Attentional and electrophysiological associations with executive function ability in young autistic children

Abstract

Difficulties in executive functioning (EF) have been consistently reported in autistic individuals, but less is known about the attentional and neural mechanisms driving these difficulties. We explored the associations between EF abilities and sustained attention, measured with eye-tracking, and spontaneous measures of EEG spectral power density in 176 2–8 year-old autistic children with a wide range of cognitive abilities. EF was measured using the Behavior Rating Inventory of Executive Function (BRIEF). We found that EF abilities were positively associated with look durations while watching complex, audiovisual stimuli involving social content and dyadic speech. We also found that EF was negatively associated with scalp-wide theta power and positively associated with frontal beta and gamma power. These results shed light on attentional and neural associations with EF abilities and underscore the role of frontal brain activity for EF in autism.

Introduction

Autism spectrum disorder (“autism”) is a heterogeneous neurodevelopmental condition characterized by difficulties in social interaction and the presence of restricted, repetitive behaviors. Challenges in executive function (EF), including difficulties with planning and cognitive flexibility, have been consistently reported in autistic children and adults^{1, 2–3}. Cognitive flexibility, inhibition, working memory, and other aspects of EF are mediated by the prefrontal cortex and allow one to adapt to environmental demands by attending to salient information and flexibly modifying one’s behaviors in response to varying feedback and situational demands^4,5. Thus, EF impairments can contribute to difficulties in regulatory behaviors that are important for interacting with others, adaptive behaviors, and academic performance. EF impairments in young autistic children have been found to be associated with difficulties in communication, play, and social relationships^6,7.

EF skills also play a role in the ability to direct and sustain attention to salient stimuli^8,9. Autistic children have been shown to have shorter look durations to dynamic visual stimuli with social content, including social stimuli, compared to neurotypical (NT) children¹⁰. However, it is unknown the extent to which these attentional impairments are associated with impairments in EF. It has been shown that young autistic children are less likely to exhibit spontaneous social orienting, social orienting in response to social bids, and sustained gaze toward others^{11, 12, 13, 14–15}. Many studies have examined how autistic individuals attend to different regions of a stimulus, whereas fewer have measured the ability to sustain focused attention irrespective of the fixation target.

In the current study, we were interested in whether difficulties in sustaining attention to visual stimuli, as reflected in shorter look duration, were associated with EF impairments in young autistic children. Furthermore, given that EF is mediated by the prefrontal cortex, and it has been hypothesized that differential maturation of the frontal lobe may contribute to EF impairment in autism^{16, 17–18}, we explored whether executive functioning was associated with patterns of frontal brain activity, as reflected in electroencephalography (EEG) recordings taken while children watched dynamic visual stimuli. We utilized eye-tracking (ET) to measure how long children were able to maintain their gaze on dynamic, audiovisual stimuli, especially during meaningful, social segments of the videos.

While findings regarding the prevalence and nature of EEG abnormalities in autism vary, general findings reported in EEG studies of autistic individuals include a pattern of excessive power in both low frequency (delta, theta) and high-frequency (beta, gamma) bands, and reduced relative power in the middle-range frequency band (alpha)^19,20. Theta and gamma bands may be good targets for EF research because they have been related to memory and task performance and attention and working memory, respectively. Wascher et al. (2014) reported that frontal theta increased with time on task and hypothesized that the increase in theta could be related to a compensatory increase in effort required to keep performance levels high²¹. Conejero et al. (2018) demonstrated that the observation of incorrect pattern configurations (i.e., error detection) produced increased frontal theta oscillatory activity in toddlers²². Taken together, these studies suggest that, in a task-directed scenario, theta acts in the frontal cortex to optimize performance.

Furthermore, oscillatory neural activity in the gamma frequency range has been shown to be associated with a variety of cognitive processes, including attention, memory, perceptual processing, and object perception. One previous study found that lower levels of gamma power were correlated with worse task performance, suggesting that increased gamma activity assists in task performance in children²³. Another study by Cantiani et al. (2019) found that infants with more favorable language outcomes showed increased left central gamma power as measured by EEG²⁴. Finally, a study assessing resting EEG power in children found that increased gamma power was associated with better executive function²⁵. Together these studies suggest that children with higher gamma power tend to have better language, attention, and inhibitory control.

While studies have explored patterns of attention, measured via ET, and brain activity, measured via EEG, in autistic children, there have been fewer studies that have examined the attentional and neural correlates of EF in this population. Given the large-scale integration of neural activity required for carrying out executive functions²⁶, it is of interest whether EF abilities are associated with distinct patterns of brain activity. Here, we explored whether specific components of EF were associated with average look durations measured via ET. Furthermore, given findings of structural and functional frontal lobe differences observed in autistic individuals compared to neurotypical individuals^27,28, we examined whether specific aspects of EF abilities were associated with patterns of frontal brain activity, or brain activity in other scalp regions.

