Brain controllability mediates the effects of early life adversity on adolescent behavior and cognition moderated by genetic risks

Brain controllability mediates the effects of early life adversity on adolescent behavior and cognition moderated by genetic risks | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Brain controllability mediates the effects of early life adversity on adolescent behavior and cognition moderated by genetic risks Huaigui Liu, Yurong Jiang, Huili Sun, Matthew Rosenblatt, Jean Ye, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8777490/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 9 You are reading this latest preprint version Abstract Early life adversity (ELA) is a robust transdiagnostic risk factor for mental health disorders, yet the neurobiological mechanisms mediating its long-term impact remain poorly understood. Network control theory offers a novel framework for capturing how structural brain networks constrain and support brain dynamics. Controllability increases over development, associates with executive function and mental health, and appears sensitive to environmental insults. Thus, it may reflect a neurobiological mediator between ELA and behavioral outcomes. We tested whether alterations in modal controllability mediate the impact of multidimensional ELA on cognitive and behavioral outcomes in youth, and whether these pathways are shaped by genetic risk for neurodevelopmental conditions. Using data from 7,970 children aged 9–11 years in the Adolescent Brain Cognitive Development (ABCD) Study, we derived five latent ELA dimensions from 67 indicators, and computed polygenic risk scores (PRS) for attention-deficit/hyperactivity disorder (ADHD PRS ) and autism spectrum disorder (ASD PRS ). Distinct ELA dimensions were associated with increased controllability in medial frontal, frontoparietal, default mode, and motor networks, as well as with externalizing symptoms and impaired crystallized cognition. Controllability partially mediated these associations, and indirect effects were significantly moderated ADHD PRS and ASD PRS . Longitudinal analyses further demonstrated that baseline controllability predicted cognitive performance two years later. These findings delineate a neurodevelopmental cascade linking early adversity and genetic vulnerability to transdiagnostic mental health risk, positioning brain controllability as a promising mechanistic marker and potential target for early intervention. Biological sciences/Neuroscience Health sciences/Biomarkers/Predictive markers Health sciences/Diseases/Psychiatric disorders/ADHD Health sciences/Diseases/Psychiatric disorders/Autism spectrum disorders Biological sciences/Psychology Figures Figure 1 Figure 2 Figure 3 Figure 4 Introduction Early life adversity (ELA) is a potent and pervasive determinant of emotional, cognitive, and behavioral development. Mounting evidence links it with elevated risk for psychiatric disorders across the lifespan [ 1 – 6 ]. However, as adversity includes a wide-range of factors (e.g., caregiver psychopathology, emotional maltreatment, family dysfunction, socioeconomic disadvantage, neighborhood crime and trauma [ 7 ]), it is not a unitary construct. Furthermore, these distinct dimensions of ELA may exert differential effects on developmental trajectories [ 8 , 9 ]. For instance, deprivation is linked to executive function deficits and significantly higher rates of ADHD [ 10 , 11 ], unpredictability is shown to correlate with more conduct problems [ 12 ], while physical abuse is associated with increased lifetime substance use [ 13 ]. These distinctions motivate multidimensional models of ELA to better capture its heterogeneity and co-occurrence. These models offer a more precise framework for studying developmental consequences of ELA. The brain is particularly vulnerable to the effects of ELA [ 14 ]. Structural and functional neuroimaging studies have consistently demonstrated that ELA disrupts neurodevelopmental trajectories in domain-specific and region-specific ways [ 15 , 16 ]. For example, deprivation-related adversity are associated with reduced total brain volume [ 17 – 19 ] and premature adult-like connectivity between the amygdala and the medial prefrontal cortex [ 20 ]. While maltreatment exhibit cortical thinning in prefrontal and temporal regions [ 21 – 23 ] and heightened amygdala connectivity with the salience, default mode, and prefrontal networks [ 24 ]. White matter abnormalities have also been reported across distinct types of adversity: frontostriatal alterations following early institutionalization, neglect and deprivation [ 25 – 27 ], frontolimbic disruptions linked to socioeconomic disadvantage and physical neglect [ 28 , 29 ], and reduced integrity of association fibers in individuals exposed to domestic violence [ 30 ]. Yet, despite this growing literature, a fundamental question remains: how do adversity-related alterations in brain networks translate into increased vulnerability for psychopathology? One promising avenue for addressing this question lies in network control theory, a computational framework grounded in systems neuroscience [ 31 ]. Rather than describing regional structure or function in isolation, network controllability quantifies the brain’s capacity to flexibly transition between diverse mental states, based on the underlying structural connections [ 31 ]. Thus, network control theory offers a biologically grounded and mathematically tractable marker of structural connections constrain and support brain dynamics. Controllability increases over young childhood through adolescence, putatively allowing for the emergence of adult-like executive functions [ 32 , 33 ]. Given that executive functions are both developmentally sensitive and highly vulnerable to adversity [ 34 , 35 ], controllability may serve as a critical link between structural network changes and behavioral outcomes. Further, controllability is altered in several psychiatric disorders [ 36 – 38 ]. However, its role in adversity-psychopathology pathways remains largely unexplored. Importantly, not all children exposed to ELA develop psychiatric disorders. Genetics may explain this variability [ 39 , 40 ]. Previous studies estimate that up to 40–80% of the variance in psychiatric disorders is heritable [ 41 ], and polygenic risk scores (PRS) derived from genome-wide association studies (GWAS) now offer a scalable approach to quantify individual-level genetic liability [ 42 ]. Prior studies have linked PRSs of psychiatric disorders to brain structure [ 43 ], but few have examined their interaction with ELA in mediating neurodevelopmental pathways. Understanding these interactions is essential, especially in youth, when both environmental exposures and brain networks remain plastic. In the present study, we leveraged data from the ABCD study to investigate how distinct dimensions of ELA affect brain network controllability, and how these changes may mediate cognitive and behavioral outcomes. We also tested whether genetic risk for ADHD and ASD moderates these brain behavior pathways, and whether the observed effects persist over a two-year longitudinal follow-up. We selected risk for ADHD and ASD as they are early-onset neurodevelopmental conditions with high heritability and relevance to the age range of the ABCD cohort. Additionally, both are associated with altered structural connections [ 44 – 47 ]. To this end, we applied dimensional modeling of ELA across 67 baseline indicators, computed controllability metrics across ten conical brain networks, and assessed behavioral and cognitive outcomes using the Child Behavior Checklist (CBCL) and NIH Toolbox at both baseline and follow-up. Analyses included linear mixed-effects models, mediation and moderated mediation frameworks, and cross-lagged panel model (CLPM) cross-lagged panel modeling to examine temporal dynamics. By integrating environmental, neural, behavioral, and genetic data in a large, diverse pediatric sample, this study aims to elucidate the mechanisms through which early adversity becomes biologically embedded, and to identify potential targets for early identification and intervention in psychiatric risk. Methods Participants All participants were recruited by the ABCD study from the ABCD consortium ( https://abcdstudy.org/index.html ) , which includes over 11,000 children aged 9–11 years, recruited from 22 sites across the United States with a diverse range of geographic, socioeconomic, ethnic, and health backgrounds [ 48 ]. We used the Annual Curated Data Release 5.1. The study protocols were approved by the Institutional Review Boards (IRBs) at each participating site, with written informed consent obtained from youth and caregivers. Details about the participant cohort, data collection procedures, and preprocessing protocols are provided at the ABCD website ( https://abcdstudy.org/scientists/protocols/ ) and also are described elsewhere [ 48 , 49 ]. Early Life Adversity Item Selection and Data Cleaning ELA was assessed at baseline using 139 variables selected from the ABCD baseline measures [ 7 , 50 ], which captured diverse adversity domains, including caregiving disruption, caregiver psychopathology, maltreatment, neighborhood violence, family and community support, socioeconomic disadvantage, and exposure to physical trauma ( Supplementary Table 1 ). All items were drawn from validated scales administered to the child, parent, or rated by trained research staff and were coded such that higher scores indicated greater adversity. We excluded 31 variables with > 50% missing and removed an additional 29 variables with zero or near-zero variance using the nearZeroVar function from the caret R package [ 51 ]. Two redundant variables were manually removed. Furthermore, eight participants lacking age information were excluded. To avoid multicollinearity arising from familial relatedness, we excluded twins and triplets, retaining at most one participant per family among non-twin siblings. This exclusion was applied prior to all analyses, resulting in a final analytic sample of N = 2,054 excluded participants. After these data cleaning procedures, the remaining variables underwent outlier detection, with outliers defined as values exceeding 4 × median absolute deviation (MAD) of continuous environmental variables [ 52 ]. Finally, 10 ELAs variables with a missing rate > 50% were excluded and a final set of 67 variables was included. Details are provided in Supplementary Fig. 1, 2 and Supplementary Table 1 . Missing values were handled using the mice (v3.16.0) [ 53 ] R package. Ten imputed datasets were generated using predictive mean matching (continuous variables) and logistic regression (categorical variables), with 10 iterations each. Factor Analysis and Validation To reduce dimensionality and identify core latent adversity factors, we conducted an exploratory factor analysis (EFA) using principal axis factoring and oblimin rotation. ELA variables were averaged across the 10 imputed datasets prior to factor analysis. Suitability for factor analysis was confirmed by Kaiser-Meyer-Olkin (KMO) values > 0.60 and significant Bartlett’s test of sphericity (p < .001). The number of factors was determined by a combination of Very Simple Structure (VSS) and Minimum Average Partial (MAP) criteria, supplemented by model fit indices, including comparative fit index (CFI), root mean square error of approximation (RMSEA) and standardized root mean square residual (SRMR) across 1–9 factor solutions. Indicators with loadings > 0.45 were retained as primary contributors to each factor. Final factor structures were validated through confirmatory factor analysis (CFA) using the lavaan R package (version 0.6–19). To reduce overfitting, we applied tenfold cross-validation: in each fold, 90% of participants were used to estimate factor loadings, while factor scores were computed in the remaining 10%. This was iterated until all participants had out-of-sample estimates. The stability of factor scores was confirmed using Spearman correlations between cross-validated and full-sample scores. Behavioral Measures Child Behavior Checklist (CBCL) Children’s emotional and behavioral symptoms were assessed using the Child Behavior Checklist (CBCL), a validated parent- or caregiver-report scale [ 54 ]. The CBCL includes eight syndrome scales: anxious/depressed, withdrawn/depressed, somatic complaints, social problems, thought problems, attention problems, rule-breaking behavior, and aggressive behavior. We used standardized T-scores of three composite