EEG functional connectivity as a prognostic biomarker of adaptive function in autistic people.
Translational psychiatry · 2026
Abstract originale
Many autistic people have challenges with adaptive function, impacting education, employment and independent-living goals. Adaptive function outcomes of autistic people vary considerably, which makes planning for future needs challenging. Here, using a developmentally sensitive approach, we investigated if cortico-cortical functional connectivity - a core neurobiological feature that differs in autism - could predict longitudinal changes in adaptive function in autistic people. Using electroencephalography in 150 autistic and 159 non-autistic participants aged 6-31 years, we investigated if mean degree and network organisation (small-world index) predict longitudinal changes in adaptive function over 19-months. We found that small-world index significantly predicted changes in adaptive function in autistic people across the entire age-range. Predictive performance was best for autistic youth (15-24-year-olds), where mean degree and small-world index explained 21 and 30% of additional variance in outcomes, respectively, outperforming measures of intelligence and autistic features. In categorising binary (improved versus not-improved) outcomes, the model containing mean degree had an AUC of 0.84 [95% CI: 0.71-0.97] in 15-24-year-olds, while that containing small-world index had an AUC of 0.76 [95% CI: 0.63-0.89] across the 6-31-year age-range. Both metrics demonstrated properties desired in prognostic biomarkers: high test-retest reliability and convergence with underlying biology (significant associations with genetic variation in brain volume-related genes). Thus, we demonstrate the first evidence that electroencephalography-derived functional connectivity metrics show promise as prognostic biomarkers of adaptive function in autistic people. Potential precision-medicine applications include stratifying participants in clinical trials and identifying those at risk of declining function in clinical settings.
Parole chiave
Leggi il testo integrale originale
Testo estratto dalla pubblicazione: sono esclusi immagini, media, supplementi e parti soggette a diritti distinti.
EEG functional connectivity as a prognostic biomarker of adaptive function in autistic people
Many autistic people have challenges with adaptive function, impacting education, employment and independent-living goals. Adaptive function outcomes of autistic people vary considerably, which makes planning for future needs challenging. Here, using a developmentally sensitive approach, we investigated if cortico-cortical functional connectivity – a core neurobiological feature that differs in autism – could predict longitudinal changes in adaptive function in autistic people. Using electroencephalography in 150 autistic and 159 non-autistic participants aged 6–31 years, we investigated if mean degree and network organisation (small-world index) predict longitudinal changes in adaptive function over 19-months. We found that small-world index significantly predicted changes in adaptive function in autistic people across the entire age-range. Predictive performance was best for autistic youth (15–24-year-olds), where mean degree and small-world index explained 21 and 30% of additional variance in outcomes, respectively, outperforming measures of intelligence and autistic features. In categorising binary (improved versus not-improved) outcomes, the model containing mean degree had an AUC of 0.84 [95% CI: 0.71–0.97] in 15–24-year-olds, while that containing small-world index had an AUC of 0.76 [95% CI: 0.63–0.89] across the 6–31-year age-range. Both metrics demonstrated properties desired in prognostic biomarkers: high test-retest reliability and convergence with underlying biology (significant associations with genetic variation in brain volume-related genes). Thus, we demonstrate the first evidence that electroencephalography-derived functional connectivity metrics show promise as prognostic biomarkers of adaptive function in autistic people. Potential precision-medicine applications include stratifying participants in clinical trials and identifying those at risk of declining function in clinical settings.
Introduction
A minority of autistic adults work full-time (12–27%), are enrolled in or have completed postsecondary education (39%) or live independently (5–12%; 1–3). Consequently, identifying approaches to support function in these domains is a research priority of autistic people [4, 5]. The neurocognitive construct most strongly associated with these abilities in autistic people is adaptive function [1, 3, 6–8]: the application of one’s abilities to navigate the everyday environment, including in a self-sufficient manner [9]. Indeed, adaptive function has been a key outcome measure in autism clinical trials [10, 11]. However, there is substantial variation in how adaptive function abilities of autistic people change over time [12, 13]. This poses a challenge in helping autistic young-people and adults plan requirements for future support or independence, and risks concealing important subgroup effects in clinical trials [14], highlighting a need for precision-medicine approaches.
Current behavioural predictors of improvements in adaptive function in autistic people – higher intelligence [1, 15, 16], higher baseline adaptive function [1, 16] and speech abilities [16] – leave most of the variance in adaptive function outcomes unexplained [9, 15, 17–19]. Existing putative neurobiological predictors – including the speed of neural response to faces (the N170 event-related potential; [14] and cortical structural differences [20] – are either limited in the domains of adaptive function outcomes they predict [14] and/or only explain a small proportion of additional variance (6% each) in outcomes [14, 20]. This highlights the need for better prognostic biomarkers.
Several studies have demonstrated that brain connectivity is different in autistic compared to non-autistic people [21–27]. Both the amount (often quantified as mean degree; 21, 28, 29) and organisation [30–33] of functional connectivity in large-scale brain networks have been associated with the extent of autistic social-communication and behavioural traits. An important feature of brain-networks across species is their small-world organisation [34–36], which facilitates specialised (segregated) and co-ordinated (integrated) function within and across neural systems [34, 37]. Altered small-world organisation has been found in autism [30–32]. Therefore, these functional connectivity metrics may show promise as prognostic biomarkers for adaptive function in autistic people.
However, previous functional connectivity findings in autistic people have been heterogenous [22, 24, 38–40], limiting progress in biomarker discovery. Considering age-effects in autism research is especially important given that autism is associated with altered rates of white matter [22, 33, 41–43], brain volume [44] and behavioural [45] maturation. Yet, most prior studies have inadequately accounted for developmental stage [40], possibly contributing to heterogenous findings.
Here, we investigated the performance of functional connectivity-derived mean degree (MD) and small-world index (SWI) as candidate prognostic biomarkers for adaptive function outcomes in autistic people. We explicitly studied age effects in a large dataset comprising participants aged 6–31 years, hypothesising that the extent to which MD/SWI predicts adaptive function outcomes differs by developmental stage. A candidate prognostic biomarker needs to be associated with underlying biological processes, statistically associated with an outcome, quantifiable in real-world settings and highly reliable [46–48]. We assessed convergence of functional connectivity metrics with biological processes by quantifying associations with polygenic variation, as autism [49, 50] and functional connectivity [51–53] are both highly heritable. We explored polygenic variation in autism- and brain-volume- related genes, the latter motivated by findings of altered brain volume [20, 44] and arguments that brain volume differences underly functional connectivity differences in autistic people [54]. We measured neural activity using electroencephalography (EEG) as it is widely available in clinical settings and cost-effective [55]. EEG-based functional connectivity also reflects the mechanisms underlying neural interactions [56–58]. Finally, we quantified the reliability of MD and SWI.
Participants and methods
Details are included in the Supplementary Methods section.
Ethics approval and consent to participate
This study was approved by ethics review boards at each study site, as detailed in Loth et al. [59]. The ethics committees were: 1) London-Central and Queen Square Health Research Authority Research Ethics Committee in the United Kingdom (13/LO/1156); 2) Regio Arnhem-Nijmegen (Radboud University Medical Centre Institute Ensuring Quality and Safety Committee on Research Involving Human Subjects Arnhem-Nijmegen) and 3) Radboud universitair medisch centrum Instituut Waarborging Kwaliteit en Veiligheid Commissie Mensgebonden Onderzoek in the Netherlands (2013/455); 4) UMM Universitätsmedizin Mannheim, Medizinische Ethik Kommission II (UMM University Medical Mannheim, Medical Ethics Commission II) in Germany (2014-540N-MA); and 5) Universita Campus Bio-Medica De Roma Comitato Etico (University Campus Bio-Medical Ethics Committee De Roma) in Italy (18/14 PAR ComET CBM).
Participant informed consent was obtained in keeping with the Declaration of Helsinki. All methods were performed in accordance with the relevant guidelines and regulations, where applicable.
Participants
We included 150 autistic (27% female) and 159 non-autistic (33% female) participants, aged 6–31 years, from five sites of the AIMS-2-TRIALS Longitudinal European Autism Project [59, 60], Table 1a. Autistic participants had an existing clinician-made diagnosis. Included participants had at least one 20 s epoch of good-quality eyes-closed resting-state EEG data and a good-quality T1 structural MRI image (for source reconstruction). Given that mental health and other developmental conditions commonly co-occur with autism [61], we did not exclude participants with these conditions. We excluded participants with neurological conditions, gross structural brain lesions, diagnosed intellectual disability, a full-scale IQ < 75 or significant systemic disease.Table 1Demographics and clinical characteristics of participants.
a) Demographics and clinical characteristics across all age-brackets
Autistic (n = 150) Non-autistic (n = 159) p Total (n = 309)
Demographics
Mean age (SD) – yrs 17.1 (5.8) 17.4 (5.9) 0.734 17.3 (5.8)
Age range – yrs 7.9 – 30.3 6.9 – 31.0 - 6.9 – 31.0
Female sex 41 (27%) 53 (33%) 0.252 94 (30%)
Full-scale IQ
Mean FSIQ (SD) 104 (15) 109 (13) 0.007 106 (14)
FSIQ range 75 – 148 75 – 142 - 75 – 148
Social Responsiveness Scale – 2
Mean SRS T-score (SD) 70 (12) 45 (5) < 0.001 59 (16)
Mental health diagnoses
Depression 11 / 135 (8%) 1/145 (1%) 0.002 12/280 (4%)
Anxiety 17 / 135 (13%) 2/146 (1%) < 0.001 19/281 (7%)
OCD 6 / 134 (5%) 0/146 (0%) 0.010 6/280 (2%)
Developmental diagnoses
ADHD 28 / 134 (21%) 5/146 (3%) < 0.001 33/280 (12%)
Dyslexia / dyscalculia 5 / 56 (9%) 5/51 (10%) 0.877 10/107 (9%)
Dyspraxia 2 / 56 (4%) 1/51 (2%) 0.614 3/107 (3%)
GDD (previous) 6 / 55 (11%) 1/51 (2%) 0.064 7/106 (7%)
b) Demographics and clinical characteristics within age-brackets
6–14-years 15–24-years 25–31-years
A(n = 62) NA(n = 54) p A(n = 74) NA(n = 87) p A(n = 14) NA(n = 18) p
Female 20 (32%) 20 (37%) 0.59 16 (22%) 27 (31%) 0.18 5 (36%) 6 (33%) 0.88
Mean FSIQ (SD) 104 (15) 111 (12) 0.01 104 (14) 107 (13) 0.12 105 (15) 107 (13) 0.75
FSIQ range 77–139 76–139 - 75–148 76–142 - 83–136 85–128 -
Mean SRS T-score (SD) 74 (11) 44 (5) < 0.01 67 (12) 45 (5) < 0.01 61 (9) 42 (1) 0.02
Mental health diagnoses
Depression 1/60 (2%) 0/51 (0%) 0.35 9/66 (14%) 0/77 (0%) < 0.01 1/9 (11%) 1/17 (6%) 0.63
Anxiety 5/60 (8%) 0/51 (0%) 0.04 9/66 (14%) 1/78 (1%) < 0.01 3/9 (33%) 1/17 (6%) 0.07
OCD 2/60 (3%) 0/51 (0%) 0.19 3/65 (5%) 0/78 (0%) 0.06 1/9 (11%) 0/17 (0%) 0.16
Developmental diagnoses
ADHD 15/60 (25%) 3/51 (6%) < 0.01 11/65 (17%) 2/78 (3%) < 0.01 2/9 (22%) 0/17 (0%) 0.04
Dyslexia / Dyscalculia 1/28 (4%) 2/20 (10%) 0.36 4/28 (14%) 2/27 (7%) 0.41 0/0 (-) 1/4 (25%) -
Dyspraxia 0/28 (0%) 0/20 (0%) 2/28 (7%) 0/27 (0%) 0.16 0/0 (-) 1/4 (25%) -
GDD (previous) 2/28 (7%) 0/20 (0%) 0.22 4/27 (15%) 1/27 (4%) 0.16 0/0 (-) 0/4 (0%) -
FSIQ full-scale IQ, OCD obsessive compulsive disorder, ADHD attention deficit / hyperactivity disorder, GDD global developmental delay, A autistic, NA non-autistic.
