ 1.38–5.66] 1.51 [ 2.00–5.02]
Low SES 5.75 [2.27–9.24] – – 4.94 [1.36–8.52] 3.92 [0.13–7.70]

Values in bold text indicate statistically significant results (p < .05). Values in Italic text are contributions of confounding variables
to the association between autistic traits and outcomes of interest. All models are adjusted for the non-independence of twin
observations. 95% CI, 95% confidence intervals; Beta, beta coefficient; IQ, intelligence quotient; NEET, not in education,
employment, or training; OR, odds ratio; ‘p’-factor, measure of general psychopathology; PTSD, post-traumatic stress disorder; SES,
socio-economic status.

CI = 0.85; 1.64) and SES (OR = 1.05, 95% CI = 0.76; This relationship was not significant when the model
1.47) were accounted for, nor when the model was was adjusted for confounders; the association was
adjusted for all possible confounders (OR = 0.98, enhanced when sex was accounted for (Beta = 3.58,
95% CI = 0.70; 1.37). 95% CI = 0.95; 6.21), but attenuated when control-
ling for IQ (Beta = 2.68, 95% CI = 0.00; 5.36) or SES
(Beta = 2.47, 95% CI = 0.21; 5.16), and in the fully
Associations of autistic traits with PTSD, general
adjusted model (Beta = 2.35, 95% CI = 0.28; 4.99).
psychopathology and functional impairment in the
trauma-exposed subsample
Functional impairment. In young people who
Full results of analyses in the trauma-exposed reported trauma exposure, those with higher autistic
subsample (n = 460) are given in Table 3 and the traits did not have a significantly increased likeli-
main unadjusted and fully adjusted results are hood of being NEET at age 18 (OR = 1.07, 95%
displayed in Figure 2C,D. CI = 0.65; 1.77; Table 3, Panel B). This remained
statistically non-significant when adjusting for con-
PTSD diagnosis. In young people who reported founders; accounting for sex had minimal impact on
trauma exposure, those with higher autistic traits the association (OR = 1.07, 95% CI = 0.64; 1.77),
had a significantly increased likelihood of meeting whereas the association was attenuated to a (non-
the PTSD diagnostic criteria by the age of 18 significant) decrease in the likelihood of being NEET
(OR = 1.75, 95% CI = 1.15; 2.68; Table 3, Panel A). when adjusting for IQ (OR = 0.8, 95% CI = 0.56;
This remained statistically significant when adjusted 1.39) or SES (OR = 0.77, 95% CI = 0.46; 1.29), as
for confounders; the association was slightly well as in the fully adjusted model (OR = 0.66, 95%
enhanced when sex (OR = 1.86, 95% CI = 1.21; CI = 0.40; 1.08).
2.88) was accounted for, and slightly attenuated
when adjusting for IQ (OR = 1.66, 95% CI = 1.08;
Sensitivity analyses
2.54) and SES (OR = 1.52, 95% CI = 1.00; 2.32), as
well as in the fully adjusted model (OR = 1.62, 95% First, we restricted the analytical sample to partic-
CI = 1.04; 2.53). ipants with complete CAST data at age 8 to test if the
replacement of missing CAST data at age 8 impacted
General psychopathology. In young people who our findings. In the restricted sample with complete
reported trauma exposure, those with higher autistic CAST data at age 8 (n = 1,213), we found similar
traits had a significantly increased likelihood of results as in the overall sample used above (see
having higher levels of psychopathology at age 18 Appendix S6, Tables S3, S4). Second, we focused the
(Beta = 3.36, 95% CI = 0.68; 6.04; Table 3, Panel C). analyses predicting PTSD on the past-year

Ó 2025 The Author(s). Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for
Child and Adolescent Mental Health.
1520 Alice M.G. Quinton et al. J Child Psychol Psychiatr 2025; 66(10): 1514–25

(A) (B)

(C) (D)

Figure 2 Univariate and multivariate regressions showing the association between autistic traits and outcome measures at age 18 for the
overall sample (N = 1,504) and trauma-exposed subsample (n = 460). All models are adjusted for the non-independence of twin
observations. Univariate regressions are from analyses unadjusted for potential confounders, and multivariate regression results are
analyses accounting for confounding effects of sex, intelligence quotient and family socio-economic status. Odds ratios and beta-
coefficients are plotted with 95% confidence limits. Figure 2 shows univariate and multivariate analyses between autistic traits and (A)
trauma exposure, PTSD diagnosis and NEET status in the overall sample, (B) ‘p’-factor in the overall sample, (C) PTSD diagnosis and NEET
status in the trauma-exposed subsample, and (D) ‘p’-factor in the trauma-exposed subsample. The dotted line on the axis indicates no
association: 1 for logistic regressions in (A) and (C), and 0 for linear regressions (B) and (D). 95% CI, 95% confidence intervals; Beta, beta-
coefficient; NEET, not in education, employment or training; OR, odds ratio; ‘p’-factor, measure of general psychopathology; PTSD,
post-traumatic stress disorder