subscales: internalizing problems (anxious/depressed, withdrawn/depressed, and somatic complaints), externalizing problems (rule-breaking and aggressive behaviors), and total problems (all eight syndrome scales combined), with higher scores indicating greater symptom severity. NIH Toolbox Cognition Battery Cognitive function was evaluated using the NIH Toolbox Cognition Battery [ 55 ], which consists of seven performance-based tasks designed to measure key cognitive domains. Crystallized intelligence was assessed by summing performance on the Picture Vocabulary Test and the Oral Reading Recognition Test. Fluid intelligence was calculated using five tasks: the Flanker Inhibitory Control and Attention Test, the List Sorting Working Memory Test, the Dimensional Change Card Sort Test, the Pattern Comparison Processing Speed Test, and the Picture Sequence Memory Test. A total cognitive composite score was derived by summing the fluid and crystallized scores. For all cognition domains, standardized T-scores were used in the analyses, with higher values reflecting better cognitive performance. Crystallized intelligence data were obtained at both the baseline and 2-year follow-up. Neuroimaging Data Acquisition and Processing Diffusion MRI (dMRI) data were acquired using standardized protocols across multiple 3T scanner platforms at ABCD sites [ 48 ]. Individual-level diffusion MRI data were downloaded in FIB format from the ABCD Study Data Portal ( https://brain.labsolver.org ), where the preprocessing pipeline is detailed (N = 9,713 for baseline, N = 7,197 for 2-year follow-up). Raw dMRI volumes were acquired using a multishell diffusion scheme with b-values of 500, 1000, 2000, and 3000 s/mm² and corresponding diffusion directions of 6, 15, 15, and 60, respectively. Images were acquired at an in-plane resolution of 1.7 mm and slice thickness of 1.7 mm, and were rotated to align with the anterior commissure–posterior commissure (AC–PC) line, resulting in isotropic voxel resolution of 1.7 mm. Diffusion data were reconstructed using generalized q-sampling imaging with a sampling length ratio of 1.25 [ 33 , 56 , 57 ]. Whole-brain tractography was performed in DSI Studio using deterministic streamline tracking with a modified FACT algorithm [ 57 ]. Tractography parameters included an angular cutoff of 60 degrees, step size of 1.0 mm, minimum streamline length of 30 mm, maximum length of 300 mm, and spin density function smoothing set to 0. QA values were used as the termination criterion, and fiber tracking continued until 10 million streamlines were reconstructed for each subject. QA was computed in native space and used for nonlinear registration to MNI space via the SPM12 algorithm. Within MNI space, spin density functions were reconstructed using three fiber orientations per voxel, and a mean diffusion distance of 1.25 mm. For each participant, a 268 × 268 structural connectivity matrix was generated by averaging the QA values of streamlines connecting pairs of regions from the 268-node Shen atlas [ 58 ]. Only connections with mean QA > 0.001 were retained to minimize noise. These 268 parcellated regions were grouped into 10 canonical networks [ 59 ]: medial frontal (MF), frontoparietal (FP), default mode (DMN), motor (Mot), visual I (Vis I), visual II (Vis II), visual association (Vas), salience (SAL), subcortical (SC), and cerebellar (CBL) networks. Network Control Theory and Controllability Metrics Dynamical Model The network control theory can simulate the consequences of a network’s underlying structural topology on its function dynamics [ 60 ]. The present study applied the network control theory framework to calculate brain controllability [ 31 – 33 ], implemented in MATLAB with in-house packages and code, from structural connectomes. Based on the previous studies [ 31 , 32 ], neural states can be mathematically described as simulated states ( x ) of network with k nodes overtime steps t using a simplified noise-free linear discrete-time and time-invariant model: $$\:\mathbf{x}\left(\varvec{t}+1\right)=\mathbf{A}\mathbf{x}\left(\varvec{t}\right)+{\mathbf{B}}_{\mathbf{k}}{\mathbf{u}}_{\mathbf{k}}\left(\varvec{t}\right)$$ where \(\:\varvec{x}\left(\varvec{t}\right)\) is a vector that describes the state of all brain regions at time \(\:t\) , while matrix A is a structural connection matrix. K is the set of nodes that can be controlled independently. The input matrix defines the control nodes. The \(\:{\mathbf{u}}_{\mathbf{K}}\left(\varvec{t}\right)\) is the control signal injected into the network via the control nodes, while B is the input matrix. Controllability Metrics Subsequently, the influence of each region on brain function is quantified using the metric of controllability. Controllability of a dynamical system refers to the possibility of driving the state of a dynamical system to a specific target state by means of an external control input. Classic results in the control theory ensure that controllability of the network from the set of network nodes k is equivalent to the controllability Gramian W K being invertible, where \(\:{\varvec{W}}_{\varvec{k}}=\:\sum\:_{\varvec{\tau\:}=0}^{\varvec{\infty\:}}{\varvec{A}}^{\varvec{\pi\:}}{\varvec{B}}_{\varvec{k}}{\varvec{B}}_{\varvec{k}}^{\varvec{T}}{\varvec{A}}^{\varvec{\pi\:}}\) . We utilize this framework to choose control nodes one at a time, and thus the input matrix B in fact reduces to a one-dimensional vector. Based on this network control theory framework, we examined modal controllability, which quantifies the ability of a brain region to drive the network into difficult-to-reach states that require substantial control energy [ 31 ]. Such transitions may correspond to cognitively demanding or less frequently engaged brain states [ 32 ], which are particularly relevant in the context of neurodevelopment and psychopathology. Modal controllability is computed from the eigenvectors [ \(\:{\varvec{V}}_{\varvec{i}\varvec{j}}\) ] of the adjacency matrix A . Here, we define the measurement of modal controllability as \(\:{\varvec{\Phi\:}}_{\varvec{i}}={\sum\:}_{\varvec{j}=1}^{\varvec{N}}(1-{\varvec{\lambda\:}}_{\varvec{j}}^{2}(\varvec{A}\left){\varvec{V}}_{\varvec{i}\varvec{j}}^{2}\right)\) of all N modes \(\:\:\:{\varvec{\lambda\:}}_{0}\left(\varvec{A}\right),...,\:\:{\varvec{\lambda\:}}_{\mathbf{N}-1}\left(\varvec{A}\right)\) from brain region i , following the definition of previous studies. PRSs Calculation Base GWAS Datasets PRS ADHD and PRS ASD were derived for two childhood psychiatric conditions: attention deficit hyperactivity disorder (ADHD) [ 61 ], and autism spectrum disorder (ASD) [ 62 ], using GWAS summary statistics from cohorts of European ancestry. These disorders were selected based on their well-powered GWAS, relevance to early development, and broad clinical presentations. The base datasets included single nucleotide polymorphisms (SNPs) with reported effect alleles, odds ratios (ORs), P-values, and allele frequencies. Full details of the GWAS samples and cohort sizes are provided in Supplementary Table 2 . Target Genotype Dataset and Quality Control Due to the reduced performance of cross-ancestry prediction [ 63 , 64 ], analyses were restricted to a subset of participants of confirmed European ancestry. Genotype data were obtained from the ABCD Study (Release 5.1) and processed using Axiom Analysis Suite (version 2.11) from raw intensities files generated by the Affymetrix Smokescreen array. Biological samples were derived from either saliva or whole blood samples. Participants were selected based on genotyping call rate, concordance between reported and genetic sex, and identity-by-state (IBS) estimates to minimize cryptic relatedness. Initial quality control (QC) was conducted by the ABCD consortium and included call rate thresholds, signal intensity checks, and marker-level filters, followed by the RICOPILI pipeline for imputation-ready QC. Ancestry-specific quality control was then applied separately to participants of European ancestry and to the remaining trans-ancestry group. SNPs were filtered based on missingness rate (< 5%), Hardy-Weinberg equilibrium (HWE 3 SD from the mean). Genotype phasing and imputation were performed using the TOPMed imputation server. Post-imputation SNPs were retained only if they had a minor allele frequency (MAF) ≥ 0.005 and imputation quality INFO score ≥ 0.9. A pruned set of SNPs (clumping threshold r² = 0.1) was used to calculate the top 10 genetic principal components (PCs) using PLINK [ 65 ]. Due to the reduced performance of cross-ancestry prediction [ 63 , 64 ], all PRS analyses were restricted to a subset of participants of confirmed European ancestry. PRS Computation PRS were computed using PRSice-2 (version 2.3.5) [ 66 ] with clumping and thresholding parameters set as clump-kb 250kb, clump-r² 0.1, and clump-p 1.0. For each participant, PRS were calculated as the sum of risk alleles (0, 1, or 2) at each locus, weighted by the corresponding log-transformed OR (logOR) from the base GWAS. Resulting scores were standardized (z-scored) prior to inclusion in downstream analyses. Covariates To reduce potential confounding, all analyses were adjusted for a consistent set of covariates, including age, sex, race, study site, handedness, pubertal status [ 67 ], and the first 5 genetic principal components. Categorical variables (sex, race, handedness) were dummy-coded prior to inclusion in statistical models. Statistical Analysis All continuous variables were standardized (Z-score) prior to analysis to account for differences in measurement scales. The Benjamini-Hochberg false discovery rate (FDR) correction was applied to adjust for multiple comparisons, and adjusted q -values were reported [ 68 ]. Association Analysis at Baseline The linear mixed-effects models were employed to examine the pairwise associations among ELA latent factors, controllability measures, and behavioral assessments at baseline. In each model, controllability or behavioral measure was used as the dependent variable, ELA factor and all covariates as fixed effects, and the study site as a random effect [ 69 ]. The estimated regression coefficient ( β ), standard error and p value were reported. Additionally, we estimated pairwise Pearson correlation between controllability and behavioral measures. Stratified Analysis To evaluate the robustness and generalizability of the observed association, sex-stratified and ancestry-stratified analyses were performed. Models were re-estimated separately for males and females, and again for participants of European versus non-European ancestry. To formally assess the similarity of results across subgroups, we calculated Pearson correlation between stratified effect sizes across all brain networks or behavioral outcomes. When low cross-group consistency was observed, additional interaction analyses between sex or ancestry group and each key independent variable were conducted in the full sample models. These interaction models allowed us to directly assess whether the strength or direction of associations differed significantly across sex or ancestry groups. Mediation and Moderated Mediation Analyses When brain controllability was significantly associated with both ELA latent factors and behavioral measures, mediation analyses were conducted to test whether the controllability measure mediated the associations between ELA factors and behavioral measures. Indirect effects were estimated using a bias-corrected bootstrap procedure with 5000 iterations, and mediation was considered statistically significant if the 95% CI did not include zero. For each significant mediation pathway, moderated mediation analyses were then conducted in the European subsample to examine whether the indirect effect varied by genetic risks for ADHD and ASD. Participants were divided into low (15%), medium (70%), and high (15%) genetic risk subgroups based on the 16th, and 84th percentiles, consistent with the default settings in the PROCESS macro for R [ 70 ]. Conditional indirect effects were estimated for each group using the same bootstrap approach. Longitudinal Analysis A cross-lagged panel model (CLPM), implemented with the lavaan package in R [ 71 ], was applied to evaluate associations between baseline brain controllability measures and 2-year follow-up behavioral outcomes, and vice versa. These analyses were performed in participants with follow-up behavioral assessments (5,115 for CBCL and 3,999 for crystallized intelligence). Prior to model fitting, both brain controllability metrics and behavioral scores (CBCL scales and cognition) at baseline and 2-year follow-up were residualized by regressing out all covariates using linear models. The resulting residuals were then standardized (z-scored) to ensure comparability across variables. Cross-lagged associations were estimated between baseline and follow-up measures of brain and behavior. Each CLPM included autoregressive paths (baseline to follow-up) and cross-lagged paths (brain to behavior and vice versa), with correlated residuals specified at both timepoints. All models were estimated using the sem function from the lavaan R package with full-information maximum likelihood (ML) estimation and robust (Huber-White) standard errors. To control for multiple testing, false discovery rate (FDR) correction was applied across all p -values derived from cross-lagged paths. Results Demographics A total of 7,970 participants were included ( Supplementary Fig. 1 ), comprising 4,141 males (52.0%) and 3,829 females (48.0%). The mean age of the participant was 9.47 years at baseline (standard deviation [SD] = 0.51; range = 9–11 years). Participants were identified as Asian (1.7%), Black (15.0%), Hispanic (20.1%), White (53.0%), and Other Race (10.2%). Detailed demographic characteristics of the participants are summarized in Table 1 . Table 1 Demographic Characteristics of Participants Included in the Study. Demographic Characteristics No. (%/SD) Age, mean (SD), y 9.47 (0.51) Sex Male 4141 (52.0) Female 3 829 Pubertal Development Scale Score -0.008 (0.97) Handedness Right-handed 6365 (79.9) Left-handed 560 (7.0) Mix-handed 1045 (13.1) Race a Asian 136 (1.7) Black 1193 (15.0) Hispanic 1602 (20.1) White 4228 (53.0) Other b 811 (10.2) Child Behavior Checklist (CBCL) c CBCL Internal 48.6 (10.6) CBCL External 45.7 (10.2) CBCL Totalprob 46.0 (11.2) NIH Toolbox Cognition d NIH FluidComp 45.9 (10.8) NIH CrystComp 51.4 (11.0) NIH TotalComp 48.1 (10.8) a Race was reported by each participant’s caregiver. b Other includes American Indian or Alaska Native, Native Hawaiian, and Pacific Islander participants. c CBCL scores are derived from caregiver report using empirically based syndrome scales: internalizing (CBCL Internal), externalizing (CBCL External), and total problems (CBCL Totalprob). d Cognitive scores were obtained from the NIH Toolbox Cognition Battery, including fluid intelligence (NIH FluidComp), crystallized intelligence (NIH CrystComp), and total composite score (NIH TotalComp). ELA Dimensions To derive latent dimensions of early life adversity (ELA), we first applied quality control to 139 candidate variables ( Supplementary Fig. 1 and Table 1 ), retaining 67 indicators for exploratory factor analysis (EFA) ( Supplementary Fig. 2 ). The data were suitable for factor analysis, as evidenced by Kaiser-Meyer-Olkin statistic = 0.827 and Bartlett test of sphericity (χ² = 266 144, p < .001). To determine the optimal number of latent factors, we used Very Simple Structure (VSS), Minimum Average Partial (MAP), and model fit indices. VSS suggested four and five factors as optimal solutions under complexity levels 1 and 2 (VSS = 0.553 and 0.614, respectively). Although the MAP statistic reached its global minimum at nine factors (MAP = 0.0048), values were near-minimal at four (MAP = 0.0058) and five (MAP = 0.0054) factors, indicating that these solutions captured the most common variance. Moreover, root mean square error of approximation (RMSEA) decreased progressively and plateaued at five factors (RMSEA = 0.027), supporting the selection of a five-factor model ( Supplementary Fig. 3 ). Confirmatory factor analysis (CFA) of the five-factor solution showed excellent model fit in 10-fold cross-validation ( Supplementary Table 3 ), with cross-validated factor scores highly correlated with full-sample scores (Spearman ρ range, 0.99-1.00), supporting the stability and robustness of the five-factor model. The five latent ELA dimensions represented distinct psychosocial contexts (Fig. 1 , Supplementary Table 4 ). Factor 1 (F1) reflected paternal psychosocial adversity, including high loadings on father-related marital, occupational, legal, behavioral, and substance use problems. Factor 2 (F2) captured family conflict and hostile climate, characterized by frequent fights, criticism, physical aggression, and low conflict resolution. Factor 3 (F3) indicated low caregiver and emotional support, with high loadings on less frequent caregiver comfort, love expression, open communication, and emotional availability. Factor 4 (F4) represented caregiver psychopathology, with high loadings on caregiver depressive, anxiety, somatic, ADHD, antisocial, and avoidant symptoms. Factor 5 (F5) indexed socioeconomic disadvantages, with high loadings for family income (negative loading), parental education (negative loading), and neighborhood safety concerns (positive loading). Together, these dimensions offer a comprehensive, multidimensional characterization of ELA during early development. Associations between ELA and Brain Controllability Three ELA dimensions were significantly associated with brain controllability (Fig. 2 , Supplementary Table 5 ). We identified consistent positive correlations between F1 (paternal psychosocial adversity) and controllability in the medial frontal, motor, and salience networks ( β = 0.028–0.036; q < .05), F4 (caregiver psychopathology) and controllability in the salience, medial frontal, visual association, motor, frontoparietal, and cerebellar networks ( β = 0.030–0.047; q < .05), and F5 (socioeconomic disadvantage) and controllability in the cerebellar and salience networks ( β = 0.039–0.057; q < .05). These findings indicate that distinct ELA dimensions could consistently increase brain controllability, with both shared and unique effects. However, F2 and F3 were not significantly associated with controllability in any network. Associations between ELA and Behavioral Outcomes All five ELA factors were significantly associated with behavioral outcomes (Fig. 2 , Supplementary Table 6 ). Higher ELA factor scores, especially F4 scores, were consistently associated with greater internalizing, externalizing, and total behavioral problems on the CBCL ( β = 0.091 to 0.564; q < .05). All, but F4, were associated with lower cognition composite scores ( β = -0.029 to -0.167; q < .05), particularly for F5 ( β = -0.120 to -0.167; q < .05). Association between Controllability and Behavioral Outcomes Higher controllability in the medial frontal, frontoparietal, default mode, motor, and salience networks was significantly associated with greater behavioral problems ( β = 0.031 to 0.052; q < .05, Fig. 3 A /B, Supplementary Table 7 ). In particular, Higher medial frontal and motor network controllability was consistently associated with higher externalizing (MF: β = 0.044; Mot: β = 0.028) and internalizing (MF: β = 0.043; Mot: β = 0.036) symptoms (all q < .05). However, SAL network controllability was also correlated with externalizing symptoms ( β = 0.030; q < .05), while FP and DMN network controllability was additionally associated with internalizing symptoms (FP: β = 0.031; DMN: β = 0.033; q < .05). Worse cognitive performance was also linked to higher controllability across seven networks, including the medial frontal, frontoparietal, default mode, motor, salience, visual association, and subcortical networks (Fig. 3 C /D, Supplementary Table 8 ). Specifically, higher controllability in the medial frontal, default mode, and motor networks was linked to lower crystallized cognition composite scores ( β = -0.027 to -0.047; q < 0.05). Stratified Analyses In sex-stratified analyses, ELA-brain controllability and brain controllability-behavior associations were generally consistent across males and females ( r = 0.552 to 0.993, q < .05) ( Supplementary Fig. 6 ). However, there were several potential sex-specific effects. For ELA, F4 and F5 showed stronger associations with controllability in the cerebellar network for females ( Supplementary Fig. 7 ). F5 showed stronger associations with CBCL total problems in males ( β = 0.282 vs 0.241 in females, Supplementary Fig. 8 ). Controllability in the subcortical, frontoparietal, and medial frontal networks were more strongly associated with cognitive total composite score in females ( Supplementary Fig. 9 ). In contrast, ancestry-stratified analyses revealed divergence in specific domains ( Supplementary Fig. 10 ). Associations between ELA factors and brain controllability showed poor concordance across European and non-European participants ( r = 0.073, p = 0.614), suggesting ancestry-specific effects. Among participants of European ancestry, F5 was positively associated with controllability in all ten networks ( β = 0.042–0.088, q < .05, Supplementary Fig. 11 ), but these effects were attenuated or absent in non-European participants. By contrast, associations between ELA and behavioral outcomes, as well as brain controllability-behavior links, were broadly consistent across ancestry groups ( Supplementary Figs. 12–13 ). However, interaction analyses revealed significant moderation by ancestry in the associations between F5 and brain controllability ( q < .05, Supplementary Fig. 14, Supplementary Table 9 ), particularly in the frontoparietal, default mode, and medial frontal networks. All interaction coefficients were positive, indicating that the association between F5 and brain controllability was stronger among participants of European ancestry. In other words, increases in socioeconomic disadvantage were more strongly related to alterations in controllability in European than in non‑European participants. These findings underscore the importance of considering population-specific contexts in developmental neuroimaging studies. Mediation and Moderated Mediation Analyses We tested a fully moderated mediation framework to examine whether brain controllability mediated the effects of ELA on cognitive and behavioral outcomes, and whether these indirect effects varied by genetic risk for neurodevelopmental disorders (Fig. 4 A). Mediation analyses revealed that brain controllability in specific brain networks partially mediated the associations between ELA factors and behavioral problems and cognitive performance (Fig. 4 B and Supplementary Table 10 ). For instance, the medial frontal network controllability partially mediated the associations of F4 (caregiver psychopathology) and NIH cognitive scores, including the total composite (proportion mediated = 3.1%, indirect β = -0.002, 95% CI = -0.004 to -0.001), and crystallized composite (proportion mediated = 3.8%, indirect β = -0.002, 95% CI = -0.004 to -0.001). Medial frontal controllability also mediated the link between F4 and CBCL total problems (proportion mediated = 0.6%, indirect β = 0.001, 95% CI = 0.0002 to 0.002). Similarly, the frontoparietal network controllability partially mediated the association between F4 and total cognition score (proportion mediated = 4.8%, indirect β = -0.001, 95% CI = -0.002 to -0.0003). To evaluate moderation by genetic risk, we stratified individuals based on PRS ADHD or PRS ASD . In individuals with higher PRS ADHD or PRS ASD , the indirect effects of F5 (socioeconomic disadvantage) on crystallized cognition via medial frontal network controllability were more pronounced compared to those with medium PRSs individuals. For example, among those with high PRS ASD , the moderated proportion mediated was 7.5% (indirect β = -0.012, 95% CI = -0.021 to -0.0006), compared to 3.7% in the medium PRS ASD group (indirect β = -0.006, 95% CI = -0.011 to -0.0003). Similarly, in high PRS ADHD individuals, the moderated proportion mediated was 3.9% (indirect β = -0.008, 95% CI = -0.017 to -0.0002) versus 3.6% in the medium PRS ADHD group (indirect β = -0.006, 95% CI = -0.011 to -0.0002). These results suggest that genetic vulnerability amplifies the impact of ELA on cognition through altered brain controllability, potentially in a dose-dependent relationship. Additional significant statistically moderated mediations are reported in Fig. 4 C /D , Supplementary Table 11 ). Longitudinal Associations Between Controllability and Behavioral Outcomes We performed cross-lagged panel modeling (CLPM) to examine bidirectional relationships between brain controllability and behavioral outcomes over time. Analyses included 5,115 participants with 2-year follow-up data for behavioral problem and in 3,999 for crystallized cognition. All combinations of 10 brain networks and 4 outcomes (3 behavioral and 1 cognitive measures) were tested, but only a subset demonstrated significant