In the present study, we conducted a secondary analysis using previously collected data from a clinical trial for autistic children²⁹. Data used in this analysis were collected at the baseline visit, before children were randomized to treatment or placebo groups, and included measures of EF, ET, and EEG from a relatively large (N = 176) group of young autistic children. EF was assessed via a well-established parent report measure^30,31. We hypothesized that higher levels of EF abilities would be associated with longer look durations during an ET task. We also hypothesized that EF abilities would be associated with frontal EEG power.

Methods

This is a secondary analysis of data from the baseline visit (prior to infusion) of a phase II randomized, placebo-controlled trial evaluating the efficacy of a single intravenous umbilical cord blood infusion for improving social communication skills in young children with autism (ClinicalTrials.gov: NCT02847182). Parents/legal guardians provided informed consent prior to study procedures. The study was approved by the Duke University Health System Institutional Review Board. Methods were carried out in accordance with institutional, state, and federal guidelines and regulations. The procedures in this study adhere to the tenants of the Declaration of Helsinki.

Study sample

Participants were 176 children (140 males and 36 females) 24 months to 97 months old (M = 64.9 months, SD = 19.5) with an existing diagnosis of autism spectrum disorder. Participants were screened using Autism Diagnostic Interview- Revised³², and autism diagnosis was confirmed using the Diagnostic and Statistical Manual of Mental Disorders Fifth Edition³³ and established by research reliable clinicians using the Autism Diagnostic Observation Schedule, Second Edition³⁴ and a structured caregiver interview. Parents also completed reports about autism related behaviors and children underwent EEG and ET assessments in the same week that the autism diagnosis was made. Mean Full Scale Ratio Intelligence Quotient for the sample was 69.0 (SD = 20.9). Demographic characteristics of the participants are presented in Table 1.

Table 1. Demographic characteristics of Sample.

N = 176 (140 males and 36 females)	Mean	Range	Standard deviation
Age (months)	64.86	27–95	19.52
Full scale ratio IQ)	70.77	23–133	20.58
ADOS-2 comparison score	8.02	2–10	1.63
EOWPVT-4 standard score	77.91	55–117	18.54
VABS-3 communication standard score	65.59	20–100	17.99
VABS-3 socialization standard score	63.02	34–102	13.98
Ethnicity n (%)
Hispanic	31 (17.6)
Non-Hispanic	145 (82.4)
Race n (%)
Non-white	29 (16.5)
White	147 (83.5)
Total pre-tax household income ($) n (%)
<50,000	10 (5.7)
50,000–99,999	31 (17.6)
100,000–149,999	45 (25.6)
150,000–199,999	30 (17.0)
200,000+	56 (31.8)
Unknown	4 (2.3)

IQ: Intelligence Quotient. ADOS-2: Autism Diagnostic Observation Schedule, Second Edition. EOWPVT-4: Expressive One-Word Picture Vocabulary Test, Fourth Edition. VABS-3: Vineland Adaptive Behavior Scales, Third Edition. Of the original 180 participants involved in the clinical trial, two pilot participants and two participants who were determined to be ineligible following randomization to treatment were excluded (n = 176).

Inclusion and exclusion

Inclusion/exclusion criteria. Inclusion criteria were (1) stable on current medications for at least 2 months prior to the infusion, (2) participants and parents/guardians were English speaking, and (3) an autologous umbilical cord blood unit or >/= ¾ HLA-matched allogeneic unrelated umbilical cord blood unit from the Carolinas Cord Blood Bank was available. Exclusionary criteria were (1) a history of prior cell therapy, (2) use of intravenous immunoglobulin (IVIG) or other anti-inflammatory medications (with the exception of NSAIDs), (3) known genetic syndrome, as indicated by medical records, (e.g., Fragile X), presence of dysmorphic features, pathogenic mutation or copy number variation associated with autism, and/or other significant medical and/or psychiatric comorbidity, (4) obvious physical dysmorphology, (5) an uncontrolled seizure disorder, (6) significantly impaired renal or liver function, (7) known active CNS infection, evidence of uncontrolled infection, and/or HIV positivity, (8) family unwilling or unable to commit to study-related assessments, and/or (9) clinically significant abnormalities in complete blood count. Genetic testing was not required as part of this study.

Measures

Cognitive functioning. Nonverbal Development Quotient (NVDQ) was assessed with the Mullen Scales of Early Learning (MSEL)³⁵ for children under four years of age and Differential Ability Scales, Second Edition (DAS-II)³⁶ for children four years of age and older. The cognitive measures were administered in-person by trained clinicians at the time of the diagnostic assessment. Studies examining nonverbal intelligence and EF, particularly in early development, demonstrate associations between nonverbal cognitive ability and EF outcomes^37,38. Thus, NVDQ was utilized as a covariate in the analyses.