diagnosis, rather than the lifetime diagnosis used in childhood, over and above the effects of estab-
above, to provide a clearer timeline for the associa- lished risk factors, such as female sex, lower IQ, and
tions examined. As expected, fewer participants had lower SES. The relationship between autistic traits
a past-year PTSD diagnosis (n = 63) than a lifetime and PTSD persisted in the subsample of young
diagnosis. Focusing on the past-year PTSD diagno- people who reported trauma exposure, suggesting
sis, we found similar results to those obtained in the that the association was not simply explained by an
analyses using the lifetime PTSD diagnosis pre- increased risk of trauma exposure but reflects
sented above (see Appendix S6, Tables S5, S6). greater vulnerability to develop PTSD in
trauma-exposed young people with higher autistic
traits. Beyond PTSD, children with higher autistic
Discussion traits in the overall sample had greater general
Our findings in a longitudinal birth cohort indicate psychopathology (‘p’), but this association was no
an elevated propensity for meeting diagnostic criteria longer significant in the subset of trauma-exposed
for PTSD in young people with higher autistic traits young people after accounting for potential

Ó 2025 The Author(s). Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for
Child and Adolescent Mental Health.
doi:10.1111/jcpp.14163 Childhood autistic traits & PTSD in young adults 1521

confounding effects. Higher autistic traits in child- trauma-exposed subgroup. Rather, the findings
hood also increased the likelihood of the young suggest that trauma-exposed children with higher
people reporting trauma exposure and being NEET autistic traits were more likely to develop PTSD than
at age 18, but these associations were no longer their low-trait peers. As discussed in the introduc-
statistically significant after accounting for low SES. tion, studies largely undertaken in adult samples
Children with higher autistic traits had higher suggest an overlap between established cognitive
rates of self-reported trauma exposure, but this vulnerabilities for PTSD and cognitive differences
association was statistically accounted for by observed in autistic individuals. Further research is
co-occurring disadvantaged socio-economic condi- necessary to identify the specific profile of autistic
tions. This is in line with the strong association of characteristics that increase the likelihood of PTSD.
socio-economic disadvantage with childhood trauma By mapping cognitive risk factors linked with
and adversity (Lacey, Howe, Kelly-Irving, Bartley, & specific autistic traits in young people, we may better
Kelly, 2022). Because studies have consistently understand the mechanisms underlying the
reported higher rates of trauma exposure among observed associations and provide more targeted
autistic versus non-autistic adults (Quinton, Ali, support.
Danese, Happ
e, & Rumball, 2024), we expected that Autistic traits were associated with poorer mental
autistic traits would be associated with reported health outcomes above and beyond PTSD, as mea-
trauma exposure in our study. However, the only sured by a general psychopathology factor, in line
case–control study that found higher rates of trauma with other population-based studies (Lundstr€ om
exposure in autistic versus non-autistic children did et al., 2011). This was observed both in the overall
not take into account SES (Paul, Gallot, Lelouche, sample and in the trauma-exposed subsample to
Bouvard, & Amestoy, 2018). Our findings suggest similar effect. However, after adjusting for potential
that socio-economic factors may help explain the confounders, the association remained significant
relationship between autistic traits and trauma only in the overall sample. The association in the
exposure. It is also possible that the observed overall sample is unsurprising given the wealth of
relationships between autistic traits, trauma expo- research on poorer mental health outcomes for
sure and socio-economic disadvantage may be due autistic people compared with non-autistic people
to overlapping influences, and our observational (Lai et al., 2019; Muniandy, Richdale, Arnold,
design cannot conclusively determine a direct mech- Trollor, & Lawson, 2021), as well as people with
anistic role for socio-economic disadvantage. Future high autistic traits compared with those with low/no
research using experimental designs and/or autistic traits (Stewart et al., 2020). Lack of statis-
repeated measures of socio-economic status (Ludwig tical significance in the adjusted association between
et al., 2011) might be able to disentangle the autistic traits and general psychopathology within
mechanisms underlying the association between the trauma-exposed subsample may reflect lower
autistic traits and trauma exposure. Furthermore, statistical power (the effect sizes in adjusted ana-
the current analyses tested whether autistic traits lyses were similar in the overall sample and trauma-
are associated with exposure to trauma, as narrowly exposed subsample) or a stronger role of the
defined in DSM-5 (American Psychiatric Associa- potential confounders examined. Our research adds
tion, 2013). Research suggests that this narrow to previous findings by indicating that autistic traits
definition may not capture the broader set of might exacerbate the impact of childhood trauma on
negative experiences that young people with high general psychopathology, although more research is
autistic traits also appraise as traumatic and might needed to disentangle causal and non-causal
trigger PTSD in this group (as seen in autistic adults, explanations.
Rumball, Happ
e, & Grey, 2020). The role of subjec- Beyond mental health outcomes, our results
tive appraisal of trauma memories has been demon- showed a small association between autistic traits
strated as a key determinant of trauma-related and NEET status at age 18 in the overall sample.
psychopathology (Baldwin, Coleman, Francis, & However, this association became non-significant
Danese, 2024; Coleman et al., 2024; Danese & when other risk factors were considered, suggesting
Widom, 2020). More clinical and epidemiological that SES and IQ played a role in this relationship. No
research is needed to map and adapt definitions of relationship was observed between autistic traits
trauma for neurodivergent young people, and these and NEET status in the trauma-exposed subsample.
findings highlight the importance of an intersec- There is minimal research on autistic traits and
tional approach to investigating the association functional outcomes in the general population, but
between autistic characteristics and trauma research with college students found autistic traits
exposure. to be associated with academic difficulties, indepen-
Children with higher autistic traits were more dent of an autism diagnosis (McLeod & Ander-
likely to develop PTSD as young adults, regardless son, 2023). Furthermore, higher autistic traits have
of the effects of sex, IQ and SES. Furthermore, the been associated with lower personal income in a
findings did not simply reflect a heightened risk of sample of 2,491 adults (Skylark & Baron-
trauma exposure, as the association held within the Cohen, 2017). We know that finding employment