longitudinal associations. Autoregressive paths revealed strong within-domain stability over the 2-year interval, with all brain controllability and behavioral/cognitive variables showing significant self-prediction (e.g., medial frontal controllability: β = 0.531, q < .001; crystallized cognition: β = 0.665, q < .001; Fig. 4 E, F; Supplementary Tables 12, 13 ). Cross-domain associations were selective and often asymmetric (Fig. 4 E, F; Supplementary Tables 12, 13 ). For example, higher medial frontal controllability at baseline predicted lower crystallized cognition at follow-up ( β = -0.037; q = .003), and lower baseline cognition also predicted increased medial frontal controllability at follow-up ( β = -0.033; q = .020), suggesting a bidirectional relationship between control dynamics and cognition development (Fig. 4 E). In contrast, total behavioral problems at baseline predicted increased motor network controllability two years later ( β = 0.028; q = .041), but not vice versa ( β = 0.013; q = .296), indicating a unidirectional influence of behavior on brain development (Fig. 4 F). Discussion In this large, demographically diverse pediatric sample, we identified associations between specific ELA dimensions to distinct brain networks and, in turn, to cognitive and behavioral outcomes. Adversity related to caregiver psychopathology and paternal psychosocial disruption displayed the strongest associations with brain controllability in the medial frontal, frontoparietal, salience, and motor networks. These same networks were also the strongest mediators linking ELA to overall behavioral problems and crystallized cognition, suggesting that behavioral regulation and cognitive control circuits may be particularly sensitive to early psychosocial stress. Notably, the effects of ELA on behavior and cognition via controllability were moderated by polygenic risk for ADHD and ASD, highlighting a gene-environment interaction in shaping neurodevelopmental trajectories. Taken together, these findings provide compelling evidence for a structured neurodevelopmental cascade, from adversity to altered brain connectivity to behavioral impairment, refined further by inherited genetic risk. This work provides a systems-level framework for delineating how early adversity shapes neurodevelopmental pathways, offering new directions for predictive models and precision-oriented early interventions integrating neurobiology and genomics. Historically, it has been difficult to disentangle the many early life experiences that drive neurodevelopment [ 72 , 73 ]. A key strength of this study lies in the identification of five distinct and interpretable latent factors of ELA. This multidimensional approach incorporates the co-occurrence of ELA and supports growing evidence that ELA is not a unitary construct [ 72 ]. These results emphasize the importance of parsing the heterogeneity in ELA to characterize its downstream impact on the brain and behavior. For example, these latent factors exhibited unique associations with controllability and later behavior, which may have been occluded if ELA was treated as a unitary construct. Although no prior studies have examined controllability in the context of ELA, our findings are broadly consistent with previous reports linking parental psychosocial adversity, psychopathology, and socioeconomic disadvantages to altered white matter integrity and functional connectivity in key regulatory networks, including the salience, medial frontal, frontoparietal, and cerebellar networks [ 74 – 78 ]. We found that paternal psychosocial adversity (F1) and caregiver psychopathology (F4) were associated with the controllability in the salience, frontoparietal, visual, and motor networks. These networks are involved in affect regulation and executive function [ 79 – 82 ]. These brain network alterations may reflect heightened allostatic demands and adaptive neurodevelopmental tuning in response to chronic caregiver-related stress. [ 15 ] Interestingly, socioeconomic disadvantage and neighborhood insecurity (F5) were associated with controllability in the salience and cerebellar networks, suggesting partially distinct neural pathways for deprivation-based versus relational adversity. These findings are consistent with prior work demonstrating that different types of ELA, such as threat versus deprivation, may exert dissociable effects on brain organization and function [ 83 , 84 ]. The recurrent involvement of the salience network across multiple ELA dimensions in our findings underscores its potential role as a common transdiagnostic hub affected by early adversity. The well-established functions of salience network in detecting salient internal and external stimuli, mediating switching between major brain networks, and coordinating allostatic responses may account for its vulnerability to early environmental stressors [ 85 ]. From a neurodevelopmental perspective, repeated stress exposure may recalibrate the controllability of salience network, potentially enhancing their influence over dynamic brain states in an effort to maintain functional homeostasis. This hypothesis is supported by emerging evidence that early life adversity may accelerate the maturation or functional coupling of salience-related circuits [ 73 , 86 – 89 ]. Consistent with this view, a study reported that young adults with a history of childhood trauma exhibited increased network switching rates, reflecting reduced temporal stability of functional connectivity, especially within the frontoparietal and cingulo-opercular networks [ 90 ]. These results suggest that early adversity may lead to more dysregulated network dynamics, potentially increasing the allostatic load required to maintain cognitive and emotional homeostasis. Notably, the ancestry‑specific moderation was observed for the association between the socioeconomic adversity (F5) and brain controllability. Participants of European ancestry showed steeper positive slopes between F5 and controllability in the medial frontal, frontoparietal, default mode, and salience networks. One possible interpretation is that, in relatively resource-advantaged contexts, socioeconomic adversity may represent a greater deviation from normative expectations, thereby exerting a stronger physiological and neurodevelopmental impact. This interpretation aligns with prior research suggesting that the magnitude of brain-environment associations depend not only on the severity of exposure but also on the distributional range and contextual meaning of socioeconomic risk [ 75 ]. In several brain networks, higher controllability associated with greater behavioral problems and worse cognitive abilities. Controllability reflects the theoretical ability of a network to drive the brain into difficult-to-reach states [ 31 ]. Thus, higher controllability may indicate greater neural instability or inefficient regulation, especially under higher allostatic loads. Further, controllability typically increases in adolescents [ 32 ]. Greater controllability in the context of ELA may also reflect accelerated development [ 9 , 91 ], which is commonly observed in response to ELAs. These associations emphasize that higher controllability should not be interpreted as uniformly advantageous or disadvantageous. Together, the findings highlight the importance of contextualizing controllability and other neuroimaging measures within developmental histories when evaluating their functional significance. Mediation analyses revealed that brain controllability served as a partial mechanism linking ELA to both cognitive deficits and behavioral problems. These indirect effects were most prominent through the medial frontal and frontoparietal networks, regions central to cognitive control and affective processing [ 92 – 94 ]. Conceptually, these results are consistent with a prior network control theory study showing that lower control energy needed to engage frontoparietal networks associates with better executive performance and partly mediates age-related gains in cognition [ 95 ]. Further, our results extend previous work by demonstrating that controllability can also act as a partial mediator of adversity-related outcomes. Notably, these mediation effects were further moderated by polygenic risk for ADHD and ASD. The inclusion of these mediation and moderated mediation models is a strength and nuances the understanding of individual variability in neurodevelopment. Overall, they support a "double-hit" model, where ELA and genetic vulnerability jointly influence neurodevelopment through alterations in brain controllability [ 96 , 97 ]. The longitudinal findings indicate that bidirectional associations between brain controllability and behavior exist, indicating that cognitive abilities and controllability mutually shape each other during adolescence. This dynamic aligns with the sensorimotor-association axis of cortical maturation, wherein higher order association regions involved in cognitive functions undergo prolonged development and plasticity [ 98 ]. In contrast, the unidirectional effect of early behavioral problems on later motor network controllability indicates that dysregulation may shape the motor network development. This finding is partially supported by structural imaging studies showing that persistent externalizing behaviors predict altered subcortical maturation and disrupted prefrontal-limbic coupling [ 99 ]. However, cross-lagged panel models account for temporal stability using autoregressive paths and do not distinguish between-person differences and within-person fluctuations. These models reflect group averages, rather than casual individual differences [ 71 ]. This study has several limitations. First, the ELA factors were based on caregiver and child reports of past experiences, which could be influenced by recall or reporting bias. Second, PRS analyses were limited to participants of European ancestry; findings may not extend to more diverse populations. Finally, our models examined only linear associations and did not account for possible nonlinear effects. Using data from the ABCD study, we revealed that ELA dimension-specific influences on brain controllability, which in turn affected behavioral outcomes in children. Genetic vulnerability further moderated these relationships, underscoring the interplay between genes and environment in shaping neurodevelopment. Ultimately, identifying individual-level brain markers (e.g., controllability) could help improve early risk detection and guide personalized intervention strategies. Declarations Acknowledgements This study was supported by funding from the Tianjin Natural Science Foundation (24ZXGQSY00040) and Tianjin Key Medical Discipline Construction Project (TJYXZDXK-3-008C) for Huaigui Liu. The ABCD study is supported by the National Institutes of Health and additional federal partners under award numbers U01DA041048, U01DA050989, U01DA051016, U01DA041022, U01DA051018, U01DA051037, U01DA050987, U01DA041174, U01DA041106, U01DA041117, U01DA041028, U01DA041134, U01DA050988, U01DA051039, U01DA041156, U01DA041025, U01DA041120, U01DA051038, U01DA041148, U01DA041093, U01DA041089, U24DA041123, and U24DA041147. A full list of federal partners is available at https://abcdstudy.org/federal-partners.html . Conflict of Interest : None reported. Data availability Data used in the preparation of this article were obtained from the Adolescent Brain Cognitive Development Study® ( https://abcdstudy . org), held in the NIMH Data Archive (NDA). Only researchers with an approved NDA Data Use Certification (DUC) may obtain ABCD Study data. Code availability We made use of publicly available software and tools. The publicly available tools used in our analyses are described in Methods. The core codes are available at https://github.com/haigui0913/ELA_Controllability_ Behavior_ABCD. References Kim B, Royle M. Annual Research Review: Mapping the multifaceted approaches and impacts of adverse childhood experiences - an umbrella review of meta-analyses. J Child Psychol Psychiatry 2025; 66(4): 399–416. Wade M, Wright L, Finegold KE. The effects of early life adversity on children's mental health and cognitive functioning. Transl Psychiatry 2022; 12(1): 244. Luby JL, Barch D, Whalen D, Tillman R, Belden A. Association Between Early Life Adversity and Risk for Poor Emotional and Physical Health in Adolescence: A Putative Mechanistic Neurodevelopmental Pathway. JAMA Pediatr 2017; 171(12): 1168–1175. Hughes K, Bellis MA, Hardcastle KA, Sethi D, Butchart A, Mikton C et al. The effect of multiple