Executive functioning. Behavior Rating Inventory of Executive Function (BRIEF) is a parent questionnaire used to assess EF in everyday environments. Both the BRIEF (used to assess school-aged children) and the BRIEF Preschool-Version (BRIEF-P) were used based on the children’s age at the start of the study. EF was measured using BRIEF-P in participants aged 2 years and 0 months to 4 years and 11 months, and EF was measured using BRIEF in participants aged 5 years and 0 months to 8 years and 1 months. For both versions of the BRIEF, higher T-scores indicate greater impairment (i.e., weaker EF), and scores >/= 65 (i.e., 1.5 + SDs above the mean) suggest clinically significant ratings. The five overlapping clinical scales used in both BRIEF and BRIEF-P (Inhibit, Shift, Emotional Control, Working Memory, Plan/Organize) and the Global Executive Composite (GEC) score were used for analysis.

Eye-tracking task. Participants watched a 3-minute video of dynamic audiovisual stimuli portraying an actress in a setting containing four toys and a table with ingredients for making sandwiches¹¹. The video included an episode during which the actress engaged in child-directed speech while looking at the camera (“Actress Dyadic Bid”) and an episode during which the actress looked at the camera, a toy became activated and made noise, and the actress looked at another non-activated toy rather than the activated, moving toy (“Actress with Activated Toys”). Both segments lasted 41 continuous seconds. Average look duration (ALD; i.e., average length of separate uninterrupted looking periods) was calculated by dividing the total viewing time on the video media by the total number of distinct periods of attention. The media region was defined as the entire video frame and ALD was calculated for the Actress Dyadic Bid and Actress with Activated Toys episodes separately.

EEG recordings. Continuous EEG data were collected while children watched 1-minute video clips of a woman reciting nursey rhymes (‘social’ video), brightly colored, sound-making dynamic toys (‘toys’ video), or floating bubbles without sound (‘bubbles’ video). EEG data were recorded from 124 electrodes with reference to Cz using a Hydrocel Geodesic Sensor Net (EGI) and Net Amps 400 amplifier in Netstation 4.5.6 with a sampling rate of 1000 Hz. Twelve frontal, twelve central, and twelve posterior electrodes were averaged into three groups; electrodes within each group were distributed evenly among the left, midline, and right hemispheres to achieve maximal scalp coverage (see Fig. 1). EEG data were filtered with a 1–100 Hz bandpass filter and a 58–62 Hz bandstop filter. Data were decomposed using Second Order Blind Identification (SOBI) as implemented in EEGLAB and re-referenced to the common average. A Fast Fourier Transformation (FFT) was performed on the rectangular windowed time series. EEG power data were reduced to averaged relative power bands: theta (5–7 Hz), alpha (8–10 Hz), low-beta (11–20 Hz), high-beta (21–30 Hz), and gamma (31–70 Hz). 145 participants provided sufficient artifact-free EEG data during the social video, 155 for the toys video, and 150 for the bubbles video.

[See PDF for image]

Fig. 1

Twelve frontal (green), twelve central (blue), and twelve posterior (yellow) electrodes were averaged into three groups, electrodes within each group were distributed evenly among the left, midline, and right hemispheres to achieve maximal scalp coverage.

Analytic strategy

Linear regression, controlling for age and NVDQ, was used to examine associations between EF and measures of attention and brain activity. The dependent variables were BRIEF subscales, and the three independent variables were: one biomarker variable (either sustained attention measured by eye-tracking or the strength of brain activity measured by EEG power), age, and NVDQ. All statistical analyses were performed in SAS Version 9.4 (SAS Institute Inc., Cary, NC). Results are reported with and without correcting for Type I error. Corrections for Type I error were made using the Benjamini-Hochberg method³⁹ at the model-level, following guidance from Cohen et al. (2013), which suggests that adjusting for model-level comparisons with sets of variables assists in protecting the individual variable-level tests⁴⁰. The p-values were ranked from smallest to largest and compared to the formula:

where i is the rank order, m is the number of regression models, and Q is the overall alpha rate acceptable to the researchers³⁹. Then, the largest p-value which is lower than the outcome of this formula and all other p-values lower than it are acceptable for interpretation. In this study, Q was set to 0.05 and m was 228, and thus all models with a biomarker (EEG or eye-tracking) p < 0.0195 were accepted. We adjusted alpha at the model-level (rather than the variable-level) so that the family-wise error rate would equal 0.05 and the variable-level statistics would remain interpretable at the p < 0.05 level. The statistical significance level from the overall omnibus F test was then compared to this value, and only eligible models were examined for further interpretation.