Age-groups
For the age-group analyses, participants were split into 3 age-groups (Table 1b): children (6–14 years, n = 116), youth (15–24 years, n = 161), and adults (25–31 years, n = 32). Youth was defined as per the United Nations World Health Organisation [62, 63]. This age-based definition aligns with a period of disruptive maturation of functional connectivity, in contrast to the conservative maturation seen in younger and older ages [64, 65].
Behavioural measures
We measured adaptive function in autistic people at two timepoints (T1 and T2) using the Vineland Adaptive Behaviour Scale (VABS), second edition [66]. The VABS is a semi-structured interview, which we conducted with participants’ parent or carer. Adaptive function across the domains of communication, daily living and socialisation skills was quantified by the adaptive behaviour composite score (ABC), standardised for age (mean 100, SD 15). Higher scores correspond with more age-normed adaptive behaviours. We calculated longitudinal changes in adaptive function (ΔABC) by subtracting T2 from T1 ABC scores, as done previously [14, 20].
The age-standardised ‘T-scores’ of the Social Responsiveness Scale-Second Edition (SRS-2; 67) were used as trait measures of autistic behaviours. Higher T-scores correspond with more autistic behaviours. We used the parent-reported instead of self-reported T-scores in our main analysis, as we had a larger sample size with the former (n = 118 versus n = 76) and the two measures are not interchangeable [68], Fig. S1. Intellectual ability was assessed using the Wechsler Abbreviated Scales of Intelligence-Second Edition [69], or shortened versions of the German, Dutch or Italian Wechsler Intelligence Scale for Children, third or fourth editions [70, 71] or Wechsler Adult Intelligence Scale third or fourth editions [72, 73].
Electroencephalography methods
Four segments of 30 s, eyes-closed recordings were captured using 60–70 EEG electrodes (site dependent). We transformed EEG timeseries from sensor space to source space using individual-participant T1-weighted MRI images and a linearly constrained minimum variance beamformer [74], as per Garcés et al. [75].
We chose 20 s eyes-closed resting-state EEG epochs to produce functional connectivity matrices. This epoch length was a compromise between maximising the number of included participants and maximising network stability; EEG-functional-connectivity-derived brain-network organisation has been shown to be stable at epoch lengths of 10–12 s and above [76, 77]. We filtered signals between 6–9 Hz (‘low-alpha’).
Functional connectivity network analyses
For each 20 s epoch, we calculated functional connectivity between brain-regions using the phase locking value (PLV; 78). Scalp-EEG-derived PLV has high test-retest reliability, concordance with its underlying structural connectivity and is sensitive to age-related connectivity changes [79].
We created weighted, undirected functional connectivity matrices with 68 nodes. Each node corresponded to a cortical region of the Desikan-Killiany atlas [80], while the edges were the PLV between pairs of nodes. The network metrics of interest were the mean degree and the small-world index, calculated using the brain connectivity toolbox [81].
The degree of each node was averaged over 68 nodes to obtain the mean degree, MDMD:1\documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$${MD}=\frac{1}{N} {\sum }_{i}{k}_{i},$$\end{document}MD=1N∑iki,where\documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$\,{k}_{i}$$\end{document}ki denotes the number of edges of node ii and i=1,… ,N=68i=1,…,N=68.
We used a modified (for weighted networks) Humphries and Gurney [82] definition of the weighted SWI [83]. The weighted SWI, σσ, was calculated by dividing the normalised weighted clustering coefficient, \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$C^{\prime}$$\end{document}C′, by the normalised weighted pathlength, \documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$L^{\prime}$$\end{document}L′:2\documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$\sigma =\frac{C^{\prime} }{L^{\prime} }$$\end{document}σ=C′L′
If a participant had more than one usable 20 s EEG epoch, MD and weighted SWI were calculated for each epoch. Each participant had 1–4 usable epochs. The per-participant connectivity metrics used in analyses were calculated from the mean of 1–4 epochs, as the mean of multiple measures is more reliable than a single measure [84].
Test-retest reliability
We quantified the intrasession test-retest reliability between two epochs recorded between 1–3 min of each other. Reliability was calculated for metrics derived from one 20 s epoch using interclass correlation coefficient (ICC (2,1)) and metrics derived from the mean of two 20 s epochs using ICC(2,k). Results were interpreted as per Ku and Li [85].
Genetics
We calculated polygenic likelihood scores (PGS) using blood samples. Using reference genome-wide association study summary statistics, PGS were produced for autism [86] and brain volume [87]. A more positive brain volume PGS is associated with more alleles for large brain volume.
General linear models predicting longitudinal changes in adaptive function
Our primary aim was to investigate if functional connectivity metrics can predict longitudinal changes in ΔABC and if prediction ability varies by age. Longitudinal ABC data was available for 40 autistic children (6–14-years), 38 autistic youth (15–24-years) and 3 autistic adults (25–31-years), Fig S2. Therefore, we omitted the adults from the age-bracket analyses predicting ΔABC.
We used general linear models to predict ΔABC, adjusting them for T1 ABC, participant age and an age x T2 – T1 (follow-up interval) interaction term (as per Pretzsch et al. [20] – as the rate of behavioural change differs by developmental stage; 45). We also used these models to determine the presence of significant age x group interaction effects and adjust for the effect of nuisance covariates (e.g. site and sex). We retained interaction effects when p < 0.05 for the interaction term [88]. We selected general linear models by comparing the Akaike information criterion (AIC) and the adjusted R2 between models with different nuisance co-variates. We evaluated the out-of-sample generalisability of the models using the connectome-based predictive modelling protocol [89].
In models where there was a main effect of functional connectivity metric on ΔABC, we explored the ability of our models to predict categorical ΔABC on a per-participant basis, using leave-one-out cross-validation. We used an ΔABC cutoff of 6 to define improving, worsening or static outcomes. This cutoff was based on evidence that ΔABC greater than ±2.7–8 [90, 91] in autistic people without intellectual disability are minimal clinically-important differences and are unlikely (p < 0.05) to occur by chance given the standard error of measurement of the ABC [90, 92, 93]. We assessed if predictive ability was better than chance using permutation analyses. Then, to illustrate predictive performance using the area under the receiver operated characteristic curve (AUC), sensitivity and specificity, we binarized outcomes to increased (≥ 6) or not-increased (<6) ABC scores. 95% confidence intervals for AUC were quantified using bootstrapping.
Finally, we illustrated the relationship between effect size and age without pre-defined age brackets. We plotted a moving estimated coefficient of the functional connectivity metric predicting ΔABC as a function of a moving average participant age, as per O’Muircheartaigh et al. [94]. To obtain estimated coefficients, we performed identical general linear models in participant groups of n = 30, ordered by participant age, increasing by 5 participants at a time.
Results
EEG epochs
There were no differences between the number of analysable EEG epochs per participant between autistic and non-autistic people (W = 11362, p = 0.66). Participants with more usable epochs had a greater mean age H(3) = 8.9, p = 0.03 (Fig. S3).
Cross-sectional group differences in functional connectivity
We quantified the extent of cortico-cortical functional connectivity using mean degree. Age-associated increases in MD were slower in autistic than in non-autistic people (β = −0.19, SE = 0.10, p = 0.045, η2 = 0.01), Fig. 1a, Table S1. Compared to non-autistic people, MD was no different in autistic 6–14-year-olds (β = 0.11, SE = 0.18, pFDR = 0.52) and 25–31-year-olds (β = −0.32, SE = 0.45, pFDR = 0.52) but lower in autistic 15–24-year-olds (β = −0.46, SE = 0.15, pFDR = 0.005), Fig. 1c. In autistic people aged 6–31-years, the presence of more parent-reported autistic features (SRS T-score) correlated with lower MD, even when adjusting for site effects (partial r = −0.31, n = 118, p < 0.001), Fig. 1e.
Age-associated decreases in SWI were slower in autistic than in non-autistic people (β = 0.21, SE = 0.10, p = 0.039, η2 = 0.01), Table S2. We found no differences in SWI in autistic 6–14-year-olds (β = −0.12, SE = 0.18, pFDR = 0.51), 15–24-year-olds (β = 0.34, SE = 0.16, pFDR = 0.10) or 25–31-year-olds (β = 0.62, SE = 0.34, pFDR = 0.13) compared to non-autistic people, Fig. 1d. Within autistic people aged 6–31-years, there was a positive correlation between parent-reported SRS T-score and SWI, even when adjusting for site effects (partial r = 0.30, n = 118, p < 0.001), Fig. 1f. Note that self-reported SRS T-scores did not correlate with MD or SWI (Fig. S4).
Functional connectivity predicts longitudinal changes in adaptive function
From the sample of 150 autistic people, we collected ABC score data at T1 from 122 people, at T2 from 90 people, and at both timepoints from 81 people, Table 2.Table 2Vineland Adaptive Behaviour Composite (ABC) Scores by timepoint and age-bracket.
6–14 years 15–24 years 25–31 years 6–31 years
Mean ABC T1 (SD) 77.9 (9.9) n = 48 68.3 (13.1) n = 63 79.1 (22.0) n = 11 73.0 (13.8) n = 122
Range 58 – 100 30 – 104 52 – 121 30 – 121
Mean ABC T2 (SD) 77.2 (10.4) n = 46 72.8 (12.9) n = 41 81.3 (1.2) n = 3 75.3 (11.6) n = 90
Range 59 – 102 20 – 95 80 – 82 20 - 102
Mean ABC change (SD) −1.6 (9.0) n = 40 3.4 (10.8) n = 38 4.0 (10.8) n = 3 0.95 (10.1) n = 81
Range −23 - +26 −37 - +30 −5 - +16 −37 - +30
Mean follow-up interval (SD) 18.9 (3.9) months 20.7 (2.3) months 18.0 (1.4) months 19.7 (3.3) months
T1 timepoint 1, T2 timepoint 2.
Cross-sectional analyses between functional connectivity and adaptive function
First, we examined if MD and SWI were associated with T1 adaptive function abilities in autistic people. Lower MD and higher SWI were associated with lower T1 ABC across 6–31-year-olds (Tables S3-S4).
Longitudinal analyses between functional connectivity and adaptive function
Next, we assessed the association between T1 functional connectivity metrics and longitudinal changes in adaptive function. The mean follow-up interval between T1 and T2 ABC measurements across all autistic participants was 19.7 months (SD = 3.5 months), Table 2, Fig. S5a. ΔABC ranged from −37 to +30, Table 2, Fig. S5b. The 25–31-year-old group only had 3 participants with longitudinal data and were excluded from age-bin analyses. The mean follow-up interval was 1.8 months (SD = 3.9 months) longer in 15–24-year-olds compared to 6–14-year-olds, t(68.5) = 2.3, p = 0.02. There were no differences in the variance of ΔABC between age-groups (F2/78 = 0.61, p = 0.55).
MD did not predict ΔABC, when 6–31-year-old autistic people were analysed together (Fig. 2a, Table S5). MD did not predict ΔABC in 6–14-year-old autistic people, Table S6.
In 15–24-year-old autistic people, higher MD was associated with improvements in ABC, with a large effect size (β = 5.6, SE = 1.5, pFDR < 0.001, ppermute = 0.004, η2 = 0.13), Fig. 2c, Table S7. Adding the variable MD to a model predicting ΔABC (already containing full-scale IQ, T1 ABC, age and sex) increased the variance explained in ΔABC by 21% (model R2adj increased from 0.29–0.50).
