Timing of childhood adversity and cognitive profiles in adults with autism spectrum disorder

Abstract

Background

This study examined whether the timing of adverse childhood experiences (ACEs) is associated with differences in intellectual ability in adults with autism spectrum disorder (ASD) without intellectual disability.

Methods

A total of 161 adults with ASD were categorized into three groups based on ACE onset: no ACEs (Group 0), early childhood ACEs (≤ 10 years; Group 1), and adolescent ACEs (11–18 years; Group 2). Intellectual ability was assessed using the Wechsler Adult Intelligence Scale-Fourth Edition (WAIS-IV).

Results

Findings revealed that Group 1 outperformed Group 2 on the picture completion and figure weights subtests, suggesting that early ACE exposure may be linked to distinct cognitive profiles in ASD. To account for the higher severity of post-traumatic stress disorder (PTSD) symptoms, including hypervigilance, in Group 1 compared to Group 2, we conducted multivariate analyses of covariance and logistic regression to control for potential confounders such as years of education and PTSD symptom severity, given that hypervigilance may influence cognitive task performance. However, the observed differences remained significant.

Conclusion

These results suggest a potential association between the timing of ACE exposure and specific cognitive subdomains in adults with ASD. Furthermore, WAIS-IV subtests, particularly picture completion and figure weights, may serve as potential markers for identifying cognitive adaptations associated with early adversity in ASD individuals.

Clinical Trial Registration

Since this study was not a clinical trial, this section does not apply.

Background

Autism Spectrum Disorder (ASD) is a neurodevelopmental disorder characterized by persistent difficulties in social communication and restricted, repetitive patterns of behavior [1]. It usually manifests in early childhood and is diagnosed via the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5) diagnostic criteria [1]. The intellectual and cognitive abilities of individuals with ASD vary widely; some individuals have co-occurring intellectual disabilities, while others demonstrate average or above-average intelligence [2]. Additionally, many individuals with ASD experience comorbid psychiatric conditions, including anxiety and attention problems [3]. Given this cognitive heterogeneity, various factors beyond the core ASD traits influence the intellectual and developmental outcomes. While genetic and neurobiological influences play a dominant role, environmental factors, including childhood adversity, may shape the cognitive development and functional outcomes of individuals with ASD.

Adverse childhood experiences (ACEs), including abuse, neglect, and household dysfunction, are well-established risk factors for long-term health and developmental problems [4–6]. Previous studies have demonstrated that the accumulation of ACEs is correlated with adverse physical, mental, and cognitive outcomes in later life [5, 7-9]. Considering their substantial impact, increasing attention has been paid to their role in neurodevelopmental disorders, particularly ASD. Emerging evidence suggested that individuals with ASD were disproportionately exposed to ACEs, which compounded their developmental challenges [10-12]. National survey data indicated that children with ASD were significantly more likely to experience multiple ACEs, even when controlled for socioeconomic factors [11]. Such early-life adversity has been linked to poorer health and mental health outcomes among those with ASD [12], which reinforces the notion that childhood trauma may be a crucial environmental contributor to the life-course trajectory of individuals with ASD.

Prior research has established a link between ACEs and adverse outcomes in ASD, and has primarily focused on cognitive function [13] and general well-being [12]. According to Hawkins et al., ACEs negatively impact brain structure and neurocognitive function, which leads to impairments in executive function, memory, learning, and language development [13]. Furthermore, neuroimaging studies have suggested that autistic individuals with a history of severe ACEs exhibit disruptions in frontal white matter integrity, which indicates that early trauma alters neurodevelopmental pathways [14]. Moreover, past studies demonstrated that distinct ACE types had differential effects on cognitive function [15–17]. A U.S. study found that exposure to deprivational ACEs during childhood was associated with a decline in short-term and long-term memory as well as executive function [15]. Additionally, post-traumatic stress disorder (PTSD) symptoms after severe ACE exposure were linked to a decline in memory, attention, and executive function [8, 18–20].

However, responses to adversity are not uniformly negative. Emerging literature suggests that the effects of ACEs may be heterogeneous and may include adaptive or resilience-related processes [21, 22]. Trauma exposure, particularly when accompanied by heightened threat sensitivity, may be associated with increased vigilance, enhanced perceptual monitoring, or heightened detection of environmental cues [23, 24]. Such adaptations may influence performance in specific cognitive domains requiring visual scanning or attention to detail [25, 26]. This dual nature of ACE effects raises the possibility that early adversity may impair certain cognitive domains while enhancing others, depending on the function and context assessment.