adverse childhood experiences on health: a systematic review and meta-analysis. Lancet Public Health 2017; 2(8): e356-e366. Carr CP, Martins CM, Stingel AM, Lemgruber VB, Juruena MF. The role of early life stress in adult psychiatric disorders: a systematic review according to childhood trauma subtypes. J Nerv Ment Dis 2013; 201(12): 1007–1020. Kessler RC, Davis CG, Kendler KS. Childhood adversity and adult psychiatric disorder in the US National Comorbidity Survey. Psychol Med 1997; 27(5): 1101–1119. Brieant A, Vannucci A, Nakua H, Harris J, Lovell J, Brundavanam D et al. Characterizing the dimensional structure of early-life adversity in the Adolescent Brain Cognitive Development (ABCD) Study. Dev Cogn Neurosci 2023; 61: 101256. McLaughlin KA, Sheridan MA, Humphreys KL, Belsky J, Ellis BJ. The Value of Dimensional Models of Early Experience: Thinking Clearly About Concepts and Categories. Perspect Psychol Sci 2021; 16(6): 1463–1472. Ellis BJ, Sheridan MA, Belsky J, McLaughlin KA. Why and how does early adversity influence development? Toward an integrated model of dimensions of environmental experience. Dev Psychopathol 2022; 34(2): 447–471. Schafer JL, McLaughlin KA, Manfro GG, Pan P, Rohde LA, Miguel EC et al. Threat and deprivation are associated with distinct aspects of cognition, emotional processing, and psychopathology in children and adolescents. Dev Sci 2023; 26(1): e13267. Zeanah CH, Egger HL, Smyke AT, Nelson CA, Fox NA, Marshall PJ et al. Institutional rearing and psychiatric disorders in Romanian preschool children. Am J Psychiatry 2009; 166(7): 777–785. Deater-Deckard K, Mullineaux PY, Beekman C, Petrill SA, Schatschneider C, Thompson LA. Conduct problems, IQ, and household chaos: a longitudinal multi-informant study. J Child Psychol Psychiatry 2009; 50(10): 1301–1308. Tomassi S, Tosato S, Mondelli V, Faravelli C, Lasalvia A, Fioravanti G et al. Influence of childhood trauma on diagnosis and substance use in first-episode psychosis. Br J Psychiatry 2017; 211(3): 151–156. May A. Experience-dependent structural plasticity in the adult human brain. Trends Cogn Sci 2011; 15(10): 475–482. Nelson CA, Sullivan EF, Valdes V. Early adversity alters brain architecture and increases susceptibility to mental health disorders. Nat Rev Neurosci 2025; 26(10): 642–656. Vaidya N, Marquand AF, Nees F, Siehl S, Schumann G. The impact of psychosocial adversity on brain and behaviour: an overview of existing knowledge and directions for future research. Mol Psychiatry 2024; 29(10): 3245–3267. Sheridan MA, Mukerji CE, Wade M, Humphreys KL, Garrisi K, Goel S et al. Early deprivation alters structural brain development from middle childhood to adolescence. Sci Adv 2022; 8(40): eabn4316. Sheridan MA, Fox NA, Zeanah CH, McLaughlin KA, Nelson CA, 3rd. Variation in neural development as a result of exposure to institutionalization early in childhood. Proc Natl Acad Sci U S A 2012; 109(32): 12927–12932. Mehta MA, Golembo NI, Nosarti C, Colvert E, Mota A, Williams SC et al. Amygdala, hippocampal and corpus callosum size following severe early institutional deprivation: the English and Romanian Adoptees study pilot. J Child Psychol Psychiatry 2009; 50(8): 943–951. Gee DG, Gabard-Durnam LJ, Flannery J, Goff B, Humphreys KL, Telzer EH et al. Early developmental emergence of human amygdala-prefrontal connectivity after maternal deprivation. Proc Natl Acad Sci U S A 2013; 110(39): 15638–15643. Lim L, Radua J, Rubia K. Gray matter abnormalities in childhood maltreatment: a voxel-wise meta-analysis. Am J Psychiatry 2014; 171(8): 854–863. Busso DS, McLaughlin KA, Brueck S, Peverill M, Gold AL, Sheridan MA. Child Abuse, Neural Structure, and Adolescent Psychopathology: A Longitudinal Study. J Am Acad Child Adolesc Psychiatry 2017; 56(4): 321–328 e321. Gold AL, Sheridan MA, Peverill M, Busso DS, Lambert HK, Alves S et al. Childhood abuse and reduced cortical thickness in brain regions involved in emotional processing. J Child Psychol Psychiatry 2016; 57(10): 1154–1164. Gerin MI, Viding E, Herringa RJ, Russell JD, McCrory EJ. A systematic review of childhood maltreatment and resting state functional connectivity. Dev Cogn Neurosci 2023; 64: 101322. Bick J, Zhu T, Stamoulis C, Fox NA, Zeanah C, Nelson CA. Effect of early institutionalization and foster care on long-term white matter development: a randomized clinical trial. JAMA Pediatr 2015; 169(3): 211–219. Hanson JL, Adluru N, Chung MK, Alexander AL, Davidson RJ, Pollak SD. Early neglect is associated with alterations in white matter integrity and cognitive functioning. Child Dev 2013; 84(5): 1566–1578. Dennison MJ, Rosen ML, Sambrook KA, Jenness JL, Sheridan MA, McLaughlin KA. Differential Associations of Distinct Forms of Childhood Adversity With Neurobehavioral Measures of Reward Processing: A Developmental Pathway to Depression. Child Dev 2019; 90(1): e96-e113. Dufford AJ, Kim P. Family Income, Cumulative Risk Exposure, and White Matter Structure in Middle Childhood. Front Hum Neurosci 2017; 11: 547. Tendolkar I, Martensson J, Kuhn S, Klumpers F, Fernandez G. Physical neglect during childhood alters white matter connectivity in healthy young males. Hum Brain Mapp 2018; 39(3): 1283–1290. Choi J, Jeong B, Polcari A, Rohan ML, Teicher MH. Reduced fractional anisotropy in the visual limbic pathway of young adults witnessing domestic violence in childhood. Neuroimage 2012; 59(2): 1071–1079. Gu S, Pasqualetti F, Cieslak M, Telesford QK, Yu AB, Kahn AE et al. Controllability of structural brain networks. Nat Commun 2015; 6: 8414. Tang E, Giusti C, Baum GL, Gu S, Pollock E, Kahn AE et al. Developmental increases in white matter network controllability support a growing diversity of brain dynamics. Nat Commun 2017; 8(1): 1252. Sun H, Jiang R, Dai W, Dufford AJ, Noble S, Spann MN et al. Network controllability of structural connectomes in the neonatal brain. Nat Commun 2023; 14(1): 5820. Johnson D, Policelli J, Li M, Dharamsi A, Hu Q, Sheridan MA et al. Associations of Early-Life Threat and Deprivation With Executive Functioning in Childhood and Adolescence: A Systematic Review and Meta-analysis. JAMA Pediatr 2021; 175(11): e212511. Perevoznikova A, Berger P, Grosse Wiesmann C, White LO, Keil J. Maltreatment and executive functioning in childhood and adolescence - A multilevel meta-analysis. Child Abuse Negl 2025; 169(Pt 2): 107663. Hahn T, Winter NR, Ernsting J, Gruber M, Mauritz MJ, Fisch L et al. Genetic, individual, and familial risk correlates of brain network controllability in major depressive disorder. Mol Psychiatry 2023; 28(3): 1057–1063. Li Q, Yao L, You W, Liu J, Deng S, Li B et al. Controllability of Functional Brain Networks and Its Clinical Significance in First-Episode Schizophrenia. Schizophr Bull 2023; 49(3): 659–668. Tian Y, Shi W, Tao Q, Yang H, Guo H, Wen B et al. Altered controllability of functional brain networks in obsessive-compulsive disorder. J Psychiatr Res 2025; 182: 522–529. Caspi A, Moffitt TE. Gene-environment interactions in psychiatry: joining forces with neuroscience. Nat Rev Neurosci 2006; 7(7): 583–590. Belsky J, Jonassaint C, Pluess M, Stanton M, Brummett B, Williams R. Vulnerability genes or plasticity genes? Mol Psychiatry 2009; 14(8): 746–754. Sullivan PF, Daly MJ, O'Donovan M. Genetic architectures of psychiatric disorders: the emerging picture and its implications. Nat Rev Genet 2012; 13(8): 537–551. Lewis CM, Vassos E. Polygenic risk scores: from research tools to clinical instruments. Genome Med 2020; 12(1): 44. Ranlund S, Rosa MJ, de Jong S, Cole JH, Kyriakopoulos M, Fu CHY et al. Associations between polygenic risk scores for four psychiatric illnesses and brain structure using multivariate pattern recognition. Neuroimage Clin 2018; 20: 1026–1036. Griffiths KR, Braund TA, Kohn MR, Clarke S, Williams LM, Korgaonkar MS. Structural brain network topology underpinning ADHD and response to methylphenidate treatment. Transl Psychiatry 2021; 11(1): 150. Adisetiyo V. Revealing the Structural Neural Circuitry of Attention Deficit Hyperactivity Disorder With Diffusion MRI: Implications for Future Diagnosis and Treatment. AJR Am J Roentgenol 2018; 210(4): 731–733. Li SJ, Wang Y, Qian L, Liu G, Liu SF, Zou LP et al. Alterations of White Matter Connectivity in Preschool Children with Autism Spectrum Disorder. Radiology 2018; 288(1): 209–217. Rudie JD, Brown JA, Beck-Pancer D, Hernandez LM, Dennis EL, Thompson PM et al. Altered functional and structural brain network organization in autism. Neuroimage Clin 2012; 2: 79–94. Hagler DJ, Jr., Hatton S, Cornejo MD, Makowski C, Fair DA, Dick AS et al. Image processing and analysis methods for the Adolescent Brain Cognitive Development Study. Neuroimage 2019; 202: 116091. Barch DM, Albaugh MD, Baskin-Sommers A, Bryant BE, Clark DB, Dick AS et al. Demographic and mental health assessments in the adolescent brain and cognitive development study: Updates and age-related trajectories. Dev Cogn Neurosci 2021; 52: 101031. Hoffman EA, Clark DB, Orendain N, Hudziak J, Squeglia LM, Dowling GJ. Stress exposures, neurodevelopment and health measures in the ABCD study. Neurobiol Stress 2019; 10: 100157. Kuhn M, Wing J, Weston S, Williams A, Keefer C, Engelhardt A et al. Package ‘caret’. 2020; 223(7): 48. Xu J, Liu N, Polemiti E, Garcia-Mondragon L, Tang J, Liu X et al. Effects of urban living environments on mental health in adults. Nat Med 2023; 29(6): 1456–1467. Van Buuren S, Groothuis-Oudshoorn KJJoss. mice: Multivariate imputation by chained equations in R. 2011; 45: 1–67. Luciana M, Bjork JM, Nagel BJ, Barch DM, Gonzalez R, Nixon SJ et al. Adolescent neurocognitive development and impacts of substance use: Overview of the adolescent brain cognitive development (ABCD) baseline neurocognition battery. Dev Cogn Neurosci 2018; 32: 67–79. Gershon RC, Wagster MV, Hendrie HC, Fox NA, Cook KF, Nowinski CJ. NIH toolbox for assessment of neurological and behavioral function. Neurology 2013; 80(11 Suppl 3): S2-6. Yeh FC, Liu L, Hitchens TK, Wu YL. Mapping immune cell infiltration using restricted diffusion MRI. Magn Reson Med 2017; 77(2): 603–612. Yeh FC, Wedeen VJ, Tseng WY. Generalized q-sampling imaging. IEEE Trans Med Imaging 2010; 29(9): 1626–1635. Shen X, Tokoglu F, Papademetris X, Constable RT. Groupwise whole-brain parcellation from resting-state fMRI data for network node identification. Neuroimage 2013; 82: 403–415. Noble S, Spann MN, Tokoglu F, Shen X, Constable RT, Scheinost D. Influences on the Test-Retest Reliability of Functional Connectivity MRI and its Relationship with Behavioral Utility. Cereb Cortex 2017; 27(11): 5415–5429. Yan G, Vertes PE, Towlson EK, Chew YL, Walker DS, Schafer WR et al. Network control principles predict neuron function in the Caenorhabditis elegans connectome. Nature 2017; 550(7677): 519–523. Demontis D, Walters GB, Athanasiadis G, Walters R, Therrien K, Nielsen TT et al. Genome-wide analyses of ADHD identify 27 risk loci, refine the genetic architecture and implicate several cognitive domains. Nature genetics 2023; 55(2): 198–208. Grove J, Ripke S, Als TD, Mattheisen M, Walters RK, Won H et al. Identification of common genetic risk variants for autism spectrum disorder. Nature genetics 2019; 51(3): 431–444. Cardon LR, Palmer LJ. Population stratification and spurious allelic association. Lancet 2003; 361(9357): 598–604. Genomes Project C, Auton A, Brooks LD, Durbin RM, Garrison EP, Kang HM et al. A global reference for human genetic variation. Nature 2015; 526(7571): 68–74. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007; 81(3): 559–575. Choi SW, O'Reilly PF. PRSice-2: Polygenic Risk Score software for biobank-scale data. Gigascience 2019; 8(7). Petersen AC, Crockett L, Richards M, Boxer A. A self-report measure of pubertal status: Reliability, validity, and initial norms. J Youth Adolesc 1988; 17(2): 117–133. Korthauer K, Kimes PK, Duvallet C, Reyes A, Subramanian A, Teng M et al. A practical guide to methods controlling false discoveries in computational biology. Genome Biol 2019; 20(1): 118. Bates D, Mächler M, Bolker B, Walker S. Fitting Linear Mixed-Effects Models Using lme4. Journal of Statistical Software 2015; 67(1): 1–48. AF H. Introduction to mediation, moderation, and conditional process analysis: A regression-based approach . Guilford Press2017. Hamaker EL, Kuiper RM, Grasman RP. A critique of the cross-lagged panel model. Psychol Methods 2015; 20(1): 102–116. Smith KE, Pollak SD. Rethinking Concepts and Categories for Understanding the Neurodevelopmental Effects of Childhood Adversity. Perspect Psychol Sci 2021; 16(1): 67–93. Gee DG. Early Adversity and Development: Parsing Heterogeneity and Identifying Pathways of Risk and Resilience. Am J Psychiatry 2021; 178(11): 998–1013. Kemp JVA, Bernier E, Lebel C, Kopala-Sibley DC. Associations Between Parental Mood and Anxiety Psychopathology and Offspring Brain Structure: A Scoping Review. Clin Child Fam Psychol Rev 2022; 25(1): 222–247. Farah MJ. The Neuroscience of Socioeconomic Status: Correlates, Causes, and Consequences. Neuron 2017; 96(1): 56–71. Noble KG, Houston SM, Brito NH, Bartsch H, Kan E, Kuperman JM et al. Family income, parental education and brain structure in children and adolescents. Nat Neurosci 2015; 18(5): 773–778. Carozza S, Kletenik I, Astle D, Schwamm L, Dhand A. Whole-brain white matter variation across childhood environments. Proc Natl Acad Sci U S A 2025; 122(15): e2409985122. Holz NE, Berhe O, Sacu S, Schwarz E, Tesarz J, Heim CM et al. Early Social Adversity, Altered Brain Functional Connectivity, and Mental Health. Biol Psychiatry 2023; 93(5): 430–441. Schimmelpfennig J, Topczewski J, Zajkowski W, Jankowiak-Siuda K. The role of the salience network in cognitive and affective deficits. Front Hum Neurosci 2023; 17: 1133367. Marek S, Dosenbach NUF. The frontoparietal network: function, electrophysiology, and importance of individual precision mapping. Dialogues Clin Neurosci 2018; 20(2): 133–140. Li L, Wang L. Linking visual-frontoparietal network neural dynamics to spontaneous cognitive processing. Neuroimage 2025; 312: 121229. Mendoza G, Merchant H. Motor system evolution and the emergence of high cognitive functions. Prog Neurobiol 2014; 122: 73–93. McLaughlin KA, Weissman D, Bitran D. Childhood Adversity and Neural Development: A Systematic Review. Annu Rev Dev Psychol 2019; 1: 277–312. Chahal R, Miller JG, Yuan JP, Buthmann JL, Gotlib IH. An exploration of dimensions of early adversity and the development of functional brain network connectivity during adolescence: Implications for trajectories of internalizing symptoms. Dev Psychopathol 2022; 34(2): 557–571. Menon V, Uddin LQ. Saliency, switching, attention and control: a network model of insula function. Brain Struct Funct 2010; 214(5–6): 655–667. Lammertink F, van den Heuvel MP, Hermans EJ, Dudink J, Tataranno ML, Benders M et al. Early-life stress exposure and large-scale covariance brain networks in extremely preterm-born infants. Transl Psychiatry 2022; 12(1): 256. Herzberg MP, McKenzie KJ, Hodel AS, Hunt RH, Mueller BA, Gunnar MR et al. Accelerated maturation in functional connectivity following early life stress: Circuit specific or broadly distributed? Dev Cogn Neurosci 2021; 48: 100922. Callaghan BL, Tottenham N. The Stress Acceleration Hypothesis: Effects of early-life adversity on emotion circuits and behavior. Curr Opin Behav Sci 2016; 7: 76–81. Chan SY, Ngoh ZM, Ong ZY, Teh A, Kee MZL, Zhou JH et al. The influence of early-life adversity on the coupling of structural and functional brain connectivity across childhood. Nat Ment Health 2024; 2(1). Huang D, Liu Z, Cao H, Yang J, Wu Z, Long Y. Childhood trauma is linked to decreased temporal stability of functional brain networks in young adults. J Affect Disord 2021; 290: 23–30. McEwen BS, Gianaros PJ. Central role of the brain in stress and adaptation: links to socioeconomic status, health, and disease. Ann N Y Acad Sci 2010; 1186: 190–222. Vincent JL, Kahn I, Snyder AZ, Raichle ME, Buckner RL. Evidence for a frontoparietal control system revealed by intrinsic functional connectivity. J Neurophysiol 2008; 100(6): 3328–3342. Menon V, D'Esposito M. The role of PFC networks in cognitive control and executive function. Neuropsychopharmacology 2022; 47(1): 90–103. de la Vega A, Chang LJ, Banich MT, Wager TD, Yarkoni T. Large-Scale Meta-Analysis of Human Medial Frontal Cortex Reveals Tripartite Functional Organization. J Neurosci 2016; 36(24): 6553–6562. Cui Z, Stiso J, Baum GL, Kim JZ, Roalf DR, Betzel RF et al. Optimization of energy state transition trajectory supports the development of executive function during youth. Elife 2020; 9. Nigg J, Nikolas M, Burt SA. Measured gene-by-environment interaction in relation to attention-deficit/hyperactivity disorder. J Am Acad Child Adolesc Psychiatry 2010; 49(9): 863–873. Chaste P, Leboyer M. Autism risk factors: genes, environment, and gene-environment interactions. Dialogues Clin Neurosci 2012; 14(3): 281–292. Sydnor VJ, Larsen B, Bassett DS, Alexander-Bloch A, Fair DA, Liston C et al. Neurodevelopment of the association cortices: Patterns, mechanisms, and implications for psychopathology. Neuron 2021; 109(18): 2820–2846. Bos MGN, Wierenga LM, Blankenstein NE, Schreuders E, Tamnes CK, Crone EA. Longitudinal structural brain development and externalizing behavior in adolescence. J Child Psychol Psychiatry 2018; 59(10): 1061–1072. Additional Declarations The authors have declared there is NO conflict of interest to disclose Supplementary Files Supplementarymaterials.docx Supplementary materials Cite Share Download PDF Status: Under Review Version 1 posted Review # 2 received at journal 01 Apr, 2026 Reviewer # 3 agreed at journal 30 Mar, 2026 Reviewer # 2 agreed at journal 26 Mar, 2026 Reviewer # 1 agreed at journal 26 Mar, 2026 Reviewers invited by journal 16 Mar, 2026 Editor assigned by journal 04 Feb, 2026 Submission checks completed at journal 04 Feb, 2026 First submitted to journal 04 Feb, 2026 Unknown event 04 Feb, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8777490","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":607213401,"identity":"946c03e4-e72a-4d82-baf1-15b0b1d1ade9","order_by":0,"name":"Huaigui Liu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA5ElEQVRIie3OQQsBQRTA8cfU7OWx1+dgfYWRWqt8GFJcHNSWC6HUHp1dfAgpZ5riImflwn0P9kBOYoTksLtHh/k3M72m+dUA6HR/mHjsBQBBmkHiAOC8L2MQzoAJNcQiz/hjUSxSpNpxefYc4AbaHeySBcZwTmGkNK4LmfXUx9De44oKgKt2KBG7ipCZD+FUHVDTjiCN04e4eKP+IOdHkaZYBi/CUh5VgDCCbPyWhC0hZ9zNTEaU97DuOqFk3ZgF13bPMk05PfmXcs405HQXRlQMAVANSXXyqOeqxPV30Ol0Ot13d58IO+xoEaqDAAAAAElFTkSuQmCC","orcid":"","institution":"Yale university","correspondingAuthor":true,"prefix":"","firstName":"Huaigui","middleName":"","lastName":"Liu","suffix":""},{"id":607213402,"identity":"27010457-d6b0-46f3-ab4d-6e0379cd02ba","order_by":1,"name":"Yurong Jiang","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Yurong","middleName":"","lastName":"Jiang","suffix":""},{"id":607213403,"identity":"f7fbaea0-9630-4ccc-bfe6-bd50965843b4","order_by":2,"name":"Huili Sun","email":"","orcid":"https://orcid.org/0000-0002-4851-3046","institution":"Yale University","correspondingAuthor":false,"prefix":"","firstName":"Huili","middleName":"","lastName":"Sun","suffix":""},{"id":607213404,"identity":"7700d9ec-07d9-4bca-b7f3-e3c8e6dc9e3b","order_by":3,"name":"Matthew Rosenblatt","email":"","orcid":"https://orcid.org/0000-0002-3894-6198","institution":"Yale University","correspondingAuthor":false,"prefix":"","firstName":"Matthew","middleName":"","lastName":"Rosenblatt","suffix":""},{"id":607213405,"identity":"38dafa4c-c4c0-4f53-bacf-ff38db84e4a2","order_by":4,"name":"Jean Ye","email":"","orcid":"https://orcid.org/0000-0001-7928-324X","institution":"Yale University, New Haven, Connecticut, USA","correspondingAuthor":false,"prefix":"","firstName":"Jean","middleName":"","lastName":"Ye","suffix":""},{"id":607213406,"identity":"590b207a-930c-428d-b29a-534823bc9ce3","order_by":5,"name":"Chunshui Yu","email":"","orcid":"https://orcid.org/0000-0001-5648-5199","institution":"Tianjin Medical University General Hospital","correspondingAuthor":false,"prefix":"","firstName":"Chunshui","middleName":"","lastName":"Yu","suffix":""},{"id":607213407,"identity":"a3be2b62-5060-4e5c-97b0-8c308262f293","order_by":6,"name":"Dustin Scheinost","email":"","orcid":"https://orcid.org/0000-0002-6301-1167","institution":"Yale School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Dustin","middleName":"","lastName":"Scheinost","suffix":""}],"badges":[],"createdAt":"2026-02-03 15:06:29","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8777490/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8777490/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":104996620,"identity":"0051ed87-5f7f-4238-a608-6e057457e2a4","added_by":"auto","created_at":"2026-03-19 16:15:31","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":5347215,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStandardized Factor Loadings of Early Life Adversities for the 5-Factor Solution.\u003c/strong\u003e Heatmap showing standardized loadings (≥|0.45| labeled) of 67 adversity indicators on five latent factors (F1-F5). Higher factor scores represent greater exposure to adversity. Color scale indicates loading magnitude and direction (red=positive, blue=negative). Categories of ELA are color-coded on the left margin. Abbreviations: F1: Paternal psychosocial adversity, F2: Family conflict/hostile environment, F3: Low caregiver and emotional support, F4: Caregiver psychopathology, F5: Socioeconomic disadvantages.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8777490/v1/5d02e2093dc8cda8fa0f46e1.png"},{"id":104996616,"identity":"8e094394-5e7d-48cc-89ba-09bcb961491c","added_by":"auto","created_at":"2026-03-19 16:15:30","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":3980252,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAssociations of ELA Factors with Brain Controllability and Behavioral outcomes. \u003c/strong\u003eSolid lines represent significant positive associations (\u003cem\u003eq\u003c/em\u003e \u0026lt; .05) between ELA factors (F1-F5) and brain controllability across 10 canonical networks, with line thickness indicating regression coefficients. Dashed lines denote significant negative associations. Similarly, associations between ELA factors and behavioral/cognitive outcomes are shown on the right, with line thickness reflecting regression coefficients. Color shading in the lower-left triangle represents Pearson’s correlation coefficients (r) between brain controllability and behavioral outcomes. These correlation results are supplementary and provided for reference only (see \u003cstrong\u003eSupplementary Fig.4, 5\u003c/strong\u003e). Abbreviations: CBL, cerebellar network, DMN, default mode network, F1: Paternal psychosocial adversity, F2: Family conflict/hostile environment, F3: Low caregiver and emotional support, F4: Caregiver psychopathology, F5: Socioeconomic disadvantages, FP, frontoparietal network, MF, medial frontal network, Mot, motor network, SAL, salience network, SC, subcortical network, Vas, visual association network, Vis I, visual network I, Vis II, visual network II.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8777490/v1/39a81b351ff1316e8bcd3da1.png"},{"id":104996625,"identity":"b194319c-543f-451c-99be-f22dec09b229","added_by":"auto","created_at":"2026-03-19 16:15:31","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":16079108,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eAssociations Between Brain Controllability and Behavioral Outcomes.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAssociations of brain controllability with behavioral symptoms (A) and cognitive measures (C). Bars represent the −log10(\u003cem\u003eq\u003c/em\u003e) of FDR-corrected \u003cem\u003ep\u003c/em\u003e-values across ten canonical brain networks. The dashed line indicates the FDR significance threshold (\u003cem\u003eq\u003c/em\u003e = .05). (B, D) Cortical surface maps of standardized regression coefficients (\u003cem\u003eβ\u003c/em\u003e) for associations between network-level brain controllability and behavioral measures. Warmer colors indicate stronger positive associations, cooler colors indicate stronger negative associations. Abbreviations: CBL, cerebellar network, DMN, default mode network, FP, frontoparietal network, MF, medial frontal network, Mot, motor network, SAL, salience network, SC, subcortical network, Vas, visual association network, Vis I, visual network I, Vis II, visual network II.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8777490/v1/24faae79cbfbaa8786967fa3.png"},{"id":104996615,"identity":"e72e367a-03b4-4d56-89f2-40ff908a0c4d","added_by":"auto","created_at":"2026-03-19 16:15:30","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":13296590,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMediation, Moderated Mediation, and Longitudinal Path Models Linking ELA, Brain Controllability, and Behavioral Outcomes.