Results

Nonverbal intelligence and age

Associations were found between the dependent variables (BRIEF subscales) and covariates age and NVDQ. Nonverbal intellectual ability demonstrated a positive relationship with the working memory subscale. Similarly, age was positively associated with the inhibit subscale. The shift subscale was positively associated with NVDQ and negatively associated with age. Planning/organizing and emotional control subscales were not associated with either age or NVDQ.

Eye-tracking and executive function

For the Actress Dyadic Bid eye-tracking condition, look durations to the video were positively associated with planning/organizing skills (p = 0.03, B=−2.51, t=−2.25, SE = 1.11, 95% CI=[−4.70,−0.31]). Eye-tracking results are presented in Table 2.

Table 2. Regression coefficients (β) examining associations between executive function and attention, with age and nonverbal IQ as covariates.

BRIEF subscalesa	Average look duration (eye-tracking)
	Actress dyadic bid	Actress with activated toys
Inhibition	−0.04	0.04
Shifting	−0.04	−0.04
Emotional control	−0.07	0.00
Working memory	−0.13	−0.02
Planning/Organizing	−0.19*	−0.05
Global executive composite	−0.13	−0.02

^aBRIEF: Behavior Rating Inventory of Executive Function; *p < 0.05 without correction for multiple comparisons.

EEG and executive function

For EEG recorded during the social video, EF abilities in the domains of inhibition (p = 0.01, B = 134.80, t = 2.50, SE = 54.02, 95% CI=[28.00,241.61]), working memory (p = 0.03, B = 134.68, t = 2.26, SE = 59.60, 95% CI=[16.84,252.52]), and global executive functioning (p = 0.04, B = 119.94, t = 2.07, SE = 57.93, 95% CI=[5.41,234.47]) were negatively associated with frontal theta power.

For EEG recorded during the toys video, inhibition skills were negatively associated with frontal theta power (p = 0.04, B = 120.02, t = 2.04, SE = 58.86, 95% CI=[3.72,236.33]) and central theta (p = 0.049, B = 127.78, t = 1.98, SE = 64.51, 95% CI=[0.30,255.26]) EEG power. Working memory (p = 0.04, B=−699.51, t=−2.12, SE = 329.58, 95% CI=[−1350.77,−48.25]) was positively associated with frontal low-beta power. Additionally, working memory (p = 0.02, B=−886.72, t=−2.36, SE = 375.27, 95% CI=[−1628.26,−145.18]), planning/organizing (p = 0.02, B=−1033.95, t=−2.41, SE = 428.13, 95% CI=[−1880.00,−187.89]), and global EF skills (p = 0.04, B=−765.07, t=−2.07, SE = 369.71, 95% CI=[−1495.62,−34.51]) were positively associated with frontal gamma power.

For EEG collected during the bubbles video, inhibition (p = 0.001, B = 172.10, t = 3.28, SE = 52.45, 95% CI=[68.44,275.76]), working memory (p = 0.009, B = 154.58, t = 2.66, SE = 58.17, 95% CI=[39.60,269.55]), planning/organizing (p = 0.04, B = 139.97, t = 2.09, SE = 67.09, 95% CI=[7.34,272.60]), and global EF (p = 0.01, B = 144.16, t = 2.51, SE = 57.46, 95% CI=[30.59,257.73]) were negatively associated with frontal theta power. Additionally, the working memory domain (p = 0.048, B = 103.90, t = 1.99, SE = 52.20, 95% CI=[0.73,207.09]) was negatively associated with posterior theta power. Finally, high-beta power was positively associated with inhibition skills (p = 0.03, B=−605.68, t=−2.22, SE = 272.43, 95% CI=[−1144.16,−67.20]).

EEG results are presented in Tables 3 and 4.

Table 3. Regression coefficients (β) reflecting associations between executive function and higher frequency frontal brain activity, with age and nonverbal IQ as covariates.

BRIEF subscalesa	Frontal EEG relative power
	beta β 11–30 Hz			gamma γ 31–70 Hz
	Bubbles	Social	Toys	Bubbles	Social	Toys
Inhibition	−0.14	−0.07	−0.13	−0.16*∇	−0.19*∇	−0.12
Shifting	−0.05	−0.03	−0.04	−0.05	−0.07	0.00
Emotional control	−0.11	−0.01	−0.12	−0.05	−0.10	−0.08
Working memory	−0.14	−0.10	−0.20*∇	−0.15	−0.20*∇	−0.06
Planning/Organizing	−0.11	−0.05	−0.21*	−0.07	−0.09	−0.02
Global executive composite	−0.13	−0.06	−0.18*∇	−0.12	−0.17*∇	−0.08

^aBRIEF: Behavior Rating Inventory of Executive Function. *p < 0.05 without correction. ∇ Association remained statistically significant after correction for multiple comparisons.