Despite these advances, the timing of ACE exposure in people with ASD remains under-explored. Most prior research has focused on whether individuals with ASD have experienced ACEs and their cumulative impact. However, few studies have examined when these adversities occurred. Developmental science suggests that their timing is crucial, as adverse events during sensitive periods of brain development may have further pervasive and lasting effects on neural and cognitive function [27, 28]. In parallel, however, trauma exposure—especially in early life—has also been linked to heightened perceptual alertness and hypervigilance, which may lead to selective enhancements in tasks involving visual scanning or detail detection. Schalinski et al. reported that specific types of early-life neglect, such as physical and emotional neglect at age 5 and 4–5, respectively, were linked to increased dissociative symptoms in adulthood [29]. Similarly, emotional neglect at ages 8–9 was associated with heightened depressive symptoms. Furthermore, Dunn et al. reported that individuals first exposed to child maltreatment between 0 and 5 years exhibited more severe depressive and PTSD symptoms in adulthood than those exposed between 6 and 18 years [30]. These findings suggest that earlier exposure to adversity may have a more detrimental impact on psychological and cognitive development than later exposure.

This study aimed to address this gap and examine how the timing of ACE exposure is associated with specific cognitive domains in adults with ASD. While early-life ACEs have been associated with cognitive impairments in areas such as memory and executive function, some trauma-related adaptations—particularly in perceptual vigilance—may enhance performance in specific subdomains. Based on this dual framework, we hypothesized that individuals with ASD who experienced ACEs in early childhood would show lower overall cognitive performance compared to those exposed later or not at all, but relatively better performance in tasks that require visual discrimination and perceptual attention, such as the Picture Completion subtest of the WAIS-IV. We aimed to clarify how the timing of ACE exposure affected cognitive function and advance our understanding of how environmental influences shape developmental outcomes in ASD. Our findings may provide essential insights for tailoring targeted interventions and support strategies, and emphasize the importance of early identification and intervention for individuals with ASD who have experienced childhood adversity.

Methods

Participants

We recruited high-functioning adults with ASD aged 18 years and older between September 2019 and December 2023 from patients who received care at the Department of Psychiatry, Nara Medical University Hospital, either as outpatients or inpatients. Initial ASD diagnoses were conducted by at least two trained psychiatrists or psychologists based on DSM-5 [1]. To ensure diagnostic accuracy, only individuals who met the ASD classification criteria via the Autism Diagnostic Observation Schedule-2 (ADOS-2), Module 4 were included [31, 32]. Additionally, the Mini-International Neuropsychiatric Interview (M.I.N.I.) was administered to screen for psychiatric comorbidities [33]. Individuals with major psychiatric disorders or suicidal ideation were excluded.

Medical history data revealed that among the participants with ASD, nine, four, one, three, two, and one had epilepsy, diabetes, hyperthyroidism, Meniere’s disease, hepatitis (one with concurrent hyperthyroidism), and a history of pancreatitis, respectively. To minimize the potential impact of intellectual disability on cognitive assessments, participants with an estimated Full-Scale Intelligence Quotient (FSIQ) < 70 were excluded. This ensured that the study focused on high-functioning individuals with ASD without co-occurring intellectual disabilities.

The study was approved by the Nara Medical University Ethics Committee, and written informed consent was obtained from all the participants. Educational history was also obtained through a standardized questionnaire. To explore the relationship between the timing of ACE initiation and intellectual ability, participants were categorized into three groups: Group 0 (n = 48), individuals who reported no ACEs before age 18, Group 1 (n = 94), those who reported ACEs exposure before age 11, and Group 2 (n = 19), individuals who first experienced ACEs between ages 11–18. These groupings allowed for an in-depth examination of how early versus later exposure to childhood adversity differentially influenced cognitive outcomes in ASD.

Psychological examinations

All psychological assessments were conducted by an experienced psychologist. The M.I.N.I. was used to screen for psychiatric disorders [33]. To confirm ASD diagnoses, Module 4 of the ADOS-2, a semi-structured interview, was administered [31, 32]. Previous research demonstrated that it was highly effective in distinguishing adults with ASD from individuals with neurotypical development [31, 32]. Assessments were conducted by a qualified psychologist, who was certified to administer the ADOS-2 for research purposes.

Additionally, the Autism Spectrum Quotient-Japanese version (AQ-J) was used to assess ASD symptoms [34, 35]. This 50-item questionnaire evaluated five core domains: social skills, attention switching, attention to detail, communication, and imagination. It served as a self-assessment tool, which allowed participants to report their perception of ASD-related traits.

To evaluate ACEs, a 9-item ACE questionnaire was administered [36]. The first three items assessed physical, psychological, and sexual abuse, while the next five items measured family dysfunction, which included parental substance abuse, domestic violence, family mental illness, parental separation, and criminal history. The final item addressed childhood neglect. In addition, participants were asked to report the age at which their ACEs first occurred (≤ 10 years or ≥ 11 years), and we categorized ACE exposure based on the two developmental periods to examine the impact of early versus later adversity on cognitive outcomes.