Ó 2025 The Author(s). Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for
Child and Adolescent Mental Health.
1522 Alice M.G. Quinton et al. J Child Psychol Psychiatr 2025; 66(10): 1514–25

has additional challenges for autistic adults, includ- calculated from the symptoms of 11 mental health
ing being burdened by the need to conceal or conditions over the past year, including PTSD.
‘camouflage’ their autistic traits (Finn, Flower, Although the p-factor and PTSD measures are not
Leong, & Hedley, 2023). Autistic people are under- fully independent, we chose to use the previously
represented in the labor market (Frank et al., 2018), published p-factor measure (Schaefer et al., 2018) to
and previous research has highlighted that poorer ensure comparability with other E-Risk studies and
mental health is one reason that autistic people’s because PTSD likely makes a limited independent
employment (Chen, Leader, Sung, & Leahy, 2015) contribution to the p-factor due to its relatively low
and education (Van Hees, Moyson, & Roeyers, 2015) prevalence and high rates of psychiatric comorbidity.
are disrupted. Given the negative findings, it is Fourth, although we controlled in the analyses for
possible that autistic traits in the general population key potential confounders (sex, IQ, SES), unmea-
do not have independent effects on broad functional sured confounders could provide alternative expla-
outcomes, such as NEET, beyond the detrimental nations for the associations observed. Finally, these
impact of low IQ, low SES and trauma exposure. data were from a sample of twins; it has been argued
that the experiences of twins may not be represen-
tative of those of singletons, and findings may not be
Limitations generalisable to people who are not twins (Røysamb
The first limitations pertain to the measure of & Tambs, 2016). However, the prevalence of trauma
childhood autistic traits. This measure was pro- and psychiatric disorders in E-Risk participants is
duced from CAST scores taken at different ages within the range reported in studies involving non-
(mainly at 8 years, but also at 9 and 12 years when twins (Lewis et al., 2019; Lewis et al., 2021), sup-
age-8 scores were not available) to maximise the porting the wider applicability of findings from this
sample size, but CAST scores at later ages might not cohort.
have provided a reliable proxy for earlier CAST
scores. However, CAST scores at these ages showed
moderate positive correlations in our sample, and Conclusions
previous work in the full TEDS cohort found that This is the first longitudinal study to show a
individual differences in autistic traits measured by significant association between autistic traits in
parent-reported CAST were stable from ages 8– childhood and PTSD diagnosis by early adulthood.
12 years (Holmboe et al., 2014). Only a small Future research needs to elucidate specific mecha-
number of children (n = 33, 2.19%) met the validated nisms of vulnerability to PTSD in young people with
cut-off (≥14) suggestive of potential autism (Williams autistic traits, so that those at risk can be identified
et al., 2005), which is in line with a recent estimate and receive targeted support in which their neuro-
that 1%–2% of the UK population are autistic (NHS divergent characteristics are accommodated.
Digital, 2020; Zeidan et al., 2022). While it is clear
that high autistic traits do not equate to an autism
diagnosis (Lord & Bishop, 2021), utilising a trait- Supporting information
wise approach has proved beneficial for identifying Additional supporting information may be found online
and including those from under-diagnosed groups, in the Supporting Information section at the end of the
such as older adults (Stewart et al., 2023), women article:
(Cardon, McQuarrie, Calton, & Gabrielsen, 2023) Appendix S1. Sample characteristics.
and those from low-income countries (Heys Appendix S2. Selecting the autistic trait measure.
et al., 2018). While useful, our findings may not Appendix S3. Missing data.
generalise to autistic young people, and it will be Appendix S4. Dimensional measures of psychopathol-
essential to replicate our findings in a diagnosed ogy within the E-Risk cohort at age 18.
autistic sample. Second, trauma exposure was Appendix S5. Associations between variables.
measured by retrospective self-report and was Appendix S6. Sensitivity analyses.
recorded without information on the specific timing
of the exposure. It is possible that the onset of
trauma-related psychopathology preceded the autis- Acknowledgements
tic trait measure, and due to overlapping character- The Environmental Risk (E-Risk) Longitudinal Twin
istics of autism and trauma responses in children Study is funded by the UK Medical Research Council
(MRC) (G1002190 and MR/X010791/1). Additional
(Stavropoulos, Bolourian, & Blacher, 2018), the
support was provided by the National Institute of Child
exposure may have impacted the expression of the
Health and Human Development (HD077482) and by
autistic trait score. However, sensitivity analysis the Jacobs Foundation. TEDS is funded by a pro-
focusing on PTSD diagnosis within the last gramme grant by the UK Medical Research Council
12 months (instead of lifetime) showed a very similar (MRC) (MR/V012878/1; MR/M021475/1). Additional
pattern of results. Third, there was a small overlap funding was provided by the US National Institutes of
between the lifetime PTSD diagnosis and general Health (AG046938). Alice Quinton’s PhD studentship is
psychopathology outcomes, as the p-factor was funded by the MRC and King’s College London (MR/