Table 4. Regression coefficients (β) obtained from linear regression between BRIEF and frontal theta and alpha EEG relative power, utilizing age and NVIQ as covariates.

BRIEF subscalesa	Frontal EEG relative power
	theta θ 5–7 Hz			alpha α 8–10 Hz
	Bubbles	Social	Toys	Bubbles	Social	Toys
Inhibition	0.26**∇	0.20*∇	0.16*∇	0.05	0.13	0.04
Shifting	0.03	0.00	−0.05	0.00	0.10	−0.02
Emotional control	0.07	0.08	0.05	−0.07	0.03	−0.08
Working memory	0.21**∇	0.19*∇	0.12	0.06	0.13	0.00
Planning/Organizing	0.17*	0.13	0.06	0.03	0.12	−0.01
Global executive composite	0.20*∇	0.17*∇	0.10	0.03	0.12	−0.02

^aBRIEF: Behavior Rating Inventory of Executive Function. *p < 0.05, **p < 0.01 without correction. ∇ Association remained statistically significant after correction for multiple comparisons.

Corrections for multiple comparisons/type 1 error

Associations between frontal theta activity and executive function abilities remained statistically significant after correcting for multiple comparisons using the Benjamini-Hochberg method. Specifically, frontal theta activity during the neutral “bubbles” video was negatively associated with inhibition, working memory, and global EF abilities. For the social video, frontal theta activity was negatively associated with inhibition.

Discussion

In the current study, we explored the relationships between executive function skills and the ability to sustain attention to a complex, dynamic audiovisual stimulus and patterns of EEG in young autistic children. We found that working memory and planning/organizing skills were positively associated with look durations during an eye-tracking task, such that children with higher scores in these EF domains tended to demonstrate enhanced sustained attention abilities measured by longer look durations.

Additionally, several consistent patterns emerged when we examined the relationships between EF skills and patterns of EEG activity. Across the three video stimuli presented, the most consistent EEG finding was that inhibition and working memory domains were negatively associated with frontal theta EEG power, such that as skills in these domains increased, frontal theta power tended to decrease. It is important to note that this finding demonstrates the associations between frontal theta EEG power and EF abilities; this study is not designed to determine whether the autism group demonstrated differences in EEG power relative to neurotypical children. Other studies have found that autistic individuals exhibit higher levels of theta power relative to neurotypical individuals²⁰. Thus, it is possible that the autistic children who displayed lower theta EEG power, and thus tended to have higher EF scores on average, in our study exhibited a pattern more similar to that of neurotypical children. These findings suggest that lower resting or passive-state theta power could reflect more mature or efficient prefrontal functioning, and that EEG may be a useful physiologic index of EF-related neural processes in autism.

We also found that inhibition, working memory, and planning/organizing were positively associated with frontal beta power. In the context of prior research, past studies have reported higher levels of low frequency power (theta) during states of low vigilance^41,42. Braboszcz and Delorme (2011) reported increased activity of low-frequency EEG power (e.g., theta) and a decrease of high-frequency EEG power (e.g., alpha, beta) when individuals began to lose focus on a task⁴³. Taken together, these studies indicate that both theta and beta bands may be related to mental states associated with attention and vigilance.

Our results also showed that working memory and planning/organizing abilities were positively associated with frontal EEG gamma power. This finding helps to add to the limited field of research done on this topic. A study by Tallon-Baudry et al. (1998) found that when individuals have to actively maintain visual stimuli in working memory, the stimuli induced more gamma oscillations as compared to a control task in which no memorization was required⁴⁴. Others have suggested that attentional modulation of gamma activity reflects the matching of attended objects with the contents of working memory⁴⁵. Additionally, increased frontal gamma activity has been previously observed during perceptual switching (i.e., visual perception of reversible or ambiguous figures)⁴⁶. Taken together, these studies support the idea that gamma power plays a role in both working memory and perceptual attention.

The observed association between parent-reported EF skills and neural biomarkers of attention and frontal brain activity has important implications. While prior research has noted discrepancies between questionnaire-based and performance-based measures of EF in autistic populations, the BRIEF was specifically designed to capture “real world” EF challenges in everyday contexts. The findings of this study suggest that caregiver reports on the BRIEF are not only ecologically valid but may also reflect meaningful underlying neurocognitive differences. This not only strengthens the case for using the BRIEF as a clinically relevant tool in the assessment of EF in autism, but it also highlights the value of combining caregiver insight physiological measures to more fully characterize EF phenotypes.