Wechsler adult intelligence scale-fourth edition

The full-scale intelligence quotient was assessed via the Wechsler Adult Intelligence Scale-Fourth Edition (WAIS-IV), a standardized, comprehensive clinical tool that measures intelligence in individuals aged 16 years 0 months to 90 years 11 months [37]. It comprised 15 subtests, with 10 core subtests used to calculate five composite scores: the Full-Scale Intelligence Quotient (FSIQ), Verbal Comprehension Index (VCI), Perceptual Reasoning Index (PRI), Working Memory Index (WMI), and Processing Speed Index (PSI). Auxiliary subtests were administered when closer examination was required or when a core subtest was deemed invalid. All 15 subtests were completed by participants to obtain a comprehensive assessment of their cognitive functioning.

Impact of Event Scale-Revised (IES-R-J)

To measure the severity of post-traumatic stress reactions, the IES-R-J was utilized [38]. The IES-R-J is commonly used to evaluate self-reported post-traumatic stress symptoms, such as intrusion (eight items), hyperarousal (six items), and avoidance (eight items), after a traumatic experience. Participants rated each questionnaire item regarding one past life event that forced them to feel great discomfort on a 5-point Likert scale that ranged from 0 (not at all) to 4 (extremely). The IES-R-J total and subscale (intrusion, hyperarousal, and avoidance) scores were calculated accordingly, and higher scores reflected greater severity. Since the participants may feel unwell during the examination, medical staff were ready to carefully monitor their condition.

Statistical analyses

First, chi-square tests and one-way analyses of variance (ANOVAs) were conducted to examine the potential differences in gender, age, years of education, general intellectual ability (WAIS-IV FSIQ), ASD symptoms (AQ-J and ADOS-2 scores), and PTSD symptoms (IES-R total scores) among the three groups categorized by ACE initiation timing.

Subsequently, multivariate analysis of covariance (MANCOVA) and logistic regression analyses were performed to investigate the relationship between ACE initiation timing and intellectual ability. Considering the well-documented positive correlation between years of education and cognitive ability [39, 40], years of education were included as a covariate in all the MANCOVA and logistic regression models to control for their potential influence on the WAIS-IV scores.

In the MANCOVA models, the WAIS-IV composite and subtest scores and ACE initiation timing were treated as dependent and independent variables, respectively. This approach allowed us to assess whether significant differences in intellectual ability profiles emerged based on the timing of childhood adversity exposure. Bonferroni correction was used for post hoc comparisons within each WAIS-IV index domain. Specifically, the four index domains—VCI, PRI, WMI, and PSI—were each treated as separate families. For each family, multiple comparisons were adjusted based on the number of core and supplemental subtests analyzed within that index (e.g., PRI: Block Design, Matrix Reasoning, Visual Puzzles, as well as supplemental subtests such as Figure Weights and Picture Completion). All p-values reported for post hoc comparisons were Bonferroni-adjusted.

Additionally, logistic regression analyses were conducted to determine whether the WAIS-IV score could predict ACE initiation timing. In these models, ACE initiation timing and the WAIS-IV composite and subtest scores were used as the dependent and independent predictors, respectively. To further refine our analyses, the IES-R total scores (as an index of PTSD symptom severity) were included as additional covariates in the subsequent MANCOVA and logistic regression models. These analyses were not intended to imply a causal direction from cognitive outcomes to ACE exposure, but rather to identify WAIS-IV subtests that may serve as potential markers of early-life adversity among individuals with ASD. Accordingly, the findings from logistic regression should be interpreted as statistical associations that complement the group comparisons conducted using MANCOVA.

A p-value of < 0.05 was considered statistically significant. Statistical analyses were conducted via SPSS Statistics version 27 (IBM Inc., Armonk, NY).

Results

Demographic and clinical data

The chi-squared test and one-way ANOVA revealed no significant differences among the three groups for gender (χ²(2) = 1.360, p = 0.507), age (F(2,158) = 0.193, p = 0.824), years of education (F(2,158) = 2.669, p = 0.072), the WAIS-IV Full-Scale IQ (FSIQ) (F(2,158) = 1.619, p = 0.201), or current ASD symptoms (AQ-J total score: F(2,158) = 1.165, p = 0.315; ADOS-2 total score: F(2,158) = 0.246, p = 0.782). However, a significant difference was found in the IES-R total scores; individuals exposed to ACEs at an earlier age had higher PTSD symptoms (F(2,158) = 7.118, p = 0.001) (Table 1).

Table 1.