Ó 2025 The Author(s). Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for
Child and Adolescent Mental Health.
doi:10.1111/jcpp.14163 Childhood autistic traits & PTSD in young adults 1523

N013700/1). Professor Helen L. Fisher is part-funded each phase of the E-Risk study. Parents gave written
by the Economic and Social Research Council (ESRC) informed consent, and twins gave assent between 5 and
Centre for Society and Mental Health at King’s College 12 years; twins then gave informed consent at age 18.
London (ES/S012567/1). Professors Francesca Happ
e Ethical approval for TEDS was provided by the King’s
and Andrea Danese are part-funded by the National College London Ethics Committee (reference: PNM/09/
Institute for Health and Care Research (NIHR) Biomed- 10–104). Written informed consent was obtained from
ical Research Centre at South London and Maudsley parents prior to data collection and from twins them-
NHS Foundation Trust and King’s College London selves from age 16 onward. The use of anonymised data
(NIHR203318). For the purpose of Open Access, the for academic purposes did not require additional ethical
author has applied a CC BY public copyright licence to approval.
any Author Accepted Manuscript version arising from
this submission. The views expressed in this publica-
tion are those of the author(s) and not necessarily those Data availability
of the ESRC, MRC, NHS, NIHR, NIH, or King’s College Data for this study came from the Environmental Risk
London. (E-Risk) Longitudinal Twin Study and the Twins Early
The authors are extremely grateful to the E-Risk Development Study (TEDS). Researchers can apply for
study mothers, fathers, twins, and the twins’ teachers managed access to both (E-Risk: [Link]
for their participation. Their thanks go to Professors com/data-access/; TEDS: [Link]
Terrie Moffitt and Avshalom Caspi, the founders of the researchers/teds-data-access-policy).
E-Risk study, and to the E-Risk team for their
dedication, hard work and insights. They acknowledge
the ongoing contribution of the participants in the
Twins Early Development Study (TEDS) and their
Correspondence
Andrea Danese, Social, Genetic and Developmental
families and thank the TEDS study team for their
Psychiatry Centre, Institute of Psychiatry, Psychology
collaboration and support.
and Neuroscience, King’s College London, 16 De
The authors have declared that they have no com-
Crespigny Park, London SE5 8AF, UK; Email: andrea.
peting or potential conflicts of interest.
danese@[Link]. Francesca Happ
e, Social, Genetic
and Developmental Psychiatry Centre, Institute of
Psychiatry, Psychology and Neuroscience, King’s
Ethical considerations College London, 16 De Crespigny Park, London SE5
The Joint South London and Maudsley and the Institute 8AF, UK; Email: [Link]@[Link].
of Psychiatry Research Ethics Committee approved