SWI predicted ΔABC when 6–31-year-old autistic people were analysed together (β = −2.4, SE = 1.0, p = 0.023, ppermute = 0.024, η2 = 0.04), Fig. 2b, Table S8. Adding the variable SWI to a model predicting ΔABC (already containing full-scale IQ, T1 ABC and age) increased the variance explained in ΔABC by 4% (model R2adj increased from 0.30–0.34).
SWI did not predict ΔABC in 6–14-year-old autistic people, Table S9.
In 15–24-year-old autistic people, lower SWI was associated with improved ABC, with a large effect size (β = −7.5, SE = 1.6, pFDR < 0.001, ppermute = 0.003, η2 = 0.18), Fig. 2d, Table S10. Adding the variable SWI to a model predicting ΔABC (already containing full-scale IQ, T1 ABC, age and sex) increased the variance explained in ΔABC by 30% (model R2adj increased from 0.29–0.59).
Associations between actual and predicted values of ΔABC for the three models that significantly predicted ΔABC are shown in Figure S6. We assessed the performance of these models on ‘unseen’ participants using 10-fold cross-validation. The cross validation R2s were between 0.39 and 0.62 (Table S11).
A plausible trivial explanation for different effect sizes between the two age-brackets could be a between-age-bracket difference in the variances of SWI, MD or ΔABC. A post hoc Levene’s test showed that this was not the case (ΔABC: F1/76 = 0.20, p = 0.66; SWI: F1/76 = 1.4, p = 0.23; MD: F1/76 = 0.50, p = 0.48). Additionally, we examined post hoc if the strong association between functional connectivity and ΔABC in 15–24-year-olds was specific to ΔABC, or if it was also found in other brain-behaviour associations. Associations between SRS T-score and SWI/MD were stronger in younger participants (Figure S7).
As all age-bracket definitions are somewhat arbitrary, we illustrated the relationship between prediction ability for ΔABC and age without pre-defined age-brackets. We plotted a moving estimated coefficient of the functional connectivity metric predicting ΔABC as a function of average age. There were main effects of MD and SWI in models predicting ΔABC when the mean age in participant bins was between 18.6 (range 15.2–23.0)–21.8 (range 17.5–30.3) years, the latter being the age-range ceiling in our sample, Fig. 2e,f, Table S12. This is consistent with better prediction ability in autistic youth compared to autistic children.
Comparison against other prognostic variables
We compared the predictive performance of MD and SWI to that of autistic features and full-scale IQ. For the models with a significant main effect of functional connectivity metric on adaptive function outcomes, MD and SWI outperformed both behavioural predictors (Supplementary Results and Table S13).
Categorical prediction of adaptive function changes
We explored the ability of the three significant regression models to categorise ΔABC (into improved, worsened and static outcomes) on a per-participant basis, using leave-one-out cross-validation. The categorisation accuracy of MD in 15–24-year-olds was 63.2%, ppermute = 0.002; SWI in 15–24-year-olds was 65.8%, ppermute = 0.001; SWI in 6–31-year-olds was 60.4%, ppermute < 0.001 (Fig. 3a). Thus, all models categorised ΔABC significantly better than chance. Then, we explored model performance in performing binary categorisation into increased (≥ 6) or not-increased (< 6) ΔABC. For the model containing MD in 15–24-year-olds, AUC = 0.84 [95% CI: 0.71–0.97], sensitivity = 0.73, specificity = 0.74; for SWI in 15–24-year-olds, AUC = 0.78 [95% CI: 0.63–0.93]; sensitivity = 0.73; specificity = 0.78; for SWI in 6–31-year-olds, AUC = 0.76 [95% CI: 0.63- 0.89]; sensitivity = 0.43; specificity = 0.93 (Fig. 3b).
Properties desired in candidate prognostic biomarkers
MD and SWI had good test-retest reliability (n = 203), both when derived from one epoch and from the mean of 2 epochs. For MD, ICC (2,1) = 0.78 [95% CI: 0.72–0.82] and ICC (2,k) = 0.88 [95% CI: 0.84–0.91]. For SWI, ICC (2,1) = 0.77 [95% CI: 0.71–0.82] and ICC (2,k) = 0.87 [95% CI: 0.83–0.90]. In the three models that significantly predicted ΔABC in autistic people, we performed a sensitivity analysis using MD/SWI derived from only one (instead of the mean of one-to-four) 20 s EEG epoch per participant, Fig. S8. The associations between MD/SWI and ΔABC were similar when only one epoch was used, although expectedly smaller in magnitude – consistent with the ICC values.
To assess for convergence with biological processes, we investigated if MD and SWI were associated with polygenic variation in autism- and brain-volume- related genes. We found no associations between polygenic scores (PGS) for autism and MD/SWI, Tables S14-S15.
In autistic people, PGS for brain volume was associated with variance in MD (n = 136, partial r2 = 0.073, pFDR = 0.004) and SWI (n = 136, partial r2 = 0.090, pFDR = 0.002), even when adjusting for age and sex, Fig. 4. Notably, no association between PGS for brain volume and MD/SWI was found in the non-autistic group, despite similar variances in PGS and MD/SWI between groups, Table S16.
Given the association between PGS for brain volume and functional connectivity metrics, we confirmed post hoc that ΔABC was not trivially associated with head circumference (a proxy of brain volume; 95), Tables S17-S19.
Discussion
We provide the first evidence that EEG-derived functional connectivity metrics can predict longitudinal changes in adaptive function in autistic people. By explicitly considering age-effects in a large, multi-site cohort, we demonstrated that mean degree and small-world index explain 21–30% of additional variance in outcomes in autistic youth (aged 15–24-years). SWI also significantly predicted adaptive function outcomes across autistic people aged 6–31 years, although it only explained 4% of additional variance in this age-bracket. Both metrics showed properties desired in prognostic biomarkers: convergence with underlying neurobiology (polygenic variation in brain volume) and high test-retest reliability.
Conflicting findings from previous autism functional connectivity studies has limited the development of functional connectivity metrics as biomarkers. Our cross-sectional analyses suggest that considering developmental stage in autism functional connectivity analyses may reconcile some of this heterogeneity, as has also been argued by Uddin et al. [40].
Utility
In 15–24-year-old autistic people and 6–31-year-old autistic people (for SWI only), we found that EEG-derived MD/SWI were better predictors of changes in adaptive function than were behavioural predictors: autistic features and/or full-scale IQ. In 15–24-year-olds, our metrics predicted greater variance in changes in adaptive function (21–30%) than the N170 [14] or cortical structure [20], which both explained 6% of additional variance. Our study expands on an important but small (n = 27) fMRI functional connectivity study in a homogenous sample of autistic males, which also predicted similar variance in outcomes [17].
In categorising outcomes as increased, static and decreased on a per-person basis, based on minimal-clinically important differences [90, 91] in ABC scores, models containing MD and SWI performed significantly better than chance, even when including participants across the 6–31-year age-range. Further, in binary categorisation into “increase” versus “no-increase” groups, the model containing MD in 15–24-year-old autistic people had an AUC of 0.84 and the model containing SWI in 6–31-year-old autistic people had an AUC of 0.76, suggesting promise as prognostic biomarkers [96].
Together, these findings suggest that EEG-derived MD and SWI may be developed as prognostic biomarkers for both continuous and categorical changes in adaptive function, especially for autistic people aged 15–24-years. In clinical contexts, this may help identify groups at risk of declining adaptive function during a key transitional period of development – enabling early, targeted interventions of adaptive functioning (e.g. the TEACCH School Transition to Employment and Postsecondary Education Program; 97). In clinical trials aiming to improve adaptive function [10, 11, 98], these biomarkers may be used to stratify participants by their likely spontaneous outcomes. This could reduce trial costs and sample size requirements [14].
Brain-behaviour associations in different age groups
We identified the age-bracket of 15–24-years as that where MD and SWI are particularly effective predictors of adaptive function outcomes. The question is, why? Firstly, during youth, precision in the development of functional connectivity could be particularly important for behavioural outcomes. Therefore, variation in functional connectivity between youth may have a large effect on outcomes. Consistently, cortical and subcortical functional connectivity organisation [99] – including that involved in social cognition [64, 65, 100] – is significantly disrupted and remodelled between the ages of 14–26 years [64, 65]. In other age-groups, conservative maturation (where strong connections become stronger and weak connections become weaker) may be more pervasive [64, 99]; therefore, variation across individuals may have a less profound impact. Secondly, functional connectivity derived MD and SWI may be key drivers of adaptive function outcomes in late adolescence, whereas other (e.g. behavioural) features could be more important at other ages. Thirdly, the effect size of MD differences between autistic and non-autistic people was greatest in the 15–24-year age-group. That connectivity is most different from non-autistic people in this age-group may also suggest that it is the most relevant.
Additionally, we consider two explanations that might arise due to peculiarities of our dataset. It could be that brain-behaviour associations in general are strongest in the 15–24-year age-group. If true, associations between MD/SWI and other behavioural features (e.g. SRS T-scores) would also be higher in this age-group than others. We found that this was not the case. Next, it could be that the variances in ΔABC, MD or SWI between age-groups were different, allowing associations to be found in some but not other groups. This was also not found to be the case.
The direction of relationships between functional connectivity and phenotype
MD was lower in autistic than non-autistic youth, in keeping with prior literature [40]. Lower MD was associated with more autistic features and worsening of age-normed adaptive function over time. Regarding SWI, previous studies using heterogeneous methods and age-groups found both decreased [30] and increased [31] small-world network organisation in autistic compared to non-autistic children. We found that greater SWI was associated with more autistic features and worsening of age-normed adaptive function over time. However, network models postulate that higher SWI reflects an optimum balance between network integration and segregation [34]. What may explain this apparent contradiction? Studies in other brain-based conditions have shown that less small-world organisation is associated with more favourable phenotypes [101, 102]. This may suggest that non-optimal connectivity between nodes may be protective in the context of altered function within nodes. Further, the argument that higher SWI equates to ‘better’ network organisation assumes constant MD [36]; this was not the case across participants in our study (see limitations).
Genetic associations with MD and SWI
Polygenic scores for brain volume [87] explained a high proportion of variance (in the context of polygenic scores; 14, 103) in MD and SWI in autistic people. Whole brain, white matter and cortical volume differences have been repeatedly found in autistic people [20, 43, 44]. Twin studies have shown that the heritability of MD is approximately 37% [104], while SWI has a heritability of approximately 50% [51, 52]. Our findings suggest that the effects of genotype on MD and SWI could explained by polygenic variation in genes contributing to brain volume.
Interestingly, the associations between PGS for brain volume and MD/SWI were only found in autistic people. This may suggest that genetic influences on brain volume affect MD and SWI more strongly in the co-presence of genetic and environmental factors that autistic people (and not non-autistic people) are exposed to. It is plausible that genetic variants for brain volume and autistic features interact in complex ways.
Study strengths and limitations
Our study design had many strengths. We included autistic people at 5 study sites with co-occurring developmental and mental health conditions – reflecting autistic people in the general population [61]. Thus, our findings may generalise to a large group of autistic people. Second, we used the Vineland Adaptive Behaviour Scale [66, 105] – shown to have high test-retest reliability [66], convergent validity and internal consistency [106] – to measure adaptive function. Third, EEG is a widely available and affordable measure of functional connectivity [55], further improving translation potential.
We showed the robustness of MD/SWI as prognostic biomarkers. MD showed sensitivity to diagnostic group and both MD and SWI varied with the extent of parent-reported autistic features. Further, MD/SWI were associated with current adaptive function in the same direction as they were associated with changes in adaptive function. The models predicting longitudinal changes in adaptive function explained a high proportion of variance in outcome even in ‘unseen’ autistic participants, as shown by 10-fold-cross-validation. Finally, we demonstrated that strong predictive performance in autistic youth was not dependent on arbitrary age-bin definitions, finding significant predictive ability of MD/SWI across several age-bins that included autistic youth.