Demographic and Clinical Characteristics by Timing of ACEs Initiation

	Group 0 (No ACEs)	Group 1 (ACEs < 11 years)	Group 2 (ACEs 11–18 years)
				F-value	p-value
Age (years)	27.10(6.23)	27.83(7.22)	27.37(5.12)	0.193	0.824
Sex (Male/Female)	34/14	62/32	15/4	1.360	0.507
Years of education	14.71(2.13)	13.83(2.23)	13.79(2.42)	2.669	0.072
WAIS-Ⅳ (FSIQ)	104.10(16.18)	99.10(15.75)	101.95(16.02)	1.619	0.201
AQ-J total score	30.92(6.57)	30.91(7.22)	33.58(8.09)	1.165	0.315
ADOS-2 total score	14.96(2.15)	14.65(2.75)	14.79(1.87)	0.246	0.782
IES-R-J total score	28.25(21.12)	42.41(22.54)	32.37(20.65)	7.118	0.001**

ACEs, adverse childhood experiences; SD, Standard Deviation; WAIS-Ⅳ, Wechsler Adult Intelligence Scale-Fourth Edition; FSIQ, Full-Scale Intelligence Quotient; AQ-J, Autism Spectrum Quotient-Japanese version; ADOS-2, Autism Diagnostic Observation Schedule-Second Edition; IES-R-J, the Japanese-language version of Impact of Event Scale-Revised

Data are presented as adjusted Mean (Standard Deviation). *p < 0.05 indicates statistical significance

Cognitive profiles and ace initiation timing

A MANCOVA was conducted using the WAIS-IV index scores as dependent variables, ACE initiation timing as an independent variable, and years of education as a covariate. Results revealed no significant differences among the three groups in the FSIQ or four WAIS-IV index scores (p > 0.05 for all comparisons) (Table 2). Additional MANCOVA using WAIS-IV subtest age-scaled scores as dependent variables revealed that among the Perceptual Reasoning subtests, only picture completion exhibited a significant group difference (F(2,154) = 3.363, p = 0.037). Post-hoc pairwise comparisons indicated that Group 1 (ACEs before age 11) had significantly higher scores than Group 2 (ACEs between ages 11–18) (p = 0.025) (Fig. 1a). A logistic regression analysis was conducted to examine the relationship between the WAIS-IV subtest scores and ACE initiation timing. The odds ratio for Group 2 relative to Group 1 and for Group 1 relative to Group 0 was 0.754 (95% CI: 0.604–0.941, p = 0.013) and 1.171 (95% CI: 1.004–1.365, p = 0.044), which indicated a statistically significant association between the ACE timing and lower picture completion scores (Fig. 1b, c).

Table 2.

Comparison of Intellectual Ability by ACE Initiation Timing (Controlling for Years of Education)

Index/subset	Multivariate	df	df error	F	η2	Group 0 (No ACEs)	Group 1 (ACEs < 11 years)	Group 2 (ACEs 11–18 years)
		10	306	0.364	0.012	Wilks‘Λ=0.977 Roy’s largest root = 0.018 p=0.961 p = 0.741
	FSIQ	2	157	0.588	0.007	102.17 (2.10)	99.90 (1.49)	102.87 (3.30)
	VCI	2	157	0.294	0.004	106.84 (1.92)	105.53 (1.36)	107.66 (3.02)
	PRI	2	157	0.102	0.001	101.26 (2.22)	100.52 (1.57)	102.10 (3.49)
	WMI	2	157	0.278	0.004	101.49 (2.51)	99.34 (1.78)	101.19 (3.95)
	PSI	2	157	1.151	0.014	94.85 (2.25)	91.56 (1.60)	96.20 (3.54)
Subset comparison
VCI Subtest	Multivariate	8	308	0.813	0.021	Wilks‘Λ=0.959 Roy’s largest root = 0.034 p=0.591 p = 0.261
	SI	2	157	1.365	0.017	10.88 (0.37)	10.70 (0.27)	11.76 (0.59)
	VC	2	157	0.217	0.003	11.74 (0.42)	11.45 (0.30)	11.78 (0.66)
	IN	2	157	0.140	0.002	11.06 (0.41)	10.94 (0.29)	10.65 (0.65)
	COM	2	157	0.620	0.008	11.63 (0.40)	11.09 (0.29)	11.06 (0.64)
PRI Subtest	Multivariate	10	300	1.825	0.057	Wilks‘Λ=0.889 Roy’s largest root = 0.077 p=0.056 p = 0.046
	BD	2	154	0.421	0.005	10.37 (0.48)	9.84 (0.33)	10.16 (0.73)
	MR	2	154	0.029	< 0.001	10.46 (0.46)	10.60 (0.31)	10.54 (0.70)
	VP	2	154	0.362	0.005	9.87 (0.47)	10.05 (0.33)	10.60 (0.72)
	FW	2	154	1.266	0.016	10.92 (0.48)	10.45 (0.33)	9.52 (0.73)
	PC	2	154	3.363*	0.042	8.76 (0.41)	9.65 (0.28)	8.09 (0.63)
WMI Subtest	Multivariate	6	310	1.465	0.028	Wilks‘Λ=0.946 Roy’s largest root = 0.054 p = 0.190 p = 0.042
	DS	2	157	0.264	0.003	10.34 (0.50)	10.08 (0.36)	10.67 (0.79)
	AR	2	157	0.383	0.005	10.27 (0.46)	9.77 (0.33)	9.84 (0.73)
	LN	2	157	2.048	0.025	9.18 (0.44)	9.87 (0.31)	10.82 (0.70)
PSI Subtest	Multivariate	6	308	0.891	0.017	Wilks‘Λ=0.966 Roy’s largest root = 0.027 p = 0.502 p = 0.245
	SS	2	156	1.438	0.018	9.05 (0.45)	8.62 (0.31)	9.88 (0.70)
	CO	2	156	0.744	0.009	9.20 (0.46)	8.56 (0.32)	9.06 (0.71)
	CA	2	156	1.809	0.023	8.23 (0.36)	7.96 (0.25)	9.12 (0.56)