Key points
• Cross-sectional studies have previously shown positive associations between autistic traits and PTSD in
adults.
• This is the first longitudinal study demonstrating a significant relationship between childhood autistic
traits and meeting PTSD diagnostic criteria by early adulthood.
• Among the trauma-exposed young people, autistic traits were also significantly associated with a
lifetime PTSD diagnosis.
• Increased risk for trauma exposure amongst those with subclinical autistic traits is likely multi-faceted
and may be accounted for by factors such as socio-economic status.
• Findings highlight the need for targeted assessments and evidence-based treatments for PTSD in
children with high autistic traits, and for future research to identify specific mechanisms of vulnerability
to PTSD in autistic individuals.

References traits. Journal of Autism and Developmental Disorders, 38,

American Psychiatric Association. (2013). Diagnostic and 840–847.
statistical manual of mental disorders (DSM-5â) (5th edn). Brewin, C.R., Dalgleish, T., & Joseph, S. (1996). A dual
Author. representation theory of posttraumatic stress disorder.
Baldwin, J.R., Coleman, O., Francis, E.R., & Danese, A. Psychological Review, 103, 670–686.
(2024). Prospective and retrospective measures of child Brosnan, M., & Gavin, J. (2023). The impact of higher levels of
maltreatment and their association with psychopathology: autistic traits on risk of hikikomori (pathological social
A systematic review and meta-analysis. JAMA Psychiatry, withdrawal) in young adults. PLoS One, 18, e0281833.
81, 769–781. Cardon, G., McQuarrie, M., Calton, S., & Gabrielsen, T.P.
Best, C.S., Moffat, V.J., Power, M.J., Owens, D.G.C., & (2023). Similar overall expression, but different profiles, of
Johnstone, E.C. (2008). The boundaries of the cognitive autistic traits, sensory processing, and mental health
phenotype of autism: Theory of mind, central coherence and between young adult males and females. Research in Autism
ambiguous figure perception in young people with autistic Spectrum Disorders, 109, 102263.

Ó 2025 The Author(s). Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for
Child and Adolescent Mental Health.
1524 Alice M.G. Quinton et al. J Child Psychol Psychiatr 2025; 66(10): 1514–25