Our study has important limitations. Although our overall sample was large, we had a modest sample size of autistic youth with longitudinal Vineland data (n = 38). Further, we had few participants in the adult group (n = 32); therefore, group-difference analyses for MD/SWI in this age-group are likely underpowered. Next, we used the Humphries [82] method of quantifying SWI, which is affected by network density (MD; 36). Networks with higher MD had lower SWI (Fig. S9). Therefore, the two network metrics used are not independent. That said, this is true for most network metrics [107, 108]. Previous research has addressed this issue by thresholding network density across participants [109]. However, this has the effect of excluding weaker functional connections. We decided against this approach as weak functional connections play an important role in cognition [110, 111].
As expected, participants with more usable EEG epochs had a higher mean age. Therefore, the functional connectivity metrics, which were averaged across epochs, were more reliable in older participants [84]. This is reassuring for our findings in autistic youth, but our negative findings in autistic children should be interpreted with some caution.
It is plausible that the ABC is measuring different adaptive behaviour constructs at different ages. For instance, the relationship between ABC and intelligence is stronger in children than in adults [18]. If this is the case, our differential findings between children and youth could represent differences is measurement of adaptive behaviour rather than differences in brain-behaviour associations. Further work is needed to clarify the relevance of this limitation of the Vineland.
Studies of MCID in the ABC score [90, 91] have predominantly included children; therefore, generalising MCID values to autistic adults should be done cautiously. Here, we used a relatively conservative MCID value of 6, consistent with the reliable change index of the ABC score.
We used chronological age as a proxy of developmental status. Developmental stage could have been captured more accurately through physiological-based measures such as Tanner (pubertal) staging; however, this would have also increased participant burden.
Future work
Adaptive function skills are the amalgamation of many cognitive functions and their interactions with the environment. Future work could explore the cognitive abilities that mediate the effect of MD and SWI on longitudinal changes in adaptive function in autistic people. For example, recent findings from our group suggest that the organisation of EEG-derived large-scale functional connectivity networks is associated with complex facial emotion recognition in an age-sensitive manner [112]. Similarly, future work should investigate if associations between MD/SWI and ΔABC are driven by certain adaptive function subdomains (i.e. communication, daily living or socialisation skills).
Most autistic people have a co-occurring mental health or neurodevelopmental condition [113], known to independently impact adaptive function [114]. Future work could investigate relationships between co-occurring conditions, functional connectivity and longitudinal changes in adaptive function. Our sensitivity analysis without participants with ADHD, the most frequent co-occurring condition in our sample, found that relationships between MD/SWI and ΔABC remained consistent (Fig. S10).
The next steps in developing MD and SWI as prognostic biomarkers of adaptive function include assessing its performance in external, larger datasets [46]. If such statistical validation is achieved, a randomised control trial is needed to investigate if prognostication improves the quality of life and/or employment, education and independent-living outcomes for autistic youth.
Conclusion
Our study is the first to demonstrate that EEG-derived functional connectivity can explain a significant proportion of the variance in longitudinal changes in adaptive function in autistic people. Therefore, EEG-derived mean degree and small-world index may be developed as prognostic biomarkers for adaptive function outcomes, especially for autistic youth. Potential precision-medicine applications include stratifying autistic people in clinical trials and identifying those at risk of declining adaptive function in clinical settings. Finally, our age-sensitive approach may reconcile the heterogenous findings in previous autism functional connectivity studies.
Supplementary information
References
1.Farley MA McMahon WM Fombonne E Jenson WR Miller J Gardner M Twenty-year outcome for individuals with autism and average or near-average cognitive abilities Autism Research 2009 2 109 18 10.1002/AUR.69 19455645 Farley MA, McMahon WM, Fombonne E, Jenson WR, Miller J, Gardner M, et al. Twenty-year outcome for individuals with autism and average or near-average cognitive abilities. Autism Research. 2009;2:109–18. 10.1002/AUR.69 19455645
2.Farley M Cottle KJ Bilder D Viskochil J Coon H McMahon W Mid-life social outcomes for a population-based sample of adults with ASD Autism Res 2018 11 142 52 10.1002/AUR.1897 29266823 PMC5924705 Farley M, Cottle KJ, Bilder D, Viskochil J, Coon H, McMahon W. Mid-life social outcomes for a population-based sample of adults with ASD. Autism Res. 2018;11:142–52. 10.1002/AUR.1897 PMC5924705 29266823
3.Forbes G Kent R Charman T Baird G Pickles A Simonoff E How do autistic people fare in adult life and can we predict it from childhood? Autism Research. 2023 16 458 73 10.1002/AUR.2868 36519265 PMC10947100 Forbes G, Kent R, Charman T, Baird G, Pickles A, Simonoff E. How do autistic people fare in adult life and can we predict it from childhood? Autism Research. 2023;16:458–73. 10.1002/AUR.2868 PMC10947100 36519265
4.Roche L Adams D Clark M Research priorities of the autism community: a systematic review of key stakeholder perspectives Autism 2020 25 336 48 10.1177/1362361320967790 33143455 Roche L, Adams D, Clark M. Research priorities of the autism community: a systematic review of key stakeholder perspectives. Autism. 2020;25:336–48. 10.1177/1362361320967790 33143455
5.Pellicano E Dinsmore A Charman T What should autism research focus upon? Community views and priorities from the United Kingdom Autism. 2014 18 756 70 10.1177/1362361314529627/ASSET/IMAGES/LARGE/10.1177_1362361314529627-FIG2.JPEG 24789871 Pellicano E, Dinsmore A, Charman T. What should autism research focus upon? Community views and priorities from the United Kingdom. Autism. 2014;18:756–70. 10.1177/1362361314529627/ASSET/IMAGES/LARGE/10.1177_1362361314529627-FIG2.JPEG PMC4230972 24789871
6.de Bildt A Sytema S Kraijer D Sparrow S Minderaa R Adaptive functioning and behaviour problems in relation to level of education in children and adolescents with intellectual disability J Intellect Disabil Res 2005 49 672 81 10.1111/J.1365-2788.2005.00711.X 16108984 de Bildt A, Sytema S, Kraijer D, Sparrow S, Minderaa R. Adaptive functioning and behaviour problems in relation to level of education in children and adolescents with intellectual disability. J Intellect Disabil Res. 2005;49:672–81. 10.1111/J.1365-2788.2005.00711.X 16108984
7.Howlin P Alcock J Burkin C An 8 year follow-up of a specialist supported employment service for high-ability adults with autism or asperger syndrome Autism 2005 9 533 49 10.1177/1362361305057871 16287704 Howlin P, Alcock J, Burkin C. An 8 year follow-up of a specialist supported employment service for high-ability adults with autism or asperger syndrome. Autism. 2005;9:533–49. 10.1177/1362361305057871 16287704
8.Magiati I Tay XW Howlin P Cognitive, language, social and behavioural outcomes in adults with autism spectrum disorders: a systematic review of longitudinal follow-up studies in adulthood Clin Psychol Rev 2014 34 78 86 10.1016/J.CPR.2013.11.002 24424351 Magiati I, Tay XW, Howlin P. Cognitive, language, social and behavioural outcomes in adults with autism spectrum disorders: a systematic review of longitudinal follow-up studies in adulthood. Clin Psychol Rev. 2014;34:78–86. 10.1016/J.CPR.2013.11.002 24424351
9.Liss M Harel B Fein D Allen D Dunn M Feinstein C Predictors and correlates of adaptive functioning in children with developmental disorders J Autism Dev Disord 2001 31 219 30 10.1023/A:1010707417274/METRICS 11450820 Liss M, Harel B, Fein D, Allen D, Dunn M, Feinstein C, et al. Predictors and correlates of adaptive functioning in children with developmental disorders. J Autism Dev Disord. 2001;31:219–30. 10.1023/A:1010707417274/METRICS 11450820
10.Sikich L Kolevzon A King BH McDougle CJ Sanders KB Kim SJ Intranasal oxytocin in children and adolescents with autism spectrum disorder New England Journal of Medicine 2021 385 1462 73 10.1056/NEJMOA2103583/SUPPL_FILE/NEJMOA2103583_DATA-SHARING.PDF 34644471 Sikich L, Kolevzon A, King BH, McDougle CJ, Sanders KB, Kim SJ, et al. Intranasal oxytocin in children and adolescents with autism spectrum disorder. New England Journal of Medicine. 2021;385:1462–73. 10.1056/NEJMOA2103583/SUPPL_FILE/NEJMOA2103583_DATA-SHARING.PDF PMC9701092 34644471
11.Jacob S Veenstra-VanderWeele J Murphy D McCracken J Smith J Sanders K Efficacy and safety of balovaptan for socialisation and communication difficulties in autistic adults in North America and Europe: a phase 3, randomised, placebo-controlled trial Lancet Psychiatry 2022 9 199 210 10.1016/S2215-0366(21)00429-6 35151410 Jacob S, Veenstra-VanderWeele J, Murphy D, McCracken J, Smith J, Sanders K, et al. Efficacy and safety of balovaptan for socialisation and communication difficulties in autistic adults in North America and Europe: a phase 3, randomised, placebo-controlled trial. Lancet Psychiatry. 2022;9:199–210. 10.1016/S2215-0366(21)00429-6 35151410
12.Seltzer MM Shattuck P Abbeduto L Greenberg JS Trajectory of development in adolescents and adults with autism Ment Retard Dev Disabil Res Rev 2004 10 234 47 10.1002/MRDD.20038 15666341 Seltzer MM, Shattuck P, Abbeduto L, Greenberg JS. Trajectory of development in adolescents and adults with autism. Ment Retard Dev Disabil Res Rev. 2004;10:234–47. 10.1002/MRDD.20038 15666341
13.Szatmari P Georgiades S Duku E Bennett TA Bryson S Fombonne E Developmental trajectories of symptom severity and adaptive functioning in an inception cohort of preschool children with autism spectrum disorder JAMA Psychiatry 2015 72 276 83 10.1001/JAMAPSYCHIATRY.2014.2463 25629657 Szatmari P, Georgiades S, Duku E, Bennett TA, Bryson S, Fombonne E, et al. Developmental trajectories of symptom severity and adaptive functioning in an inception cohort of preschool children with autism spectrum disorder. JAMA Psychiatry. 2015;72:276–83. 10.1001/JAMAPSYCHIATRY.2014.2463 25629657
14.Mason L Moessnang C Chatham C Ham L Tillmann J Dumas G Stratifying the autistic phenotype using electrophysiological indices of social perception Sci Transl Med 2022 14 eabf8987 10.1126/SCITRANSLMED.ABF8987 35976994 Mason L, Moessnang C, Chatham C, Ham L, Tillmann J, Dumas G, et al. Stratifying the autistic phenotype using electrophysiological indices of social perception. Sci Transl Med. 2022;14:eabf8987 10.1126/SCITRANSLMED.ABF8987 35976994