A limitation of the present study is that other demographics besides age and IQ were not controlled for, some of which may be relevant for EF development, such as socio-economic status⁴⁷. Furthermore, EF was measured using parent-report which could be subject to biases stemming from limited caregiver insight or opportunity to observe EF behaviors across contexts; however, the associations between parent-reported EF and neural associations described in the present study provide support for the ecological validity of these measures. Finally, the sample included a relatively wide age range (2–8 years), during which EF and attentional processes are still rapidly developing. This developmental fluidity may introduce variability that complicates the interpretation of results. While this heterogeneity reflects the real-world variability observed in young autistic populations, it may also limit the generalizability and replicability of our results across more narrowly defined age bands. In this present study, the large age range provided a challenge for establishing the frequency bands upper and lower limits, as the EEG power spectra is still developing in this age window. While we controlled for age and reported its associations with EF, future studies should consider studying the interaction effects of age on neural activity and executive functioning.

In conclusion, we found that autistic children’s ability to sustain their attention to social audiovisual stimuli was positively associated with EF skills. We also found that EF skills were positively associated with frontal brain activity in fast beta/gamma bands, and negatively associated with frontal brain activity in the slower theta band. While these findings need to be replicated in future studies, they offer hypotheses for future work regarding the neural basis of EF in autism and suggest that therapeutic approaches and/or behavioral modification programs that promote EF and frontal brain function could be useful for autistic children. Given the role of EF skills in regulatory behavior, interventions targeting this domain could have broader implications for improving children’s social, communicative, and adaptive behavior as well as academic competence. Interventions to improve EF abilities in autistic children have shown promise in increasing some domains of EF abilities⁴⁸.

The findings from the present study could be useful in designing future clinical trials of EF interventions where inclusion of an eye-tracking or EEG biomarker is desired. Eye-tracking patterns and/or EEG frontal theta power could be incorporated as a predictor of response to intervention or a measure of outcome. Such studies would provide insight into mechanisms underlying change in response to EF interventions, and help identify which children might be most responsive to an EF intervention. Designing such interventions that could improve EF function in autistic children could have significant implications for their social, adaptive, and academic outcomes.

Acknowledgements

This research was supported by grants from the Marcus Foundation and NICHD. C. Leahy conducted this project for her medical school thesis at Duke University Medical School. We thank the Marcus Foundation, NICHD, and the children that participated in the research studies and their families.

Author contributions

C. L., S.M., G.D., designed the study. G.D. wrote the funding proposal. S.M., J.H., K.L.H.C., E.T., L.F., S.V., and J.G. gathered the clinical data. C.L., S.M., J.G., and G.D. wrote the manuscript. C.L., S.M., S.C., M.S., G.D., performed the analysis and interpreted the data. C.L., S.M., J.G., and G.D. wrote the main paper and C.L., S.M., J.G., and G.D. revised the manuscript. All authors read and approved the final manuscript.

Data availability

Groups interested in direct use of the data can contact the authors for permission and availability. Data is stored in a secured partition at Duke Health. To request access to data from this study, please contact Geraldine Dawson PhD ([email protected]).

Declarations

Competing interests

Dr. Dawson is on the Scientific Advisory Boards of the Nonverbal Learning Disability Project, and Tris Pharma, Inc., is a consultant to Apple, Inc., and received book royalties from Guilford Press, Oxford University Press, and Springer Nature Press. Dr. Dawson has developed technology, data, and/or products that have been licensed to Apple, Inc. and Cryocell, Inc. and Dawson and Duke University have benefited financially. Dr. Dawson has received funding from the Marcus Foundation, Cord Blood Association, National Institutes of Health (NIH), and Simons Foundation. Dr. Carpenter has had funding by the National Institutes of Health (NIH), the Department of Defense, and the Brain and Behavior Foundation. Dr. Carpenter is a standing member on the Programmatic Panel for the Department of Defense Congressionally Directed Medical Research Programs (CDMRP) Autism Research Program and has served as an ad hoc reviewer on NIH review panels; she has received reimbursement for her time on these panels. Dr. Tenenbaum is a consultant to Maplight Therapeutics. The remaining authors declare no competing interests.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

1. Garon, N; Smith, IM; Bryson, SE. Early executive dysfunction in ASD: simple versus complex skills: early executive dysfunction. Autism Res.; 2018; 11, pp. 318-330. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29149519]

2. Granader, Y et al. Characterizing the factor structure of parent reported executive function in autism spectrum disorders: the impact of cognitive inflexibility. J Autism Dev. Disord; 2014; 44, pp. 3056-3062. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24972681][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6084425]

3. Hill, EL. Evaluating the theory of executive dysfunction in autism. Dev Rev.; 2004; 24, pp. 189-233.

4. Miller, EK; Cohen, JD. An integrative theory of prefrontal cortex function. Annu. Rev. Neurosci.; 2001; 24, pp. 167-202.1:CAS:528:DC%2BD3MXls1Shsro%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/11283309]