Chandler, M., & Barrett, R. (2013). UK estimate of young people Kerns, C.M., Newschaffer, C.J., & Berkowitz, S.J. (2015).
not in education, employment or training. London: Office for Traumatic childhood events and autism Spectrum disorder.
National Statistics. Journal of Autism and Developmental Disorders, 45, 3475–
Chen, J.L., Leader, G., Sung, C., & Leahy, M. (2015). Trends in 3486.
employment for individuals with Autism Spectrum Disorder: Kerns, C.M., Rast, J.E., & Shattuck, P.T. (2020). Prevalence
A review of the research literature. Review Journal of Autism and correlates of caregiver-reported mental health condi-
and Developmental Disorders, 2, 115–127. tions in youth with autism spectrum disorder in the United
Coleman, O., Baldwin, J.R., Dalgleish, T., Rose-Clarke, K., States. Journal of Clinical Psychiatry, 82, 11637.
Widom, C.S., & Danese, A. (2024). Research Review: Why do Koenen, K.C., Moffitt, T.E., Poulton, R., Martin, J., & Caspi, A.
prospective and retrospective measures of maltreatment (2007). Early childhood factors associated with the develop-
differ? A narrative review. Journal of Child Psychology and ment of post-traumatic stress disorder: Results from a
Psychiatry, 65, 1662–1677. longitudinal birth cohort. Psychological Medicine, 37, 181–
Dahoun, T., Peel, A., Baldwin, J., Coleman, O., Lewis, S.J., 192.
Wertz, J., . . . & Danese, A. (2024). Genetic and environment Lacey, R.E., Howe, L.D., Kelly-Irving, M., Bartley, M., & Kelly,
influences on childhood victimization: A systematic review Y. (2022). The clustering of adverse childhood experiences in
and meta-analysis. Molecular Psychiatry. [Link] the Avon Longitudinal Study of Parents and Children: Are
10.1038/s41380-024-02868-z gender and poverty important? Journal of Interpersonal
Danese, A., McLaughlin, K.A., Samara, M., & Stover, C.S. Violence, 37, 2218–2241.
(2020). Psychopathology in children exposed to trauma: Lai, M.-C., Kassee, C., Besney, R., Bonato, S., Hull, L., Mandy,
Detection and intervention needed to reduce downstream W., . . . & Ameis, S.H. (2019). Prevalence of co-occurring
burden. British Medical Journal, 371, m3073. mental health diagnoses in the autism population: A
Danese, A., & Widom, C.S. (2020). Objective and subjective systematic review and meta-analysis. The Lancet Psychiatry,
experiences of child maltreatment and their relationships 6, 819–829.
with psychopathology. Nature Human Behaviour, 4, 811– Lewis, S.J., Arseneault, L., Caspi, A., Fisher, H.L., Matthews,
818. T., Moffitt, T.E., . . . & Danese, A. (2019). The epidemiology of
Ehlers, A., & Clark, D.M. (2000). A cognitive model of trauma and post-traumatic stress disorder in a representa-
posttraumatic stress disorder. Behaviour Research and tive cohort of young people in England and Wales. Lancet
Therapy, 38, 319–345. Psychiatry, 6, 247–256.
Finn, M., Flower, R.L., Leong, H.M., & Hedley, D. (2023). ‘If I’m Lewis, S.J., Koenen, K.C., Ambler, A., Arseneault, L., Caspi, A.,
just me, I doubt I’d get the job’: A qualitative exploration of Fisher, H.L., . . . & Danese, A. (2021). Unravelling the
autistic people’s experiences in job interviews. Autism, 27, contribution of complex trauma to psychopathology and
2086–2097. cognitive deficits: A cohort study. The British Journal of
Frank, F., Jablotschkin, M., Arthen, T., Riedel, A., Fangmeier, Psychiatry, 219, 448–455.
T., H€olzel, L.P., & van Tebartz Elst, L. (2018). Education and Lord, C., & Bishop, S.L. (2021). Let’s be clear that “Autism
employment status of adults with autism spectrum disor- spectrum disorder symptoms” are not always related to