ACEs, adverse childhood experiences; FSIQ, Full Scale Intelligence Quotient; VCI, Verbal Comprehension Index; PRI, Perceptual Reasoning Index; WMI, Working Memory Index; PSI, Processing Speed Index; BD, Block Design; SI, Similarities; DS, Digit Span; MR, Matrix Reasoning; VC, Vocabulary; AR, Arithmetic; SS, Symbol Search; VP, Visual Puzzles; IN, Information; CO, Coding; LN, Letter Number sequencing; FW, Figure Weights; COM, Comprehension; CA, Cancellation; PC, Picture Completion

Data are presented as adjusted Mean (Standard Error). *p < 0.05 indicates statistical significance

Fig. 1.

Relationship between Timing of Adverse Childhood Experiences (ACEs) and WAIS-IV Picture Completion Scores in Adults with Autism Spectrum Disorder (a) WAIS-IV Picture Completion subtest scores among adults with autism spectrum disorder (ASD), stratified by the timing of ACE initiation: Group 0 (no ACEs), Group 1 (ACEs initiated before age 11), and Group 2 (ACEs initiated between ages 11–18). Group 1 demonstrated significantly higher scores compared with Group 2 (p = 0.037). (b, c) Logistic regression results show that Group 1 had higher Picture Completion scores than both Group 0 (p = 0.044) and Group 2 (p = 0.013). Error bars indicate standard errors in (a) and credible interval in (b, c). *p < 0.05

Covariate Control: PTSD symptoms

While MANCOVA showed no significant differences in the FSIQ and composite scores across the groups as well, picture completion remained significantly different among the groups (F(2,153) = 4.247, p = 0.016) even when controlled for the IES-R total scores as an additional covariate (Table 3). Further pairwise comparisons revealed that Group 1 exhibited significantly higher picture completion scores compared with both groups 0 (p = 0.037) and Group 2 (p = 0.014) even when controlled for the IES-R total scores (Fig. 2a). These findings suggest that individuals who experienced ACEs before age 11 may develop heightened perceptual vigilance, potentially as a response to early environmental instability. The significant difference between Groups 1 and 2 reinforced the notion that earlier ACE exposure may lead to stronger perceptual reasoning skills compared with later exposure. To further investigate the relationships between ACE initiation timing and the WAIS-IV subtest performance, an additional logistic regression analysis was conducted. Results demonstrated that Group 1 had significantly higher odds of outperforming Group 0 in picture completion, with an odds ratio of 1.196 (95% CI: 1.020–1.403, p = 0.028) (Fig. 2b). Additionally, significant associations were observed for both figure weights (OR = 0.768, 95% CI: 0.596–0.990, p = 0.042) and picture completion (OR = 0.745, 95% CI: 0.591–0.939, p = 0.013) between Groups 2 and 1 (Fig. 2c). These findings further support the notion that cognitive performance varied depending on the reported timing of ACE exposure, particularly in perceptual reasoning and executive function, independent of PTSD symptoms.

Table 3.