15.McGovern CW Sigman M Continuity and change from early childhood to adolescence in autism Journal of Child Psychology and Psychiatry 2005 46 401 8 10.1111/J.1469-7610.2004.00361.X 15819649 McGovern CW, Sigman M. Continuity and change from early childhood to adolescence in autism. Journal of Child Psychology and Psychiatry. 2005;46:401–8. 10.1111/J.1469-7610.2004.00361.X 15819649
16.Billstedt E Carina Gillberg I Gillberg C Autism in adults: symptom patterns and early childhood predictors. Use of the DISCO in a community sample followed from childhood Journal of Child Psychology and Psychiatry 2007 48 1102 10 10.1111/J.1469-7610.2007.01774.X 17995486 Billstedt E, Carina Gillberg I, Gillberg C. Autism in adults: symptom patterns and early childhood predictors. Use of the DISCO in a community sample followed from childhood. Journal of Child Psychology and Psychiatry. 2007;48:1102–10. 10.1111/J.1469-7610.2007.01774.X 17995486
17.Plitta M Barnes KA Wallace GL Kenworthy L Martin A Resting-state functional connectivity predicts longitudinal change in autistic traits and adaptive functioning in autism Proc Natl Acad Sci USA 2015 112 E6699 706 10.1073/pnas.1510098112 PMC4672806 26627261 Plitta M, Barnes KA, Wallace GL, Kenworthy L, Martin A. Resting-state functional connectivity predicts longitudinal change in autistic traits and adaptive functioning in autism. Proc Natl Acad Sci USA. 2015;112:E6699–706. 10.1073/pnas.1510098112 PMC4672806 26627261
18.Tillmann J San José Cáceres A Chatham CH Crawley D Holt R Oakley B Investigating the factors underlying adaptive functioning in autism in the EU-AIMS Longitudinal European autism project Autism Res 2019 12 645 57 10.1002/AUR.2081 30741482 Tillmann J, San José Cáceres A, Chatham CH, Crawley D, Holt R, Oakley B, et al. Investigating the factors underlying adaptive functioning in autism in the EU-AIMS Longitudinal European autism project. Autism Res. 2019;12:645–57. 10.1002/AUR.2081 PMC6519242 30741482
19.Kanne SM Gerber AJ Quirmbach LM Sparrow SS Cicchetti DV Saulnier CA The role of adaptive behavior in autism spectrum disorders: implications for functional outcome J Autism Dev Disord 2011 41 1007 18 10.1007/s10803-010-1126-4 21042872 Kanne SM, Gerber AJ, Quirmbach LM, Sparrow SS, Cicchetti DV, Saulnier CA. The role of adaptive behavior in autism spectrum disorders: implications for functional outcome. J Autism Dev Disord. 2011;41:1007–18. 10.1007/s10803-010-1126-4 21042872
20.Pretzsch CM Schäfer T Lombardo MV Warrier V Mann C Bletsch A Neurobiological correlates of change in adaptive behavior in autism Am J Psychiatry 2022 179 336 49 10.1176/appi.ajp.21070711 35331004 Pretzsch CM, Schäfer T, Lombardo MV, Warrier V, Mann C, Bletsch A, et al. Neurobiological correlates of change in adaptive behavior in autism. Am J Psychiatry. 2022;179:336–49. 10.1176/appi.ajp.21070711 35331004
21.Holiga Š Hipp JF Chatham CH Garces P Spooren W D’Ardhuy XL Patients with autism spectrum disorders display reproducible functional connectivity alterations Sci Transl Med 2019 11 eaat9223 10.1126/SCITRANSLMED.AAT9223 30814340 Holiga Š, Hipp JF, Chatham CH, Garces P, Spooren W, D’Ardhuy XL, et al. Patients with autism spectrum disorders display reproducible functional connectivity alterations. Sci Transl Med. 2019;11:eaat9223 10.1126/SCITRANSLMED.AAT9223 30814340
22.O’Reilly C Lewis JD Elsabbagh M Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies PLoS One 2017 12 e0175870 10.1371/JOURNAL.PONE.0175870 28467487 O’Reilly C, Lewis JD, Elsabbagh M. Is functional brain connectivity atypical in autism? A systematic review of EEG and MEG studies. PLoS One. 2017;12:e0175870 10.1371/JOURNAL.PONE.0175870 PMC5414938 28467487
23.Murias M Webb SJ Greenson J Dawson G Resting state cortical connectivity reflected in EEG coherence in individuals with autism Biol Psychiatry 2007 62 270 3 10.1016/j.biopsych.2006.11.012 17336944 Murias M, Webb SJ, Greenson J, Dawson G. Resting state cortical connectivity reflected in EEG coherence in individuals with autism. Biol Psychiatry. 2007;62:270–3. 10.1016/j.biopsych.2006.11.012 PMC2001237 17336944
24.Courchesne E Campbell K Solso S Brain growth across the life span in autism: age-specific changes in anatomical pathology Brain Res 2011 1380 138 45 10.1016/j.brainres.2010.09.101 20920490 PMC4500507 Courchesne E, Campbell K, Solso S. Brain growth across the life span in autism: age-specific changes in anatomical pathology. Brain Res. 2011;1380:138–45. 10.1016/j.brainres.2010.09.101 PMC4500507 20920490
25.Barttfeld P Wicker B Cukier S Navarta S Lew S Sigman M A big-world network in ASD: Dynamical connectivity analysis reflects a deficit in long-range connections and an excess of short-range connections Neuropsychologia 2011 49 254 63 10.1016/j.neuropsychologia.2010.11.024 21110988 Barttfeld P, Wicker B, Cukier S, Navarta S, Lew S, Sigman M. A big-world network in ASD: Dynamical connectivity analysis reflects a deficit in long-range connections and an excess of short-range connections. Neuropsychologia. 2011;49:254–63. 10.1016/j.neuropsychologia.2010.11.024 21110988
26.Courchesne E Pierce K Why the frontal cortex in autism might be talking only to itself: local over-connectivity but long-distance disconnection Curr Opin Neurobiol 2005 15 225 30 10.1016/J.CONB.2005.03.001 15831407 Courchesne E, Pierce K. Why the frontal cortex in autism might be talking only to itself: local over-connectivity but long-distance disconnection. Curr Opin Neurobiol. 2005;15:225–30. 10.1016/J.CONB.2005.03.001 15831407
27.Mehra C Absoud M Commentary on… the overlapping and distinct resting functional connectivity between autism spectrum disorder and attention-deficit hyperactivity disorder The British Journal of Psychiatry 2019 214 345 6 10.1192/BJP.2019.77 31014412 Mehra C, Absoud M. Commentary on… the overlapping and distinct resting functional connectivity between autism spectrum disorder and attention-deficit hyperactivity disorder. The British Journal of Psychiatry. 2019;214:345–6. 10.1192/BJP.2019.77 31014412
28.Oldehinkel M Mennes M Marquand A Charman T Tillmann J Ecker C Altered connectivity between cerebellum, visual, and sensory-motor networks in autism spectrum disorder: results from the EU-AIMS longitudinal european autism project Biol Psychiatry Cogn Neurosci Neuroimaging 2019 4 260 70 10.1016/J.BPSC.2018.11.010 30711508 Oldehinkel M, Mennes M, Marquand A, Charman T, Tillmann J, Ecker C, et al. Altered connectivity between cerebellum, visual, and sensory-motor networks in autism spectrum disorder: results from the EU-AIMS longitudinal european autism project. Biol Psychiatry Cogn Neurosci Neuroimaging. 2019;4:260–70. 10.1016/J.BPSC.2018.11.010 30711508
29.Ilioska I Oldehinkel M Llera A Chopra S Looden T Chauvin R Connectome-wide mega-analysis reveals robust patterns of atypical functional connectivity in autism Biol Psychiatry 2023 94 29 39 10.1016/J.BIOPSYCH.2022.12.018 36925414 Ilioska I, Oldehinkel M, Llera A, Chopra S, Looden T, Chauvin R, et al. Connectome-wide mega-analysis reveals robust patterns of atypical functional connectivity in autism. Biol Psychiatry. 2023;94:29–39. 10.1016/J.BIOPSYCH.2022.12.018 36925414
30.Soma D Hirosawa T Hasegawa C An KM Kameya M Hino S Atypical resting state functional neural network in children with autism spectrum disorder: graph theory approach Front Psychiatry 2021 12 790234 10.3389/FPSYT.2021.790234/BIBTEX 34970170 Soma D, Hirosawa T, Hasegawa C, An KM, Kameya M, Hino S, et al. Atypical resting state functional neural network in children with autism spectrum disorder: graph theory approach. Front Psychiatry. 2021;12:790234 10.3389/FPSYT.2021.790234/BIBTEX PMC8712628 34970170
31.Zhang J Zhang Z Sun H Ma Y Yang J Chen K Personalized functional network mapping for autism spectrum disorder and attention-deficit/hyperactivity disorder Translational Psychiatry 2024 14 1 11 10.1038/s41398-024-02797-z PMC10861462 38346949 Zhang J, Zhang Z, Sun H, Ma Y, Yang J, Chen K, et al. Personalized functional network mapping for autism spectrum disorder and attention-deficit/hyperactivity disorder. Translational Psychiatry. 2024;14:1–11. 10.1038/s41398-024-02797-z PMC10861462 38346949
32.Barttfeld P Amoruso L Ais J Cukier S Bavassi L Tomio A Organization of brain networks governed by long-range connections index autistic traits in the general population J Neurodev Disord 2013 5 1 9 10.1186/1866-1955-5-16/FIGURES/3 PMC3698083 23806204 Barttfeld P, Amoruso L, Ais J, Cukier S, Bavassi L, Tomio A, et al. Organization of brain networks governed by long-range connections index autistic traits in the general population. J Neurodev Disord. 2013;5:1–9. 10.1186/1866-1955-5-16/FIGURES/3 PMC3698083 23806204
33.Watanabe T Rees G Age-associated changes in rich-club organisation in autistic and neurotypical human brains Scientific Reports 2015 5 1 13 10.1038/srep16152 PMC4633620 26537477 Watanabe T, Rees G. Age-associated changes in rich-club organisation in autistic and neurotypical human brains. Scientific Reports. 2015;5:1–13. 10.1038/srep16152 PMC4633620 26537477
34.Bassett DS Bullmore ET Small-world brain networks revisited Neuroscientist 2017 23 499 516 10.1177/1073858416667720 27655008 PMC5603984 Bassett DS, Bullmore ET. Small-world brain networks revisited. Neuroscientist. 2017;23:499–516. 10.1177/1073858416667720 PMC5603984 27655008
35.White JG Southgate E Thomson JN Brenner S The structure of the nervous system of the nematode caenorhabditis elegans Philos Trans R Soc Lond B Biol Sci 1986 314 1 340 10.1098/RSTB.1986.0056 22462104 White JG, Southgate E, Thomson JN, Brenner S. The structure of the nervous system of the nematode caenorhabditis elegans. Philos Trans R Soc Lond B Biol Sci. 1986;314:1–340. 10.1098/RSTB.1986.0056 22462104
36.Muldoon SF Bridgeford EW Bassett DS Small-world propensity and weighted brain networks Scientific Reports 2016 6 1 13 10.1038/srep22057 PMC4766852 26912196 Muldoon SF, Bridgeford EW, Bassett DS. Small-world propensity and weighted brain networks. Scientific Reports. 2016;6:1–13. 10.1038/srep22057 PMC4766852 26912196
37.Medaglia JD Lynall ME Bassett DS Cognitive network neuroscience J Cogn Neurosci 2015 27 1471 91 10.1162/JOCN_A_00810 25803596 Medaglia JD, Lynall ME, Bassett DS. Cognitive network neuroscience. J Cogn Neurosci. 2015;27:1471–91. 10.1162/JOCN_A_00810 PMC4854276 25803596
38.Nomi JS Uddin LQ Developmental changes in large-scale network connectivity in autism Neuroimage Clin 2015 7 732 41 10.1016/j.nicl.2015.02.024 25844325 Nomi JS, Uddin LQ. Developmental changes in large-scale network connectivity in autism. Neuroimage Clin. 2015;7:732–41. 10.1016/j.nicl.2015.02.024 PMC4375789 25844325
39.Vasa RA, Mostofsky SH, Ewen JB The disrupted connectivity hypothesis of autism spectrum disorders: time for the next phase in research. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging. Elsevier Inc; 2016 [cited 2021 Jan 29]. p. 245–52. 10.1016/j.bpsc.2016.02.003 PMC5222574 28083565