ders in Germany – A cross-sectional-survey. BMC Psychia- autism spectrum disorder. American Journal of Psychiatry,
try, 18, 1–10. 178, 680–682.
Golan, O., Haruvi-Lamdan, N., Laor, N., & Horesh, D. (2022). Ludwig, J., Sanbonmatsu, L., Gennetian, L., Adam, E.,
The comorbidity between autism spectrum disorder and Duncan, G.J., Katz, L.F., . . . & McDade, T.W. (2011).
post-traumatic stress disorder is mediated by brooding Neighborhoods, obesity, and diabetes – A randomized social
rumination. Autism, 26, 538–544. experiment. The New England Journal of Medicine, 365,
Happ
e, F., & Frith, U. (2006). The weak coherence account: 1509–1519.
Detail-focused cognitive style in autism spectrum disorders. Lundstr€ om, S., Chang, Z., Kerekes, N., Gumpert, C.H.,
Journal of Autism and Developmental Disorders, 36, 5–25. Rastam, M., Gillberg, C., . . . & Anckars€ ater, H. (2011).
Happ
e, F., & Frith, U. (2020). Annual Research Review: Autistic-like traits and their association with mental health
Looking back to look forward – changes in the concept of problems in two nationwide twin cohorts of children and
autism and implications for future research. Journal of Child adults. Psychological Medicine, 41, 2423–2433.
Psychology and Psychiatry, 61, 218–232. [Link] Marinopoulou, M., Billstedt, E., Wessman, C., Bornehag, C.G.,
10.1111/jcpp.13176 & Hallerb€ ack, M.U. (2025). Association between intellectual
Haruvi-Lamdan, N., Lebendiger, S., Golan, O., & Horesh, D. functioning and autistic traits in the general population of
(2019). Are PTSD and autistic traits related? An examination children. Child Psychiatry and Human Development, 56,
among typically developing Israeli adults. Comprehensive 264–275. [Link]
Psychiatry, 89, 22–27. Mazefsky, C.A., Borue, X., Day, T.N., & Minshew, N.J. (2014).
Heys, M., Gibbons, F., Haworth, E., Medeiros, E., Tumba- Emotion regulation patterns in adolescents with high-
hangphe, K.M., Wickenden, M., . . . & Pellicano, E. (2018). functioning autism spectrum disorder: Comparison to
The estimated prevalence of autism in school-aged children typically developing adolescents and association with psy-
living in rural Nepal using a population-based screening chiatric symptoms. Autism Research, 7, 344–354.
tool. Journal of Autism and Developmental Disorders, 48, McLeod, J.D., & Anderson, E.M. (2023). Autistic traits and
3483–3498. college adjustment. Journal of Autism and Developmental
Holmboe, K., Rijsdijk, F.V., Hallett, V., Happ
e, F., Plomin, R., & Disorders, 53, 3475–3492.
Ronald, A. (2014). Strong genetic influences on the stability Moffitt, T.E., & E-Risk Study Team. (2002). Teen-aged mothers
of autistic traits in childhood. Journal of the American in contemporary Britain. Journal of Child Psychology and
Academy of Child and Adolescent Psychiatry, 53, 221–230. Psychiatry, 43, 727–742.
Hoover, D.W., & Kaufman, J. (2018). Adverse childhood Muniandy, M., Richdale, A.L., Arnold, S.R.C., Trollor, J.N., &
experiences in children with autism spectrum disorder. Lawson, L.P. (2021). Inter-relationships between trait resil-
Current Opinion in Psychiatry, 31, 128–132. ience, coping strategies, and mental health outcomes in
Huckins, L.M., Johnson, J.S., Cancelmo, L., Diab, O., autistic adults. Autism Research, 14, 2156–2168.
Schaffer, J., Cahn, L., . . . & Feder, A. (2021). Polygenic Napolitano, A., Schiavi, S., La Rosa, P., Rossi-Espagnet, M.C.,
regulation of PTSD severity and outcomes among World Petrillo, S., Bottino, F., . . . & Vicari, S. (2022). Sex differences
Trade Center responders. medRxiv. [Link] in autism spectrum disorder: Diagnostic, neurobiological,
1101/2020.12.06.20244772 and behavioral features. Frontiers in Psychiatry, 13, 889636.