Comparison of Intellectual Ability by ACE Initiation Timing (Controlling for Years of Education and IES-R-J Total Score)

Index/subset	Multivariate	df	df error	F	η2	Group 0 (No ACEs)	Group 1 (ACEs < 11 years)	Group 2 (ACEs 11–18 years)
		10	304	0.561	0.018	Wilks‘Λ=0.964 Roy’s largest root = 0.031 p=0.845 p = 0.457
	FSIQ	2	156	0.192	0.002	101.43 (2.12)	100.37 (1.50)	102.40 (3.29)
	VCI	2	156	0.076	0.001	106.18 (1.95)	105.94 (1.38)	107.24 (3.01)
	PRI	2	156	0.079	0.001	100.18 (2.23)	101.21 (1.57)	101.41 (3.45)
	WMI	2	156	0.065	0.001	100.82 (2.55)	99.77 (1.80)	100.76 (3.95)
	PSI	2	156	1.170	0.015	94.97 (2.30)	91.48 (1.63)	96.28 (3.57)
Subset comparison
VCI Subtest	Multivariate	8	306	0.734	0.019	Wilks‘Λ=0.963 Roy’s largest root = 0.037 p=0.661 p = 0.232
	SI	2	156	1.142	0.014	10.81 (0.38)	10.74 (0.27)	11.71 (0.59)
	VC	2	156	0.027	< 0.001	11.60 (0.43)	11.53 (0.30)	11.69 (0.66)
	IN	2	156	0.198	0.003	10.93 (0.42)	11.02 (0.30)	10.57 (0.65)
	COM	2	156	0.225	0.003	11.41 (0.40)	11.23 (0.29)	10.92 (0.63)
PRI Subtest	Multivariate	10	298	1.801	0.057	Wilks‘Λ=0.889 Roy’s largest root = 0.085 p=0.060 p = 0.031
	BD	2	153	0.024	< 0.001	10.12 (0.48)	10.00 (0.33)	9.99 (0.72)
	MR	2	153	0.201	0.003	10.32 (0.47)	10.68 (0.32)	10.45 (0.70)
	VP	2	153	0.467	0.006	9.72 (0.48)	10.14 (0.33)	10.50 (0.72)
	FW	2	153	1.352	0.017	10.68 (0.48)	10.59 (0.33)	9.36 (0.72)
	PC	2	153	4.247*	0.053	8.63 (0.42)	9.73 (0.29)	8.00 (0.63)
WMI Subtest	Multivariate	6	308	1.337	0.025	Wilks‘Λ=0.950 Roy’s largest root = 0.052 p = 0.240 p = 0.049
	DS	2	156	0.102	0.001	10.17 (0.51)	10.19 (0.36)	10.56 (0.79)
	AR	2	156	0.245	0.003	10.21 (0.47)	9.81 (0.34)	9.80 (0.73)
	LN	2	156	2.210	0.028	9.11 (0.45)	9.92 (0.32)	10.77 (0.70)
PSI Subtest	Multivariate	6	306	0.968	0.019	Wilks‘Λ=0.963 Roy’s largest root = 0.025 p = 0.447 p = 0.276
	SS	2	155	1.210	0.015	9.00 (0.46)	8.65 (0.32)	9.85 (0.70)
	CO	2	155	1.122	0.014	9.31 (0.46)	8.49 (0.32)	9.13 (0.71)
	CA	2	155	1.839	0.023	8.26 (0.37)	7.95 (0.26)	9.14 (0.56)

ACEs, adverse childhood experiences; IES-R-J, the Japanese-language version of Impact of Event Scale-Revised; FSIQ, Full Scale Intelligence Quotient; VCI, Verbal Comprehension Index; PRI, Perceptual Reasoning Index; WMI, Working Memory Index; PSI, Processing Speed Index; BD, Block Design; SI, Similarities; DS, Digit Span; MR, Matrix Reasoning; VC, Vocabulary; AR, Arithmetic; SS, Symbol Search; VP, Visual Puzzles; IN, Information; CO, Coding; LN, Letter Number sequencing; FW, Figure Weights; COM, Comprehension; CA, Cancellation; PC, Picture Completion

Data are presented as adjusted Mean (Standard Error). *p < 0.05 indicates statistical significance

Fig. 2.

Effects of ACE Timing on WAIS-IV Picture Completion and Figure Weights Subtest Performance Adjusted for PTSD Symptoms (a) WAIS-IV Picture Completion and Figure Weights subtest scores among adults with autism spectrum disorder grouped by the timing of ACE initiation (Group 0: no ACEs, Group 1: ACEs before age 11, Group 2: ACEs at ages 11–18), adjusted for PTSD symptom severity (IES-R total score). After adjustment, Group 1 continued showing significantly higher Picture Completion scores than Group 2 (p < 0.016). (b, c) Logistic regression results also indicate that Group 1 had higher Picture Completion scores than both Group 0 (p = 0.013) and Group 2 (p = 0.028). Error bars indicate standard errors in (a) and credible interval in (b, c). *p < 0.05