40.Uddin LQ Supekar K Menon V Reconceptualizing functional brain connectivity in autism from a developmental perspective Frontiers in Human Neuroscience 2013 7 458 10.3389/fnhum.2013.00458 23966925 Uddin LQ, Supekar K, Menon V. Reconceptualizing functional brain connectivity in autism from a developmental perspective. Frontiers in Human Neuroscience. 2013;7:458 10.3389/fnhum.2013.00458 PMC3735986 23966925
41.Andrews DS Lee JK Harvey DJ Waizbard-Bartov E Solomon M Rogers SJ A longitudinal study of white matter development in relation to changes in autism severity across early childhood Biol Psychiatry 2021 89 424 32 10.1016/J.BIOPSYCH.2020.10.013 33349451 PMC7867569 Andrews DS, Lee JK, Harvey DJ, Waizbard-Bartov E, Solomon M, Rogers SJ, et al. A longitudinal study of white matter development in relation to changes in autism severity across early childhood. Biol Psychiatry. 2021;89:424–32. 10.1016/J.BIOPSYCH.2020.10.013 PMC7867569 33349451
42.Wolff JJ Gu H Gerig G Elison JT Styner M Gouttard S Differences in white matter fiber tract development present from 6–24 months in infants with autism Am J Psychiatry 2012 169 589 600 10.1176/APPI.AJP.2011.11091447 22362397 Wolff JJ, Gu H, Gerig G, Elison JT, Styner M, Gouttard S, et al. Differences in white matter fiber tract development present from 6–24 months in infants with autism. Am J Psychiatry. 2012;169:589–600. 10.1176/APPI.AJP.2011.11091447 PMC3377782 22362397
43.Hazlett HC Poe MD Gerig G Styner M Chappell C Smith RG Early brain overgrowth in autism associated with an increase in cortical surface area before age 2 Arch Gen Psychiatry 2011 68 467 10.1001/ARCHGENPSYCHIATRY.2011.39 21536976 Hazlett HC, Poe MD, Gerig G, Styner M, Chappell C, Smith RG, et al. Early brain overgrowth in autism associated with an increase in cortical surface area before age 2. Arch Gen Psychiatry. 2011;68:467 10.1001/ARCHGENPSYCHIATRY.2011.39 PMC3315057 21536976
44.Courchesne E Karns CM Davis HR Ziccardi R Carper RA Tigue ZD Unusual brain growth patterns in early life in patients with autistic disorder: an MRI study Neurology 2001 57 245 54 10.1212/WNL.57.2.245 11468308 Courchesne E, Karns CM, Davis HR, Ziccardi R, Carper RA, Tigue ZD, et al. Unusual brain growth patterns in early life in patients with autistic disorder: an MRI study. Neurology. 2001;57:245–54. 10.1212/WNL.57.2.245 11468308
45.Jones W Klin A Attention to eyes is present but in decline in 2–6-month-old infants later diagnosed with autism Nature 2013 504 427 31 10.1038/nature12715 24196715 Jones W, Klin A. Attention to eyes is present but in decline in 2–6-month-old infants later diagnosed with autism. Nature. 2013;504:427–31. 10.1038/nature12715 PMC4035120 24196715
46.Ewen J, Babiloni C, Collins G, Ethridge L, Gotman J, Ikeda A, et al. The GREENBEAN Checklist for Reporting Studies Evaluating the Effectiveness of EEG-Based Biomarkers. Clin Neurophysiol. 2025;2110777. 10.1016/j.clinph.2025.2110777 PMC12302412 40554380
47.FDA-NIH Biomarker Working Group. BEST (Biomarkers, EndpointS, and other Tools) Resource. BEST (Biomarkers, EndpointS, and other Tools) Resource. Silver Spring: Food and Drug Administration (US); 2016 [cited 2024 Jul 26]. https://www.ncbi.nlm.nih.gov/books/NBK326791/.27010052
48.Davis KD Aghaeepour N Ahn AH Angst MS Borsook D Brenton A Discovery and validation of biomarkers to aid the development of safe and effective pain therapeutics: challenges and opportunities Nature Reviews Neurology 2020 16 381 400 10.1038/s41582-020-0362-2 32541893 Davis KD, Aghaeepour N, Ahn AH, Angst MS, Borsook D, Brenton A, et al. Discovery and validation of biomarkers to aid the development of safe and effective pain therapeutics: challenges and opportunities. Nature Reviews Neurology. 2020;16:381–400. 10.1038/s41582-020-0362-2 PMC7326705 32541893
49.Tick B Bolton P Happé F Rutter M Rijsdijk F Heritability of autism spectrum disorders: a meta-analysis of twin studies J Child Psychol Psychiatry 2016 57 585 95 10.1111/JCPP.12499 26709141 PMC4996332 Tick B, Bolton P, Happé F, Rutter M, Rijsdijk F. Heritability of autism spectrum disorders: a meta-analysis of twin studies. J Child Psychol Psychiatry. 2016;57:585–95. 10.1111/JCPP.12499 PMC4996332 26709141
50.Genovese A Butler MG The autism spectrum: behavioral, psychiatric and genetic associations Genes (Basel) 2023 14 677 10.3390/GENES14030677 36980949 Genovese A, Butler MG. The autism spectrum: behavioral, psychiatric and genetic associations. Genes (Basel). 2023;14:677 10.3390/GENES14030677 PMC10048473 36980949
51.Smit DJA Stam CJ Posthuma D Boomsma DI De Geus EJC Heritability of “small‐world” networks in the brain: A graph theoretical analysis of resting‐state EEG functional connectivity Hum Brain Mapp 2008 29 1368 10.1002/HBM.20468 18064590 Smit DJA, Stam CJ, Posthuma D, Boomsma DI, De Geus EJC. Heritability of “small‐world” networks in the brain: A graph theoretical analysis of resting‐state EEG functional connectivity. Hum Brain Mapp. 2008;29:1368 10.1002/HBM.20468 PMC6870849 18064590
52.Sinclair B Hansell NK Blokland GAM Martin NG Thompson PM Breakspear M Heritability of the network architecture of intrinsic brain functional connectivity Neuroimage 2015 121 243 10.1016/J.NEUROIMAGE.2015.07.048 26226088 Sinclair B, Hansell NK, Blokland GAM, Martin NG, Thompson PM, Breakspear M, et al. Heritability of the network architecture of intrinsic brain functional connectivity. Neuroimage. 2015;121:243 10.1016/J.NEUROIMAGE.2015.07.048 PMC4837693 26226088
53.Reineberg AE Hatoum AS Hewitt JK Banich MT Friedman NP Genetic and environmental influence on the human functional connectome Cereb Cortex 2020 30 2099 113 10.1093/CERCOR/BHZ225 31711120 Reineberg AE, Hatoum AS, Hewitt JK, Banich MT, Friedman NP. Genetic and environmental influence on the human functional connectome. Cereb Cortex. 2020;30:2099–113. 10.1093/CERCOR/BHZ225 PMC7175002 31711120
54.Just MA Keller TA Malave VL Kana RK Varma S Autism as a neural systems disorder: a theory of frontal-posterior underconnectivity Neurosci Biobehav Rev 2012 36 1292 313 10.1016/J.NEUBIOREV.2012.02.007 22353426 Just MA, Keller TA, Malave VL, Kana RK, Varma S. Autism as a neural systems disorder: a theory of frontal-posterior underconnectivity. Neurosci Biobehav Rev. 2012;36:1292–313. 10.1016/J.NEUBIOREV.2012.02.007 PMC3341852 22353426
55.Mclane HC Berkowitz AL Patenaude BN Mckenzie ED Wolper E Wahlster S Availability, accessibility, and affordability of neurodiagnostic tests in 37 countries Neurology 2015 85 1614 10.1212/WNL.0000000000002090 26446063 Mclane HC, Berkowitz AL, Patenaude BN, Mckenzie ED, Wolper E, Wahlster S, et al. Availability, accessibility, and affordability of neurodiagnostic tests in 37 countries. Neurology. 2015;85:1614 10.1212/WNL.0000000000002090 PMC4642148 26446063
56.Singer W Neuronal synchrony: a versatile code for the definition of relations? Neuron 1999 24 49 65 10.1016/S0896-6273(00)80821-1 10677026 Singer W. Neuronal synchrony: a versatile code for the definition of relations? Neuron. 1999;24:49–65. 10.1016/S0896-6273(00)80821-1 10677026
57.Varela F Lachaux JP Rodriguez E Martinerie J The brainweb: phase synchronization and large-scale integration Nat Rev Neurosci 2001 2 229 39 10.1038/35067550 11283746 Varela F, Lachaux JP, Rodriguez E, Martinerie J. The brainweb: phase synchronization and large-scale integration. Nat Rev Neurosci. 2001;2:229–39. 10.1038/35067550 11283746
58.Rodriguez E George N Lachaux JP Martinerie J Renault B Varela FJ Perception’s shadow: long-distance synchronization of human brain activity Nature 1999 397 430 3 10.1038/17120 9989408 Rodriguez E, George N, Lachaux JP, Martinerie J, Renault B, Varela FJ. Perception’s shadow: long-distance synchronization of human brain activity. Nature. 1999;397:430–3. 10.1038/17120 9989408
59.Loth E Charman T Mason L Tillmann J Jones EJH Wooldridge C The EU-AIMS Longitudinal European Autism Project (LEAP): design and methodologies to identify and validate stratification biomarkers for autism spectrum disorders Mol Autism 2017 8 24 10.1186/S13229-017-0146-8 28649312 Loth E, Charman T, Mason L, Tillmann J, Jones EJH, Wooldridge C, et al. The EU-AIMS Longitudinal European Autism Project (LEAP): design and methodologies to identify and validate stratification biomarkers for autism spectrum disorders. Mol Autism. 2017;8:24 10.1186/S13229-017-0146-8 PMC5481887 28649312
60.Charman T Loth E Tillmann J Crawley D Wooldridge C Goyard D The EU-AIMS Longitudinal European Autism Project (LEAP): Clinical characterisation Mol Autism 2017 8 1 21 10.1186/S13229-017-0145-9/TABLES/8 PMC5481972 28649313 Charman T, Loth E, Tillmann J, Crawley D, Wooldridge C, Goyard D, et al. The EU-AIMS Longitudinal European Autism Project (LEAP): Clinical characterisation. Mol Autism. 2017;8:1–21. 10.1186/S13229-017-0145-9/TABLES/8 PMC5481972 28649313
61.Lai MC Kassee C Besney R Bonato S Hull L Mandy W Prevalence of co-occurring mental health diagnoses in the autism population: a systematic review and meta-analysis Lancet Psychiatry 2019 6 819 29 10.1016/S2215-0366(19)30289-5 31447415 Lai MC, Kassee C, Besney R, Bonato S, Hull L, Mandy W, et al. Prevalence of co-occurring mental health diagnoses in the autism population: a systematic review and meta-analysis. Lancet Psychiatry. 2019;6:819–29. 10.1016/S2215-0366(19)30289-5 31447415
62.Sawyer SM Azzopardi PS Wickremarathne D Patton GC The age of adolescence Lancet Child Adolesc Health 2018 2 223 8 10.1016/S2352-4642(18)30022-1 30169257 Sawyer SM, Azzopardi PS, Wickremarathne D, Patton GC. The age of adolescence. Lancet Child Adolesc Health. 2018;2:223–8. 10.1016/S2352-4642(18)30022-1 30169257
63.Nations U Youth | United Nations. [cited 2026 Jun 25]. https://www.un.org/en/global-issues/youth
64.Váša F Romero-Garcia R Kitzbichler MG Seidlitz J Whitaker KJ Vaghi MM Conservative and disruptive modes of adolescent change in human brain functional connectivity Proc Natl Acad Sci USA 2020 117 3248 53 10.1073/PNAS.1906144117/SUPPL_FILE/PNAS.1906144117.SAPP.PDF 31992644 Váša F, Romero-Garcia R, Kitzbichler MG, Seidlitz J, Whitaker KJ, Vaghi MM, et al. Conservative and disruptive modes of adolescent change in human brain functional connectivity. Proc Natl Acad Sci USA. 2020;117:3248–53. 10.1073/PNAS.1906144117/SUPPL_FILE/PNAS.1906144117.SAPP.PDF PMC7022153 31992644
65.Klapwijk ET Goddings AL Burnett Heyes S Bird G Viner RM Blakemore SJ Increased functional connectivity with puberty in the mentalising network involved in social emotion processing Horm Behav 2013 64 314 10.1016/J.YHBEH.2013.03.012 23998674 Klapwijk ET, Goddings AL, Burnett Heyes S, Bird G, Viner RM, Blakemore SJ. Increased functional connectivity with puberty in the mentalising network involved in social emotion processing. Horm Behav. 2013;64:314 10.1016/J.YHBEH.2013.03.012 PMC4540076 23998674
66.Sparrow SS, Balla DA, Cicchetti DV Vineland II: Vineland adaptive behavior scales. American Guidance Service Publishing; 2005.