Ó 2025 The Author(s). Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for
Child and Adolescent Mental Health.
doi:10.1111/jcpp.14163 Childhood autistic traits & PTSD in young adults 1525

NHS Digital. (2020). Autism waiting time statistics. NDRS. Schaefer, J.D., Moffitt, T.E., Arseneault, L., Danese, A., Fisher,
Available from: [Link] H.L., Houts, R., . . . & Caspi, A. (2018). Adolescent victimi-
information/publications/statistical/autism-statistics/q1- zation and early-adult psychopathology: Approaching
april-to-june-2020-21/data-quality-copy causal inference using a longitudinal twin study to rule out
Odgers, C.L., Caspi, A., Bates, C.J., Sampson, R.J., & Moffitt, noncausal explanations. Clinical Psychological Science, 6,
T.E. (2012). Systematic social observation of children’s 352–371.
neighborhoods using Google Street View: A reliable and Scott, F.J., Baron-Cohen, S., Bolton, P., & Brayne, C. (2002).
cost-effective method. Journal of Child Psychology and The CAST (Childhood asperger syndrome test): Preliminary
Psychiatry, 53, 1009–1017. development of a UK screen for mainstream primary-school-
Paul, A., Gallot, C., Lelouche, C., Bouvard, M.P., & Amestoy, A. age children. Autism, 6, 9–31.
(2018). Victimisation in a French population of children and Skylark, W.J., & Baron-Cohen, S. (2017). Initial evidence that
youths with autism spectrum disorder: A case control study. non-clinical autistic traits are associated with lower income.
Child and Adolescent Psychiatry and Mental Health, 12, 48. Molecular Autism, 8, 61.
Peel, A.J., Purves, K.L., Baldwin, J.R., Breen, G., Coleman, Stavropoulos, K.K.-M., Bolourian, Y., & Blacher, J. (2018).
J.R.I., Pingault, J.B., . . . & Eley, T.C. (2022). Genetic and Differential diagnosis of autism spectrum disorder and post
early environmental predictors of adulthood self-reports of traumatic stress disorder: Two clinical cases. Journal of
trauma. British Journal of Psychiatry, 221, 613–620. Clinical Medicine, 7, 71.
Quinton, A.M.G., Ali, D., Danese, A., Happ
e, F., & Rumball, F. Stewart, G.R., Corbett, A., Ballard, C., Creese, B., Aarsland,
(2024). The assessment and treatment of post-traumatic D., Hampshire, A., . . . & Happ
e, F. (2023). The cognitive
stress disorder in autistic people: A systematic review. profile of middle-aged and older adults with high vs. low
Review Journal of Autism and Developmental Disorders. autistic traits. Autism Research, 16, 429–440.
[Link] Stewart, G.R., Corbett, A., Ballard, C., Creese, B., Aarsland,
Ratanatharathorn, A., Koenen, K.C., Chibnik, L.B., Weisskopf, D., Hampshire, A., . . . & Happ
e, F. (2020). Sleep problems
M.G., Rich-Edwards, J.W., & Roberts, A.L. (2021). Polygenic and mental health difficulties in older adults who endorse
risk for autism, attention-deficit hyperactivity disorder, high autistic traits. Research in Autism Spectrum Disorders,
schizophrenia, major depressive disorder, and neuroticism 77, 101633.
is associated with the experience of childhood abuse. Trouton, A., Spinath, F.M., & Plomin, R. (2002). Twins early
Molecular Psychiatry, 26, 1696–1705. development study (TEDS): A multivariate, longitudinal
Reuben, A., Sugden, K., Arseneault, L., Corcoran, D.L., genetic investigation of language, cognition and behavior
Danese, A., Fisher, H.L., . . . & Caspi, A. (2020). Association problems in childhood. Twin Research, 5, 444–448.
of neighborhood disadvantage in childhood with DNA Trzesniewski, K.H., Moffitt, T.E., Caspi, A., Taylor, A., &
methylation in young adulthood. The Journal of the Amer- Maughan, B. (2006). Revisiting the association between
ican Medical Association (JAMA) Network Open, 3, e206095. reading achievement and antisocial behavior: New evidence
Robins, L., Cottler, L., Bucholz, K., & Compton, W. (1995). of an environmental explanation from a twin study. Child
Diagnostic Interview Schedule for DSM-IV. St. Louis: Wash- Development, 77, 72–88.
ington University School of Medicine. Van Hees, V., Moyson, T., & Roeyers, H. (2015). Higher
Robinson, E.B., Koenen, K.C., McCormick, M.C., Munir, K., education experiences of students with autism spectrum
Hallett, V., Happ
e, F., . . . & Ronald, A. (2011). Evidence that disorder: Challenges, benefits and support needs. Journal of
autistic traits show the same etiology in the general Autism and Developmental Disorders, 45, 1673–1688.
population and at the quantitative extremes (5%, 2.5%, Wang, Y., Zhang, Y., Liu, L., Cui, J., Wang, J., Shum, D.H.K.,
and 1%). Archives of General Psychiatry, 68, 1113–1121. . . . & Chan, R.C.K. (2017). A meta-analysis of working
Røysamb, E., & Tambs, K. (2016). The beauty, logic and memory impairments in autism spectrum disorders. Neuro-
limitations of twin studies. Norsk Epidemiologi, 26, 35–46. psychology Review, 27, 46–61.
Rumball, F., Antal, K., Happ
e, F., & Grey, N. (2021). Co- Warrier, V., & Baron-Cohen, S. (2021). Childhood trauma, life-
occurring mental health symptoms and cognitive processes time self-harm, and suicidal behaviour and ideation are
in trauma-exposed ASD adults. Research in Developmental associated with polygenic scores for autism. Molecular
Disabilities, 110, 103836. doi: 10.1016/[Link].2020. Psychiatry, 26, 1670–1684.
103836. Weiland, R.F., Polderman, T.J., Hoekstra, R.A., Smit, D.J., &
Rumball, F., Brook, L., Happ
e, F., & Karl, A. (2021). Begeer, S. (2020). The Dutch Sensory Perception Quotient-
Heightened risk of posttraumatic stress disorder in adults Short in adults with and without autism. Autism, 24, 2071–
with autism spectrum disorder: The role of cumulative 2080.
trauma and memory deficits. Research in Developmental Williams, J., Scott, F., Stott, C., Allison, C., Bolton, P., Baron-
Disabilities, 110, 103848. Cohen, S., & Brayne, C. (2005). The CAST (Childhood
Rumball, F., Happ
e, F., & Grey, N. (2020). Experience of Asperger Syndrome Test): Test accuracy. Autism, 9, 45–68.
trauma and PTSD symptoms in autistic adults: Risk of PTSD Williams, R.L. (2000). A note on robust variance estimation for
development following DSM-5 and non-DSM-5 traumatic life cluster-correlated data. Biometrics, 56, 645–646.
events. Autism Research, 13, 2122–2132. Zeidan, J., Fombonne, E., Scorah, J., Ibrahim, A., Durkin,
Sallis, H.M., Croft, J., Havdahl, A., Jones, H.J., Dunn, E.C., M.S., Saxena, S., . . . & Elsabbagh, M. (2022). Global
Smith, G.D., . . . & Munaf o, M.R. (2021). Genetic liability to prevalence of autism: A systematic review update. Autism
schizophrenia is associated with exposure to traumatic Research, 15, 778–790.
events in childhood. Psychological Medicine, 51, 1814–1821.
Sattler, J.M. (1988). In 3rd edn (Ed.), Assessment of children. Accepted for publication: 25 February 2025
Jerome M. Sattler.

Ó 2025 The Author(s). Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for
Child and Adolescent Mental Health.