\u003c/strong\u003e (A) The conceptual model of a fully moderated mediation framework, where ELA factors influence behavioral outcomes via brain controllability (indirect effect: Path A × Path B), with PRS moderating all three paths. (B) Mediation pathways from ELA factors to behavioral outcomes through brain controllability. (C, D) Moderated mediation pathways in individuals stratified by PRS\u003csub\u003eASD\u003c/sub\u003e (C) and PRS\u003csub\u003eADHD \u003c/sub\u003e(D) levels (low, medium, high). In (B-D), colors represent standardized regression coefficients (\u003cem\u003eβ\u003c/em\u003e), with significant indirect effects denoted by asterisks. (E) Cross-lagged panel model showing longitudinal bidirectional associations between controllability in the MF network and NIH crystallized cognition composite scores across baseline and 2-year follow-up. (F) CLPM results for the Mot network and CBCL total problem scores. Solid arrows indicate significant paths (\u003cem\u003eq\u003c/em\u003e \u0026lt; .05); Yellow solid arrows indicate significant positive effects; green dashed arrows indicate significant negative effects; gray arrows denote non-significant paths. Abbreviations: F1: Paternal psychosocial adversity, F2: Family conflict/hostile environment, F3: Low caregiver and emotional support, F4: Caregiver psychopathology, F5: Socioeconomic disadvantages, FP, frontoparietal network, MF, medial frontal network, Mot, motor network, SAL, salience network, Vas, visual association network.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-8777490/v1/7d1d83c0308bfe547b176438.png"},{"id":105035925,"identity":"734728a2-f6b8-47ff-b6c1-6b17a78d6a2f","added_by":"auto","created_at":"2026-03-20 07:27:00","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":39051343,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8777490/v1/4b9e0f7b-ad35-42a0-a1f8-ed0804ac224d.pdf"},{"id":104996629,"identity":"d1032513-0c4b-4624-877b-a32398a2c002","added_by":"auto","created_at":"2026-03-19 16:15:32","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":3432797,"visible":true,"origin":"","legend":"Supplementary materials","description":"","filename":"Supplementarymaterials.docx","url":"https://assets-eu.researchsquare.com/files/rs-8777490/v1/c1f8dee80c8a4ec82839c388.docx"}],"financialInterests":"The authors have declared there is \u003cb\u003eNO\u003c/b\u003e conflict of interest to disclose","formattedTitle":"Brain controllability mediates the effects of early life adversity on adolescent behavior and cognition moderated by genetic risks","fulltext":[{"header":"Introduction","content":"\u003cp\u003eEarly life adversity (ELA) is a potent and pervasive determinant of emotional, cognitive, and behavioral development. Mounting evidence links it with elevated risk for psychiatric disorders across the lifespan [\u003cspan additionalcitationids=\"CR2 CR3 CR4 CR5\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]. However, as adversity includes a wide-range of factors (e.g., caregiver psychopathology, emotional maltreatment, family dysfunction, socioeconomic disadvantage, neighborhood crime and trauma [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]), it is not a unitary construct. Furthermore, these distinct dimensions of ELA may exert differential effects on developmental trajectories [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. For instance, deprivation is linked to executive function deficits and significantly higher rates of ADHD [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e], unpredictability is shown to correlate with more conduct problems [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e], while physical abuse is associated with increased lifetime substance use [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. These distinctions motivate multidimensional models of ELA to better capture its heterogeneity and co-occurrence. These models offer a more precise framework for studying developmental consequences of ELA.\u003c/p\u003e \u003cp\u003eThe brain is particularly vulnerable to the effects of ELA [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Structural and functional neuroimaging studies have consistently demonstrated that ELA disrupts neurodevelopmental trajectories in domain-specific and region-specific ways [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. For example, deprivation-related adversity are associated with reduced total brain volume [\u003cspan additionalcitationids=\"CR18\" citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e] and premature adult-like connectivity between the amygdala and the medial prefrontal cortex [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. While maltreatment exhibit cortical thinning in prefrontal and temporal regions [\u003cspan additionalcitationids=\"CR22\" citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e] and heightened amygdala connectivity with the salience, default mode, and prefrontal networks [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. White matter abnormalities have also been reported across distinct types of adversity: frontostriatal alterations following early institutionalization, neglect and deprivation [\u003cspan additionalcitationids=\"CR26\" citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e], frontolimbic disruptions linked to socioeconomic disadvantage and physical neglect [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e], and reduced integrity of association fibers in individuals exposed to domestic violence [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Yet, despite this growing literature, a fundamental question remains: how do adversity-related alterations in brain networks translate into increased vulnerability for psychopathology?\u003c/p\u003e \u003cp\u003eOne promising avenue for addressing this question lies in network control theory, a computational framework grounded in systems neuroscience [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Rather than describing regional structure or function in isolation, network controllability quantifies the brain\u0026rsquo;s capacity to flexibly transition between diverse mental states, based on the underlying structural connections [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Thus, network control theory offers a biologically grounded and mathematically tractable marker of structural connections constrain and support brain dynamics. Controllability increases over young childhood through adolescence, putatively allowing for the emergence of adult-like executive functions [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. Given that executive functions are both developmentally sensitive and highly vulnerable to adversity [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e], controllability may serve as a critical link between structural network changes and behavioral outcomes. Further, controllability is altered in several psychiatric disorders [\u003cspan additionalcitationids=\"CR37\" citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. However, its role in adversity-psychopathology pathways remains largely unexplored.\u003c/p\u003e \u003cp\u003eImportantly, not all children exposed to ELA develop psychiatric disorders. Genetics may explain this variability [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. Previous studies estimate that up to 40\u0026ndash;80% of the variance in psychiatric disorders is heritable [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e], and polygenic risk scores (PRS) derived from genome-wide association studies (GWAS) now offer a scalable approach to quantify individual-level genetic liability [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. Prior studies have linked PRSs of psychiatric disorders to brain structure [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e], but few have examined their interaction with ELA in mediating neurodevelopmental pathways. Understanding these interactions is essential, especially in youth, when both environmental exposures and brain networks remain plastic.\u003c/p\u003e \u003cp\u003eIn the present study, we leveraged data from the ABCD study to investigate how distinct dimensions of ELA affect brain network controllability, and how these changes may mediate cognitive and behavioral outcomes. We also tested whether genetic risk for ADHD and ASD moderates these brain behavior pathways, and whether the observed effects persist over a two-year longitudinal follow-up. We selected risk for ADHD and ASD as they are early-onset neurodevelopmental conditions with high heritability and relevance to the age range of the ABCD cohort. Additionally, both are associated with altered structural connections [\u003cspan additionalcitationids=\"CR45 CR46\" citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. To this end, we applied dimensional modeling of ELA across 67 baseline indicators, computed controllability metrics across ten conical brain networks, and assessed behavioral and cognitive outcomes using the Child Behavior Checklist (CBCL) and NIH Toolbox at both baseline and follow-up. Analyses included linear mixed-effects models, mediation and moderated mediation frameworks, and cross-lagged panel model (CLPM) cross-lagged panel modeling to examine temporal dynamics. By integrating environmental, neural, behavioral, and genetic data in a large, diverse pediatric sample, this study aims to elucidate the mechanisms through which early adversity becomes biologically embedded, and to identify potential targets for early identification and intervention in psychiatric risk.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eParticipants\u003c/h2\u003e \u003cp\u003eAll participants were recruited by the ABCD study from the ABCD consortium (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://abcdstudy.org/index.html\u003c/span\u003e\u003cspan address=\"https://abcdstudy.org/index.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e, which includes over 11,000 children aged 9\u0026ndash;11 years, recruited from 22 sites across the United States with a diverse range of geographic, socioeconomic, ethnic, and health backgrounds [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. We used the Annual Curated Data Release 5.1. The study protocols were approved by the Institutional Review Boards (IRBs) at each participating site, with written informed consent obtained from youth and caregivers. Details about the participant cohort, data collection procedures, and preprocessing protocols are provided at the ABCD website (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://abcdstudy.org/scientists/protocols/\u003c/span\u003e\u003cspan address=\"https://abcdstudy.org/scientists/protocols/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e)\u003c/span\u003e and also are described elsewhere [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eEarly Life Adversity\u003c/h3\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eItem Selection and Data Cleaning\u003c/h2\u003e \u003cp\u003eELA was assessed at baseline using 139 variables selected from the ABCD baseline measures [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e, \u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e], which captured diverse adversity domains, including caregiving disruption, caregiver psychopathology, maltreatment, neighborhood violence, family and community support, socioeconomic disadvantage, and exposure to physical trauma (\u003cb\u003eSupplementary Table\u0026nbsp;1\u003c/b\u003e). All items were drawn from validated scales administered to the child, parent, or rated by trained research staff and were coded such that higher scores indicated greater adversity.\u003c/p\u003e \u003cp\u003eWe excluded 31 variables with \u0026gt;\u0026thinsp;50% missing and removed an additional 29 variables with zero or near-zero variance using the \u003cem\u003enearZeroVar\u003c/em\u003e function from the caret R package [\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]. Two redundant variables were manually removed. Furthermore, eight participants lacking age information were excluded. To avoid multicollinearity arising from familial relatedness, we excluded twins and triplets, retaining at most one participant per family among non-twin siblings. This exclusion was applied prior to all analyses, resulting in a final analytic sample of N\u0026thinsp;=\u0026thinsp;2,054 excluded participants. After these data cleaning procedures, the remaining variables underwent outlier detection, with outliers defined as values exceeding 4 \u0026times; median absolute deviation (MAD) of continuous environmental variables [\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e]. Finally, 10 ELAs variables with a missing rate\u0026thinsp;\u0026gt;\u0026thinsp;50% were excluded and a final set of 67 variables was included. Details are provided in \u003cb\u003eSupplementary Fig.