67.Bruni TP Test Review: Social Responsiveness Scale–Second Edition (SRS-2). 2014;32:365–9. 10.1177/0734282913517525
68.Taylor SC Gehringer BN Dow HC Langer A Rawot E Smernoff Z Contrasting views of autism spectrum traits in adults, especially in self-reports vs. informant-reports for women high in autism spectrum traits Journal of Autism and Developmental Disorders 2022 54 1088 100 10.1007/S10803-022-05822-6 36484966 Taylor SC, Gehringer BN, Dow HC, Langer A, Rawot E, Smernoff Z, et al. Contrasting views of autism spectrum traits in adults, especially in self-reports vs. informant-reports for women high in autism spectrum traits. Journal of Autism and Developmental Disorders. 2022;54:1088–100. 10.1007/S10803-022-05822-6 PMC9734875 36484966
69.Wechsler D APA PsycTests. 2011 [cited 2024 Jul 31]. Wechsler Abbreviated Scale of Intelligence--Second Edition. https://psycnet.apa.org/doiLanding?doi=10.1037%2Ft15171-00010.1177/10731911231207785 37902058
70.The Wechsler Intelligence Scale for Children—Fourth Edition. [cited 2024 Jul 31]. https://psycnet.apa.org/record/2010-04803-007
71.Woolger C Wechsler Intelligence Scale for Children-Third Edition (wisc-iii). Understanding Psychological Assessment. 2001;219–33. 10.1007/978-1-4615-1185-4_11
72.Ryan JJ, Lopez SJ Wechsler adult intelligence scale-III. Understanding Psychological Assessment. 2001;19–42. 10.1007/978-1-4615-1185-4_2
73.Wechsler D Wechsler Adult Intelligence Scale-Fourth Edition. 2008. 10.1037/T15169-000
74.Anzolin A Presti P Van De Steen F Astolfi L Haufe S Marinazzo D Quantifying the effect of demixing approaches on directed connectivity estimated between reconstructed EEG sources Brain Topogr 2019 32 655 74 10.1007/S10548-019-00705-Z 30972604 Anzolin A, Presti P, Van De Steen F, Astolfi L, Haufe S, Marinazzo D. Quantifying the effect of demixing approaches on directed connectivity estimated between reconstructed EEG sources. Brain Topogr. 2019;32:655–74. 10.1007/S10548-019-00705-Z 30972604
75.Garcés P Baumeister S Mason L Chatham CH Holiga S Dukart J Resting state EEG power spectrum and functional connectivity in autism: a cross-sectional analysis Molecular Autism 2022 13 1 16 10.1186/S13229-022-00500-X PMC9118870 35585637 Garcés P, Baumeister S, Mason L, Chatham CH, Holiga S, Dukart J, et al. Resting state EEG power spectrum and functional connectivity in autism: a cross-sectional analysis. Molecular Autism. 2022;13:1–16. 10.1186/S13229-022-00500-X PMC9118870 35585637
76.Fraschini M Demuru M Crobe A Marrosu F Stam CJ Hillebrand A The effect of epoch length on estimated EEG functional connectivity and brain network organisation J Neural Eng 2016 13 036015 10.1088/1741-2560/13/3/036015 27137952 Fraschini M, Demuru M, Crobe A, Marrosu F, Stam CJ, Hillebrand A. The effect of epoch length on estimated EEG functional connectivity and brain network organisation. J Neural Eng. 2016;13:036015 10.1088/1741-2560/13/3/036015 27137952
77.Pullon RM Yan L Sleigh JW Warnaby CE Granger causality of the electroencephalogram reveals abrupt global loss of cortical information flow during propofol-induced loss of responsiveness Anesthesiology 2020 133 774 10.1097/ALN.0000000000003398 32930729 Pullon RM, Yan L, Sleigh JW, Warnaby CE. Granger causality of the electroencephalogram reveals abrupt global loss of cortical information flow during propofol-induced loss of responsiveness. Anesthesiology. 2020;133:774 10.1097/ALN.0000000000003398 PMC7495984 32930729
78.Lachaux JP Rodriguez E Martinerie J Varela FJ Measuring phase synchrony in brain signals Hum Brain Mapp 1999 8 194 10.1002/(SICI)1097-0193(1999)8:4<194::AID-HBM4>3.0.CO;2-C 10619414 Lachaux JP, Rodriguez E, Martinerie J, Varela FJ. Measuring phase synchrony in brain signals. Hum Brain Mapp. 1999;8:194.10.1002/(SICI)1097-0193(1999)8:4<194::AID-HBM4>3.0.CO;2-C PMC6873296 10619414
79.Mehra C Beyh A Laiou P Garces P Jones EJ Mason L Zero-phase-delay synchrony between interacting neural populations: implications for functional connectivity derived biomarkers Imaging Neuroscience 2025 10.1162/IMAG.A.985 41230416 Mehra C, Beyh A, Laiou P, Garces P, Jones EJ, Mason L, et al. Zero-phase-delay synchrony between interacting neural populations: implications for functional connectivity derived biomarkers. Imaging Neuroscience. 2025. 10.1162/IMAG.A.985 PMC12603661 41230416
80.Desikan RS Ségonne F Fischl B Quinn BT Dickerson BC Blacker D An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest Neuroimage 2006 31 968 80 10.1016/J.NEUROIMAGE.2006.01.021 16530430 Desikan RS, Ségonne F, Fischl B, Quinn BT, Dickerson BC, Blacker D, et al. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. Neuroimage. 2006;31:968–80. 10.1016/J.NEUROIMAGE.2006.01.021 16530430
81.Rubinov M Kötter R Hagmann P Sporns O Brain connectivity toolbox: a collection of complex network measurements and brain connectivity datasets Neuroimage 2009 47 S169 10.1016/S1053-8119(09)71822-1 Rubinov M, Kötter R, Hagmann P, Sporns O. Brain connectivity toolbox: a collection of complex network measurements and brain connectivity datasets. Neuroimage. 2009;47:S169 10.1016/S1053-8119(09)71822-1
82.Humphries MD Gurney K Network ‘Small-World-Ness’: a quantitative method for determining canonical network equivalence PLoS One 2008 3 e0002051 10.1371/JOURNAL.PONE.0002051 18446219 Humphries MD, Gurney K. Network ‘Small-World-Ness’: a quantitative method for determining canonical network equivalence. PLoS One. 2008;3:e0002051 10.1371/JOURNAL.PONE.0002051 PMC2323569 18446219
83.Bolaños M Bernat EM He B Aviyente S A weighted small world network measure for assessing functional connectivity J Neurosci Methods 2013 212 133 42 10.1016/J.JNEUMETH.2012.10.004 23085279 Bolaños M, Bernat EM, He B, Aviyente S. A weighted small world network measure for assessing functional connectivity. J Neurosci Methods. 2013;212:133–42. 10.1016/J.JNEUMETH.2012.10.004 23085279
84.Noble S Scheinost D Constable RT A decade of test-retest reliability of functional connectivity: a systematic review and meta-analysis Neuroimage 2019 203 116157 10.1016/J.NEUROIMAGE.2019.116157 31494250 Noble S, Scheinost D, Constable RT. A decade of test-retest reliability of functional connectivity: a systematic review and meta-analysis. Neuroimage. 2019;203:116157 10.1016/J.NEUROIMAGE.2019.116157 PMC6907736 31494250
85.Koo TK Li MY A guideline of selecting and reporting intraclass correlation coefficients for reliability research J Chiropr Med 2016 15 155 63 10.1016/J.JCM.2016.02.012 27330520 Koo TK, Li MY. A guideline of selecting and reporting intraclass correlation coefficients for reliability research. J Chiropr Med. 2016;15:155–63. 10.1016/J.JCM.2016.02.012 PMC4913118 27330520
86.Grove J Ripke S Als TD Mattheisen M Walters RK Won H Identification of common genetic risk variants for autism spectrum disorder Nat Genet 2019 51 431 444 10.1038/s41588-019-0344-8 30804558 Grove J, Ripke S, Als TD, Mattheisen M, Walters RK, Won H, et al. Identification of common genetic risk variants for autism spectrum disorder. Nat Genet. 2019;51:431–444. 10.1038/s41588-019-0344-8 PMC6454898 30804558
87.Biton A Traut N Poline JB Aribisala BS Bastin ME Bülow R Polygenic architecture of human neuroanatomical diversity Cerebral Cortex 2020 30 2307 20 10.1093/CERCOR/BHZ241 32109272 Biton A, Traut N, Poline JB, Aribisala BS, Bastin ME, Bülow R, et al. Polygenic architecture of human neuroanatomical diversity. Cerebral Cortex. 2020;30:2307–20. 10.1093/CERCOR/BHZ241 PMC7175006 32109272
88.Beck CW Bliwise NG Interactions are critical CBE Life Sci Educ 2014 13 371 2 10.1187/CBE.14-05-0086 25185220 Beck CW, Bliwise NG. Interactions are critical. CBE Life Sci Educ. 2014;13:371–2. 10.1187/CBE.14-05-0086 PMC4152198 25185220
89.Shen X Finn ES Scheinost D Rosenberg MD Chun MM Papademetris X Using connectome-based predictive modeling to predict individual behavior from brain connectivity Nature Protocols 2017 12 506 18 10.1038/nprot.2016.178 28182017 Shen X, Finn ES, Scheinost D, Rosenberg MD, Chun MM, Papademetris X, et al. Using connectome-based predictive modeling to predict individual behavior from brain connectivity. Nature Protocols. 2017;12:506–18. 10.1038/nprot.2016.178 PMC5526681 28182017
90.Chatham CH Taylor KI Charman T Liogier D’ardhuy X Eule E Fedele A Adaptive Behavior in Autism: Minimal Clinically-Important Differences on the Vineland-II Autism Res 2018 11 270 10.1002/AUR.1874 28941213 PMC5997920 Chatham CH, Taylor KI, Charman T, Liogier D’ardhuy X, Eule E, Fedele A, et al. Adaptive Behavior in Autism: Minimal Clinically-Important Differences on the Vineland-II. Autism Res. 2018;11:270 10.1002/AUR.1874 PMC5997920 28941213
91.Clinch S Hudgens S Gibbons E Willgoss T Smith J Polek E Quantitative and qualitative exploration of meaningful change on the vineland adaptive behavior scales (Vineland™-II) in children and adolescents with autism without intellectual disability following participation in a clinical trial Patient Relat Outcome Meas 2023 14 337 54 10.2147/PROM.S385542 38027418 Clinch S, Hudgens S, Gibbons E, Willgoss T, Smith J, Polek E, et al. Quantitative and qualitative exploration of meaningful change on the vineland adaptive behavior scales (Vineland™-II) in children and adolescents with autism without intellectual disability following participation in a clinical trial. Patient Relat Outcome Meas. 2023;14:337–54. 10.2147/PROM.S385542 PMC10675867 38027418
92.Jacobson NS Truax P Clinical significance: a statistical approach to defining meaningful change in psychotherapy research J Consult Clin Psychol 1991 59 12 9 10.1037/0022-006X.59.1.12 2002127 Jacobson NS, Truax P. Clinical significance: a statistical approach to defining meaningful change in psychotherapy research. J Consult Clin Psychol. 1991;59:12–9. 10.1037/0022-006X.59.1.12 2002127
93.Maassen GH The standard error in the jacobson and truax reliable change index: the classical approach to the assessment of reliable change Journal of the International Neuropsychological Society 2004 10 888 93 10.1017/S1355617704106097 15637779 Maassen GH. The standard error in the jacobson and truax reliable change index: the classical approach to the assessment of reliable change. Journal of the International Neuropsychological Society. 2004;10:888–93. 10.1017/S1355617704106097 15637779
94.O’Muircheartaigh J Dean DC Dirks H Waskiewicz N Lehman K Jerskey BA Interactions between white matter asymmetry and language during neurodevelopment Journal of Neuroscience 2013 33 16170 7 10.1523/JNEUROSCI.1463-13.2013 24107949 O’Muircheartaigh J, Dean DC, Dirks H, Waskiewicz N, Lehman K, Jerskey BA, et al. Interactions between white matter asymmetry and language during neurodevelopment. Journal of Neuroscience. 2013;33:16170–7. 10.1523/JNEUROSCI.1463-13.2013 PMC3792458 24107949
95.Courchesne E Carper R Akshoomoff N Evidence of brain overgrowth in the first year of life in autism JAMA 2003 290 337 44 10.1001/JAMA.290.3.337 12865374 Courchesne E, Carper R, Akshoomoff N. Evidence of brain overgrowth in the first year of life in autism. JAMA. 2003;290:337–44. 10.1001/JAMA.290.3.337 12865374
96.Mandrekar JN Receiver operating characteristic curve in diagnostic test assessment Journal of Thoracic Oncology 2010 5 1315 6 10.1097/JTO.0B013E3181EC173D 20736804 Mandrekar JN. Receiver operating characteristic curve in diagnostic test assessment. Journal of Thoracic Oncology. 2010;5:1315–6. 10.1097/JTO.0B013E3181EC173D 20736804
97.T-STEP - TEACCH School Transition to Employment and Postsecondary Education | TEACCH® Autism Program. [cited 2025 May 15]. https://teacch.com/adult_services/tstep/.