5. Diamond, A. Executive functions. Ann. Rev. Psychol.; 2013; 64, pp. 135-168.

6. Gilotty, L; Kenworthy, L; Sirian, L; Black, DO; Wagner, AE. Adaptive skills and executive function in autism spectrum disorders. Child. Neuropsychol.; 2002; 8, pp. 241-248. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/12759821]

7. Jarrold, C. A review of research into pretend play in autism. Autism; 2003; 7, pp. 379-390. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/14678677]

8. Berninger, VW; Richards, TL; Abbott, RD. Brain and behavioral assessment of executive functions for Self-Regulating levels of Language in reading brain. J. Nat. Sci.; 2017; 3, e464. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29104930][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5662027]

9. Brandes-Aitken, A; Braren, S; Swingler, M; Voegtline, K; Blair, C. Sustained attention in infancy: A foundation for the development of multiple aspects of self-regulation for children in poverty. J. Exp. Child. Psychol.; 2019; 184, pp. 192-209. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31039447][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6528818]

10. Isaev, DY et al. Relative average look duration and its association with neurophysiological activity in young children with autism spectrum disorder. Sci. Rep.; 2020; 10, 1912.1:CAS:528:DC%2BB3cXkvVaju7o%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32024855][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7002421]

11. Chawarska, K; Shic, F. Looking but not seeing: atypical visual scanning and recognition of faces in 2 and 4-Year-Old children with autism spectrum disorder. J. Autism Dev. Disord; 2009; 39, pp. 1663-1672. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19590943][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4878114]

12. Dawson, G; Meltzoff, AN; Osterling, J; Rinaldi, J; Brown, E. Children with autism fail to orient to naturally occurring social stimuli. J. Autism Dev. Disord; 1998; 28, pp. 479-485.1:STN:280:DyaK1M7jtFShtA%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/9932234]

13. Leekam, SR; Ramsden, CAH. Dyadic orienting and joint attention in preschool children with autism. J. Autism Dev. Disord; 2006; 36, pp. 185-197. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16502142]

14. Ozonoff, S et al. A prospective study of the emergence of early behavioral signs of autism. J. Am. Acad. Child. Adolesc. Psychiatry; 2010; 49, .e1–2 pp. 256-266. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20410715][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2923050]

15. Trepagnier, C; Sebrechts, MM; Peterson, R. Atypical face gaze in autism. Cyberpsychol Behav.; 2002; 5, pp. 213-217. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/12123243]

16. Hardan, AY; Jou, RJ; Keshavan, MS; Varma, R; Minshew, NJ. Increased frontal cortical folding in autism: a preliminary MRI study. Psychiatry Res. Neuroimaging; 2004; 131, pp. 263-268.

17. Levitt, JG. Cortical sulcal maps in autism. Cereb. Cortex; 2003; 13, pp. 728-735. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/12816888]

18. Zilbovicius, M et al. Delayed maturation of the frontal cortex in childhood autism. Am. J. Psychiatry; 1995; 152, pp. 248-252.1:STN:280:DyaK2M7kt1emuw%3D%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/7840359]

19. Stroganova, TA et al. Abnormal EEG lateralization in boys with autism. Clin. Neurophysiol.; 2007; 118, pp. 1842-1854. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17581774]

20. Wang, J et al. Resting state EEG abnormalities in autism spectrum disorders. J. Neurodev Disord; 2013; 5, 24. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24040879][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3847481]

21. Wascher, E et al. Frontal theta activity reflects distinct aspects of mental fatigue. Biol. Psychol.; 2014; 96, pp. 57-65. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24309160]

22. Conejero, Á; Abundis-Gutiérrez, A; Rueda, MR. Frontal Theta activation associated with error detection in toddlers: influence of Familial socioeconomic status. Dev. Sci.; 2018; 21, 1.

23. Dockstader, C; Wang, F; Bouffet, E; Mabbott, DJ. Gamma deficits as a neural signature of cognitive impairment in children treated for brain tumors. J. Neuroscience: Official J. Soc. Neurosci.; 2014; 34, 26 pp. 8813-8824.1:CAS:528:DC%2BC2cXhvV2ltb%2FL

24. Cantiani, C. et al. Reduced left-lateralized Pattern of event-related EEG Oscillations in Infants at Familial Risk for Language and Learning Impairment. NeuroImage: Clinical22, 101778 (2019).