Discussion

Our findings highlight observed differences in cognitive profiles among adults with ASD categorized by the reported timing of ACE exposure. Specifically, differences in the picture completion and figure weights subtest scores were observed across groups categorized by ACE timing. Participants who experienced ACEs before age 11 (Group 1) demonstrated significantly higher scores in these perceptual reasoning tasks compared with those who experienced ACEs between ages 11–18 (group 2). These results suggested that early-life adversity may have lasting cognitive effects in people with ASD, particularly in visual search abilities, vigilance behavior, and executive function. However, it should also be noted that Picture Completion and Figure Weights tasks engage a broader set of cognitive processes, including attentional selectivity, abstract reasoning, and pattern recognition [41]. Thus, the observed performance may reflect a multifaceted interaction between trauma-related adaptations and intrinsic cognitive capacities in ASD.

The present study distinguished ACE exposure occurring at ≤ 10 years and ≥ 11 years based on developmental considerations, especially focusing on the period of adolescence. Developmental frameworks commonly define adolescence as beginning around age 10, marking the transition from late childhood to early adolescence accompanied by the onset of puberty and substantial biological, psychological, and social changes [42]. This developmental transition is associated with extensive neurodevelopmental remodeling, particularly within fronto-limbic and socio-emotional networks that support higher-order cognitive and social processing [43]. Furthermore, the effects of early-life adversity are thought to depend on the developmental timing of exposure within sensitive periods of brain development [44, 45]. Therefore, distinguishing ACE exposure occurring during childhood versus early adolescence may provide insight into the differential neurodevelopmental consequences of adversity.

Notably, no significant differences were observed in the FSIQ or other WAIS-IV index scores across the groups, which suggested that the observed cognitive effects may be localized to specific subdomains of intellectual ability. This aligned with previous research results that trauma exposure selectively impacted aspects of cognitive function rather than uniformly affecting global intelligence [7-9, 13]. Lack of significant differences in ASD symptom severity across the groups further reinforced that these cognitive disparities were likely related to ACE timing rather than inherent ASD characteristics.

Furthermore, the differences in figure weight scores observed between Groups 1 and 2 introduced important considerations regarding executive function and reasoning abilities. The figure weights subtest is widely recognized as an assessment of quantitative reasoning and working memory [41]. Since Group 1 outperformed Group 2, early adversity may have contributed to an increased reliance on visual-perceptual reasoning in individuals with ASD, potentially as a compensatory mechanism for executive function weaknesses [41]. This finding was further supported by prior research results that individuals with ASD exhibited heightened visuospatial processing relative to neurotypical controls, which may be exaggerated by trauma exposure [46, 47].

Comparison with previous research

Significant elevation in the picture completion scores among individuals with early ACE exposure was consistent with prior research that linked hyperarousal symptoms of PTSD to enhanced perceptual vigilance [48, 49]. PTSD-related hypervigilance has been associated with increased attention to environmental details, which may partially explain the superior performance of Group 1 in detecting the missing visual components [23-26]. Nonetheless, Picture Completion also relies on attentional control, perceptual organization, and conceptual reasoning. It is therefore likely that trauma-related hypervigilance interacted with pre-existing individual differences in cognitive style to produce the observed patterns. Moreover, this finding aligned with those of previous studies, which found that trauma could lead to compensatory cognitive adaptations, where individuals developed heightened perceptual sensitivity as a survival mechanism [23, 24].

However, the significant difference in figure weight scores between Groups 1 and 2 was unexpected. Traditionally, the figure weights subtest, which assessed quantitative and analogical reasoning, was not associated with PTSD-related hypervigilance. A potential explanation was that executive attention processes differed between these groups [50]. Individuals with ASD exhibit variability in executive function, particularly in inhibitory control and working memory, which could mediate figure weights performance [51]. The finding that Group 1 outperformed Group 2 suggested that early-life adversity may differentially influence the development of executive attention in people with ASD; this hypothesis warrants further exploration.

Our results also parallel those from neuroimaging studies. Structural magnetic resonance imaging (MRI) analyses have identified reduced visual cortex volume in individuals with histories of severe childhood maltreatment, particularly sexual abuse and parental domestic violence [52, 53]. Furthermore, reduced visual memory capacity was linked to early trauma exposure, which supported the observed discrepancies in perceptual reasoning scores [54]. Additionally, functional MRI (fMRI) studies demonstrated that individuals with ASD processed visuospatial information differently than individuals with neurotypical development and had greater reliance on perceptual areas and reduced activation of reasoning-related brain regions [55]. These neural differences may partially explain why individuals with ASD with early ACE history performed better in perceptual reasoning tasks, as their cognitive strategies fundamentally differed from those of trauma-exposed individuals with neurotypical development.