98.Frye RE Slattery J Delhey L Furgerson B Strickland T Tippett M Folinic acid improves verbal communication in children with autism and language impairment: a randomized double-blind placebo-controlled trial Molecular Psychiatry 2018 23 247 56 10.1038/mp.2016.168 27752075 PMC5794882 Frye RE, Slattery J, Delhey L, Furgerson B, Strickland T, Tippett M, et al. Folinic acid improves verbal communication in children with autism and language impairment: a randomized double-blind placebo-controlled trial. Molecular Psychiatry. 2018;23:247–56. 10.1038/mp.2016.168 PMC5794882 27752075
99.Gracia-Tabuenca Z Moreno MB Barrios FA Alcauter S Development of the brain functional connectome follows puberty-dependent nonlinear trajectories Neuroimage 2021 229 117769 10.1016/J.NEUROIMAGE.2021.117769 33482398 Gracia-Tabuenca Z, Moreno MB, Barrios FA, Alcauter S. Development of the brain functional connectome follows puberty-dependent nonlinear trajectories. Neuroimage. 2021;229:117769 10.1016/J.NEUROIMAGE.2021.117769 33482398
100.Kelly AMC Di Martino A Uddin LQ Shehzad Z Gee DG Reiss PT Development of anterior cingulate functional connectivity from late childhood to early adulthood Cerebral Cortex 2009 19 640 57 10.1093/CERCOR/BHN117 18653667 Kelly AMC, Di Martino A, Uddin LQ, Shehzad Z, Gee DG, Reiss PT, et al. Development of anterior cingulate functional connectivity from late childhood to early adulthood. Cerebral Cortex. 2009;19:640–57. 10.1093/CERCOR/BHN117 18653667
101.Netoff TI Clewley R Arno S Keck T White JA Epilepsy in small-world networks Journal of Neuroscience 2004 24 8075 83 10.1523/JNEUROSCI.1509-04.2004 15371508 Netoff TI, Clewley R, Arno S, Keck T, White JA. Epilepsy in small-world networks. Journal of Neuroscience. 2004;24:8075–83. 10.1523/JNEUROSCI.1509-04.2004 PMC6729784 15371508
102.Yuasa K Hirosawa T Soma D Furutani N Kameya M Sano M Eyes-state-dependent alterations of magnetoencephalographic connectivity associated with delayed recall in alzheimer’s disease via graph theory approach Front Psychiatry 2023 14 1272120 10.3389/FPSYT.2023.1272120/FULL 37941968 Yuasa K, Hirosawa T, Soma D, Furutani N, Kameya M, Sano M, et al. Eyes-state-dependent alterations of magnetoencephalographic connectivity associated with delayed recall in alzheimer’s disease via graph theory approach. Front Psychiatry. 2023;14:1272120 10.3389/FPSYT.2023.1272120/FULL PMC10628524 37941968
103.Lewis CM Vassos E Polygenic risk scores: from research tools to clinical instruments Genome Med 2020 12 1 11 10.1186/S13073-020-00742-5/TABLES/2 PMC7236300 32423490 Lewis CM, Vassos E. Polygenic risk scores: from research tools to clinical instruments. Genome Med. 2020;12:1–11. 10.1186/S13073-020-00742-5/TABLES/2 PMC7236300 32423490
104.Barber AD Hegarty CE Lindquist M Karlsgodt KH Heritability of functional connectivity in resting state: assessment of the dynamic mean, dynamic variance, and static connectivity across networks Cereb Cortex 2021 31 2834 44 10.1093/CERCOR/BHAA391 33429433 Barber AD, Hegarty CE, Lindquist M, Karlsgodt KH. Heritability of functional connectivity in resting state: assessment of the dynamic mean, dynamic variance, and static connectivity across networks. Cereb Cortex. 2021;31:2834–44. 10.1093/CERCOR/BHAA391 PMC8325018 33429433
105.Sparrow S, Balla D, Cicchetti D Vineland Adpative Behavior Scales: Interview edition survey form manual. Circle Press, MN. 1984 [cited 2024 Jul 31].
106.Perry A Flanagan HE Dunn Geier J Freeman NL Brief report: the Vineland adaptive behavior scales in young children with autism spectrum disorders at different cognitive levels J Autism Dev Disord 2009 39 1066 78 10.1007/S10803-009-0704-9 19234777 Perry A, Flanagan HE, Dunn Geier J, Freeman NL. Brief report: the Vineland adaptive behavior scales in young children with autism spectrum disorders at different cognitive levels. J Autism Dev Disord. 2009;39:1066–78. 10.1007/S10803-009-0704-9 19234777
107.Bañuelos C, Verstynen T How structural brain network topologies associate with cognitive abilities in a value-based decision-making task. bioRxiv [Preprint]. 2019. Available from: https://www.biorxiv.org/content/10.1101/743245v1.full
108.Richardson MP Large scale brain models of epilepsy: dynamics meets connectomics. journal of neurology, neurosurgery and psychiatry BMJ Publishing Group 2012 83 1238 48 10.1136/jnnp-2011-301944 22917671 Richardson MP. Large scale brain models of epilepsy: dynamics meets connectomics. journal of neurology, neurosurgery and psychiatry. BMJ Publishing Group. 2012;83:1238–48. 10.1136/jnnp-2011-301944 22917671
109.Bassett DS Nelson BG Mueller BA Camchong J Lim KO Altered resting state complexity in schizophrenia Neuroimage 2012 59 2196 207 10.1016/J.NEUROIMAGE.2011.10.002 22008374 PMC3254701 Bassett DS, Nelson BG, Mueller BA, Camchong J, Lim KO. Altered resting state complexity in schizophrenia. Neuroimage. 2012;59:2196–207. 10.1016/J.NEUROIMAGE.2011.10.002 PMC3254701 22008374
110.Schneidman E Berry MJ Segev R Bialek W Weak pairwise correlations imply strongly correlated network states in a neural population Nature 2006 440 1007 12 10.1038/nature04701 16625187 Schneidman E, Berry MJ, Segev R, Bialek W. Weak pairwise correlations imply strongly correlated network states in a neural population. Nature. 2006;440:1007–12. 10.1038/nature04701 PMC1785327 16625187
111.Santarnecchi E Galli G Polizzotto NR Rossi A Rossi S Efficiency of weak brain connections support general cognitive functioning Hum Brain Mapp 2014 35 4566 82 10.1002/HBM.22495 24585433 Santarnecchi E, Galli G, Polizzotto NR, Rossi A, Rossi S. Efficiency of weak brain connections support general cognitive functioning. Hum Brain Mapp. 2014;35:4566–82. 10.1002/HBM.22495 PMC6869093 24585433
112.de Jonge E Garcés P de Bildt A Groen Y Jones EJH Mason L Atypical Resting-State EEG graph metrics of network efficiency across development in autism and their association with social cognition: results from the LEAP Study J Autism Dev Disord 2025 9 1 17 10.1007/S10803-025-06731-0/FIGURES/3 PMC13346277 39951212 de Jonge E, Garcés P, de Bildt A, Groen Y, Jones EJH, Mason L, et al. Atypical Resting-State EEG graph metrics of network efficiency across development in autism and their association with social cognition: results from the LEAP Study. J Autism Dev Disord. 2025;9:1–17. 10.1007/S10803-025-06731-0/FIGURES/3 PMC13346277 39951212
113.Micai M Fatta LM Gila L Caruso A Salvitti T Fulceri F Prevalence of co-occurring conditions in children and adults with autism spectrum disorder: a systematic review and meta-analysis Neurosci Biobehav Rev 2023 155 105436 10.1016/J.NEUBIOREV.2023.105436 37913872 Micai M, Fatta LM, Gila L, Caruso A, Salvitti T, Fulceri F, et al. Prevalence of co-occurring conditions in children and adults with autism spectrum disorder: a systematic review and meta-analysis. Neurosci Biobehav Rev. 2023;155:105436 10.1016/J.NEUBIOREV.2023.105436 37913872
114.Isaksson J Eklund F Remnélius KL Black MH Bölte S Neurodevelopmental conditions and adaptive functioning – a co-twin control study J Child Psychol Psychiatry 2025 67 777 87 10.1111/JCPP.70073 41188694 Isaksson J, Eklund F, Remnélius KL, Black MH, Bölte S. Neurodevelopmental conditions and adaptive functioning – a co-twin control study. J Child Psychol Psychiatry. 2025;67:777–87. 10.1111/JCPP.70073 PMC13102047 41188694
Come citare questo studio
Mehra C, Laiou P, Garcés P, Ewen JB, Loth E, Johnson MH, Mason L, Jones EJ, Charman T, Bourgeron T, Buitelaar J, Absoud M, Richardson MP, Murphy D, O'Muircheartaigh J. (2026). EEG functional connectivity as a prognostic biomarker of adaptive function in autistic people.. Translational psychiatry. DOI: 10.1038/s41398-026-04422-7
Consulta la pubblicazione originale ↗