25. Tarullo, AR et al. Gamma power in rural Pakistani children: links to executive function and verbal ability. Dev. Cogn. Neurosci.; 2017; 26, pp. 1-8. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28436831][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6987759]

26. Uhlhaas, PJ; Singer, W. Neural synchrony in brain disorders: relevance for cognitive dysfunctions and pathophysiology. Neuron; 2006; 52, pp. 155-168.1:CAS:528:DC%2BD28XhtFentr7K [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17015233]

27. Carper, RA; Courchesne, E. Localized enlargement of the frontal cortex in early autism. Biol. Psychiatry; 2005; 57, pp. 126-133. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15652870]

28. Murias, M; Webb, SJ; Greenson, J; Dawson, G. Resting state cortical connectivity reflected in EEG coherence in individuals with autism. Biol. Psychiatry; 2007; 62, pp. 270-273. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/17336944][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2001237]

29. Dawson, G et al. A phase II randomized clinical trial of the safety and efficacy of intravenous umbilical cord blood infusion for treatment of children with autism spectrum disorder. J. Pediatr.; 2020; 222, pp. 164-173e5.1:CAS:528:DC%2BB3cXpslSqtLo%3D [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32444220]

30. Gioia, G. A., Espy, K. A. & Isquith, P. K. Behavior Rating Inventory of Executive Function-Preschool Version (PAR, Inc, 2003).

31. Gioia, G. A., Isquith, P. K., Guy, S. C. & Kenworthy, L. Behavior Rating Inventory of Executive Function: Professional Manual (PAR, Inc, 2000).

32. Rutter, M., Le Couteur, A., Lord, C. & Faggioli, R. Autism Diagnostic Interview-Revised (ADI-R). (Western Psychological Services, 2005).

33. American Psychiatric Association. Diagnostic and statistical manual of mental disorders (5th Edition).

34. Lord, C. et al. Autism Diagnostic Observation Schedule (ADOS-2). (Western Psychological Services, 2012).

35. Mullen, E. Mullen Scales of Early Learning (NCS Pearson, 2002).

36. Elliot, C. D. Differential Ability Scales - Second Edition. (2007).

37. Everaert, E et al. Nonverbal executive functioning in relation to vocabulary and morphosyntax in preschool children with and without developmental Language disorder. J. Speech Lang. Hear. Res.; 2023; 66, 6 pp. 1872-1889.

38. Bertollo, JR; Yerys, BE. More than IQ: executive function explains adaptive behavior above and beyond nonverbal IQ in youth with autism and lower IQ. Am. J. Intellect. Dev. Disabil.; 2019; 124, 3 pp. 191-205. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31026201][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6815227]

39. Benjamini, Y; Hochberg, Y. Controlling the false discovery rate: A practical and powerful approach to multiple testing. J. R Stat. Soc. Ser. B Methodol.; 1995; 57, pp. 289-300.1325392

40. Cohen, A; Ma, Y; Sackrowitz, H. Nonparametric multiple testing procedures. J. Stat. Plan. Inference; 2013; 143, pp. 1753-1765.3082231

41. Gonçalves, ÓF; Carvalho, S; Mendes, AJ; Leite, J; Boggio, PS. Neuromodulating attention and Mind-Wandering processes with a single session real time EEG. Appl. Psychophysiol. Biofeedback; 2018; 43, pp. 143-151. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29797155]

42. Smit, AS; Fortuyn, D; Eling, HA; Coenen, AM. L. Diurnal spectral EEG fluctuations in narcoleptic patients during rest and reaction time tasks. J. Sleep. Res.; 2005; 14, pp. 455-461. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16364147]

43. Braboszcz, C; Delorme, A. Lost in thoughts: neural markers of low alertness during Mind wandering. NeuroImage; 2011; 54, pp. 3040-3047. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20946963]

44. Tallon-Baudry, C; Bertrand, O; Peronnet, F; Pernier, J. Induced gamma-band activity during the delay of a visual short-term memory task in humans. J. Neurosci. Off J. Soc. Neurosci.; 1998; 18, pp. 4244-4254.1:CAS:528:DyaK1cXjsV2qtLw%3D

45. Herrmann, CS; Lenz, D; Junge, S; Busch, NA; Maess, B. Memory-matches evoke human gamma-responses. BMC Neurosci.; 2004; 5, 13. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/15084225][PubMedCentral: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC419345]

46. Başar-Eroglu, C; Strüber, D; Kruse, P; Başar, E; Stadler, M. Frontal gamma-band enhancement during multistable visual perception. Int. J. Psychophysiol.; 1996; 24, pp. 113-125. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/8978438]

47. Lawson, G. M., Hook, C. J. & Farah, M. J. A Meta-Analysis of the relationship between socioeconomic status and executive function performance among children. Developmental Science21(2), e12529 (2018).

48. de Vries, M; Prins, PJ; Schmand, BA; Geurts, HM. Working memory and cognitive flexibility-training for children with an autism spectrum disorder: a randomized controlled trial. J. Child. Psychol. Psychiatry; 2015; 56, 5 pp. 566-576. [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/25256627]