Theoretical and practical implications

These findings contribute to theoretical models of ASD cognition and trauma-related adaptations. Results suggest that individuals with ASD may exhibit cognitive compensations in response to early-life adversity, particularly in perceptual vigilance and reasoning. Importantly, these findings should not be interpreted as evidence that trauma exclusively enhances specific cognitive functions. Rather, Picture Completion and Figure Weights performance may reflect a convergence of attentional selectivity, executive reasoning, and perceptual sensitivity. This complexity underscores the importance of integrating trauma adaptation theories with broader cognitive developmental frameworks in ASD. This aligned with the dual-process model of trauma adaptation, which posited that trauma survivors often develop heightened bottom-up processing (e.g., hypervigilance) while experiencing deficits in top-down executive control [23-26]. Our findings suggest that individuals with ASD with early ACEs may exhibit a distinct cognitive profile, which combines heightened perceptual sensitivity with relatively intact reasoning abilities.

Clinically, these findings suggest that the cognitive assessments of individuals with ASD should consider ACE history (≤ 10 years) when interpreting the WAIS-IV subtest results. High scores on the picture completion and figure weights subtests should not necessarily be interpreted as indicators of cognitive strength; rather, they should be viewed as potential markers of trauma-related cognitive adaptation.

Regarding education, this study underscores the need for trauma-informed teaching strategies in individuals with ASD. Given the link between early adversity and cognitive adaptations, personalized learning approaches considering heightened perceptual sensitivity may enhance engagement and learning outcomes. Additionally, interventions that target executive function, such as metacognitive training, may help individuals with ASD with ACE histories optimize their reasoning abilities in ways that generalize beyond structured testing environments [56-58].

Limitations and future research directions

The study’s limitations encompass its observational design, which constrains the capacity to establish causal relationships. Additionally, the reliance on self-reported data for adverse childhood experiences may introduce recall bias. Future research should examine how both the cumulative burden and qualitative characteristics of ACE exposure may differentially affect cognitive functioning. Furthermore, the cultural specificity of the sample limits the generalizability of the findings to other populations.

Given the complex interplay between ACEs, ASD, and cognitive adaptations, future research should explore the neurobiological correlates of trauma-related cognitive changes, incorporate longitudinal methodologies, and consider cross-cultural variations in ASD symptomatology and cognitive development. In addition, these findings should be replicated in a cohort with more balanced sample sizes, using more specific cognitive scales. These insights are essential for refining intervention strategies tailored to individuals with ASD who have experienced early-life adversity.

Conclusions

This study identified group differences in specific cognitive subdomains among adults with ASD based on the reported timing of ACE exposure. While our findings suggest a possible link between early adversity and cognitive profiles, the cross-sectional nature of the study precludes causal inference. Our findings emphasize the importance of trauma-informed assessments and personalized interventions to support cognitive adaptation in individuals with ASD. Future research should incorporate longitudinal designs and cross-cultural perspectives to elucidate the underlying mechanisms and refine intervention strategies tailored to individuals with ASD and early adversity.

Abbreviations

ACEs

Adverse childhood experiences

ASD

Autism spectrum disorder

WAIS-IV

Wechsler Adult Intelligence Scale-Fourth Edition

PTSD

Post-traumatic stress disorder

DSM-5

Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition

ADOS-2

Autism Diagnostic Observation Schedule-2

M.I.N.I

Mini-International Neuropsychiatric Interview

FSIQ

Full-Scale Intelligence Quotient

AQ-J

Autism Spectrum Quotient-Japanese version

VCI

Verbal Comprehension Index

PRI

Perceptual Reasoning Index

WMI

Working Memory Index

PSI

Processing Speed Index

IES-R-J

the Japanese-language version of the Impact of Event Scale-Revised

Author contributions

RI, KY, NK, NI, and MM were involved in data collection. RI, KY, and MM wrote the manuscript. NI and MM supervised this study. All authors have read and approved the final version of this manuscript.

Funding

This study received funding from the Grants-in-Aid for Scientific Research, KAKENHI 23H04173 and 24K02386 from the Japan Society for the Promotion of Science (M.M.) and Japan Agency for Medical Research and Development (AMED) under Grant Numbers 21wm04250XXs0101, 21uk1024002h0002, 22gm6310015h0002, 22gm1510009h0001, 23gm190004s0301, and 24wm0625510h0001 (M.M.).

Data availability

Data are available from the corresponding author upon request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Nara Medical University. Written informed consent was obtained from all participants prior to their inclusion in the study.

Consent for publication

All participants provided written informed consent for participation and publication of anonymized data.

Competing interests

The authors declare no competing interests.