\u0026nbsp;1, 2\u003c/b\u003e and \u003cb\u003eSupplementary Table\u0026nbsp;1\u003c/b\u003e. Missing values were handled using the \u003cem\u003emice\u003c/em\u003e (v3.16.0) [\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e] R package. Ten imputed datasets were generated using predictive mean matching (continuous variables) and logistic regression (categorical variables), with 10 iterations each.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eFactor Analysis and Validation\u003c/h3\u003e\n\u003cp\u003eTo reduce dimensionality and identify core latent adversity factors, we conducted an exploratory factor analysis (EFA) using principal axis factoring and \u003cem\u003eoblimin\u003c/em\u003e rotation. ELA variables were averaged across the 10 imputed datasets prior to factor analysis. Suitability for factor analysis was confirmed by Kaiser-Meyer-Olkin (KMO) values\u0026thinsp;\u0026gt;\u0026thinsp;0.60 and significant Bartlett\u0026rsquo;s test of sphericity (p \u0026lt; .001). The number of factors was determined by a combination of Very Simple Structure (VSS) and Minimum Average Partial (MAP) criteria, supplemented by model fit indices, including comparative fit index (CFI), root mean square error of approximation (RMSEA) and standardized root mean square residual (SRMR) across 1\u0026ndash;9 factor solutions.\u003c/p\u003e \u003cp\u003eIndicators with loadings\u0026thinsp;\u0026gt;\u0026thinsp;0.45 were retained as primary contributors to each factor. Final factor structures were validated through confirmatory factor analysis (CFA) using the \u003cem\u003elavaan\u003c/em\u003e R package (version 0.6\u0026ndash;19). To reduce overfitting, we applied tenfold cross-validation: in each fold, 90% of participants were used to estimate factor loadings, while factor scores were computed in the remaining 10%. This was iterated until all participants had out-of-sample estimates. The stability of factor scores was confirmed using Spearman correlations between cross-validated and full-sample scores.\u003c/p\u003e\n\u003ch3\u003eBehavioral Measures\u003c/h3\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eChild Behavior Checklist (CBCL)\u003c/h2\u003e \u003cp\u003eChildren\u0026rsquo;s emotional and behavioral symptoms were assessed using the Child Behavior Checklist (CBCL), a validated parent- or caregiver-report scale [\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e]. The CBCL includes eight syndrome scales: anxious/depressed, withdrawn/depressed, somatic complaints, social problems, thought problems, attention problems, rule-breaking behavior, and aggressive behavior. We used standardized T-scores of three composite subscales: internalizing problems (anxious/depressed, withdrawn/depressed, and somatic complaints), externalizing problems (rule-breaking and aggressive behaviors), and total problems (all eight syndrome scales combined), with higher scores indicating greater symptom severity.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eNIH Toolbox Cognition Battery\u003c/h3\u003e\n\u003cp\u003eCognitive function was evaluated using the NIH Toolbox Cognition Battery [\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e], which consists of seven performance-based tasks designed to measure key cognitive domains. Crystallized intelligence was assessed by summing performance on the Picture Vocabulary Test and the Oral Reading Recognition Test. Fluid intelligence was calculated using five tasks: the Flanker Inhibitory Control and Attention Test, the List Sorting Working Memory Test, the Dimensional Change Card Sort Test, the Pattern Comparison Processing Speed Test, and the Picture Sequence Memory Test. A total cognitive composite score was derived by summing the fluid and crystallized scores. For all cognition domains, standardized T-scores were used in the analyses, with higher values reflecting better cognitive performance. Crystallized intelligence data were obtained at both the baseline and 2-year follow-up.\u003c/p\u003e\n\u003ch3\u003eNeuroimaging Data Acquisition and Processing\u003c/h3\u003e\n\u003cp\u003eDiffusion MRI (dMRI) data were acquired using standardized protocols across multiple 3T scanner platforms at ABCD sites [\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. Individual-level diffusion MRI data were downloaded in FIB format from the ABCD Study Data Portal (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://brain.labsolver.org\u003c/span\u003e\u003cspan address=\"https://brain.labsolver.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), where the preprocessing pipeline is detailed (N\u0026thinsp;=\u0026thinsp;9,713 for baseline, N\u0026thinsp;=\u0026thinsp;7,197 for 2-year follow-up). Raw dMRI volumes were acquired using a multishell diffusion scheme with b-values of 500, 1000, 2000, and 3000 s/mm\u0026sup2; and corresponding diffusion directions of 6, 15, 15, and 60, respectively. Images were acquired at an in-plane resolution of 1.7 mm and slice thickness of 1.7 mm, and were rotated to align with the anterior commissure\u0026ndash;posterior commissure (AC\u0026ndash;PC) line, resulting in isotropic voxel resolution of 1.7 mm.\u003c/p\u003e \u003cp\u003eDiffusion data were reconstructed using generalized q-sampling imaging with a sampling length ratio of 1.25 [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e, \u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e]. Whole-brain tractography was performed in DSI Studio using deterministic streamline tracking with a modified FACT algorithm [\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e]. Tractography parameters included an angular cutoff of 60 degrees, step size of 1.0 mm, minimum streamline length of 30 mm, maximum length of 300 mm, and spin density function smoothing set to 0. QA values were used as the termination criterion, and fiber tracking continued until 10\u0026nbsp;million streamlines were reconstructed for each subject.\u003c/p\u003e \u003cp\u003eQA was computed in native space and used for nonlinear registration to MNI space via the SPM12 algorithm. Within MNI space, spin density functions were reconstructed using three fiber orientations per voxel, and a mean diffusion distance of 1.25 mm. For each participant, a 268 \u0026times; 268 structural connectivity matrix was generated by averaging the QA values of streamlines connecting pairs of regions from the 268-node Shen atlas [\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e]. Only connections with mean QA\u0026thinsp;\u0026gt;\u0026thinsp;0.001 were retained to minimize noise. These 268 parcellated regions were grouped into 10 canonical networks [\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e]: medial frontal (MF), frontoparietal (FP), default mode (DMN), motor (Mot), visual I (Vis I), visual II (Vis II), visual association (Vas), salience (SAL), subcortical (SC), and cerebellar (CBL) networks.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eNetwork Control Theory and Controllability Metrics\u003c/h2\u003e \u003cdiv id=\"Sec12\" class=\"Section3\"\u003e \u003ch2\u003eDynamical Model\u003c/h2\u003e \u003cp\u003eThe network control theory can simulate the consequences of a network\u0026rsquo;s underlying structural topology on its function dynamics [\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e]. The present study applied the network control theory framework to calculate brain controllability [\u003cspan additionalcitationids=\"CR32\" citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e], implemented in MATLAB with in-house packages and code, from structural connectomes. Based on the previous studies [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e], neural states can be mathematically described as simulated states (\u003cb\u003ex\u003c/b\u003e) of network with \u003cb\u003ek\u003c/b\u003e nodes overtime steps \u003cb\u003et\u003c/b\u003e using a simplified noise-free linear discrete-time and time-invariant model:\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\mathbf{x}\\left(\\varvec{t}+1\\right)=\\mathbf{A}\\mathbf{x}\\left(\\varvec{t}\\right)+{\\mathbf{B}}_{\\mathbf{k}}{\\mathbf{u}}_{\\mathbf{k}}\\left(\\varvec{t}\\right)$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003ewhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{x}\\left(\\varvec{t}\\right)\\)\u003c/span\u003e\u003c/span\u003e is a vector that describes the state of all brain regions at time \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:t\\)\u003c/span\u003e\u003c/span\u003e, while matrix \u003cb\u003eA\u003c/b\u003e is a structural connection matrix. \u003cb\u003eK\u003c/b\u003e is the set of nodes that can be controlled independently. The input matrix defines the control nodes. The \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\mathbf{u}}_{\\mathbf{K}}\\left(\\varvec{t}\\right)\\)\u003c/span\u003e\u003c/span\u003e is the control signal injected into the network via the control nodes, while \u003cb\u003eB\u003c/b\u003e is the input matrix.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eControllability Metrics\u003c/h2\u003e \u003cp\u003eSubsequently, the influence of each region on brain function is quantified using the metric of controllability. Controllability of a dynamical system refers to the possibility of driving the state of a dynamical system to a specific target state by means of an external control input. Classic results in the control theory ensure that controllability of the network from the set of network nodes \u003cem\u003ek\u003c/em\u003e is equivalent to the controllability Gramian \u003cem\u003eW\u003c/em\u003e\u003csub\u003eK\u003c/sub\u003e being invertible, where \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varvec{W}}_{\\varvec{k}}=\\:\\sum\\:_{\\varvec{\\tau\\:}=0}^{\\varvec{\\infty\\:}}{\\varvec{A}}^{\\varvec{\\pi\\:}}{\\varvec{B}}_{\\varvec{k}}{\\varvec{B}}_{\\varvec{k}}^{\\varvec{T}}{\\varvec{A}}^{\\varvec{\\pi\\:}}\\)\u003c/span\u003e\u003c/span\u003e. We utilize this framework to choose control nodes one at a time, and thus the input matrix \u003cb\u003eB\u003c/b\u003e in fact reduces to a one-dimensional vector.\u003c/p\u003e \u003cp\u003eBased on this network control theory framework, we examined modal controllability, which quantifies the ability of a brain region to drive the network into difficult-to-reach states that require substantial control energy [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Such transitions may correspond to cognitively demanding or less frequently engaged brain states [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e], which are particularly relevant in the context of neurodevelopment and psychopathology. Modal controllability is computed from the eigenvectors [\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varvec{V}}_{\\varvec{i}\\varvec{j}}\\)\u003c/span\u003e\u003c/span\u003e] of the adjacency matrix \u003cb\u003eA\u003c/b\u003e. Here, we define the measurement of modal controllability as \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varvec{\\Phi\\:}}_{\\varvec{i}}={\\sum\\:}_{\\varvec{j}=1}^{\\varvec{N}}(1-{\\varvec{\\lambda\\:}}_{\\varvec{j}}^{2}(\\varvec{A}\\left){\\varvec{V}}_{\\varvec{i}\\varvec{j}}^{2}\\right)\\)\u003c/span\u003e\u003c/span\u003e of all N modes\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\:\\:{\\varvec{\\lambda\\:}}_{0}\\left(\\varvec{A}\\right),...,\\:\\:{\\varvec{\\lambda\\:}}_{\\mathbf{N}-1}\\left(\\varvec{A}\\right)\\)\u003c/span\u003e\u003c/span\u003e from brain region \u003cb\u003ei\u003c/b\u003e, following the definition of previous studies.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003ePRSs Calculation\u003c/h2\u003e \u003cdiv id=\"Sec15\" class=\"Section3\"\u003e \u003ch2\u003eBase GWAS Datasets\u003c/h2\u003e \u003cp\u003ePRS\u003csub\u003eADHD\u003c/sub\u003e and PRS\u003csub\u003eASD\u003c/sub\u003e were derived for two childhood psychiatric conditions: attention deficit hyperactivity disorder (ADHD) [\u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e61\u003c/span\u003e], and autism spectrum disorder (ASD) [\u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e62\u003c/span\u003e], using GWAS summary statistics from cohorts of European ancestry. These disorders were selected based on their well-powered GWAS, relevance to early development, and broad clinical presentations. The base datasets included single nucleotide polymorphisms (SNPs) with reported effect alleles, odds ratios (ORs), P-values, and allele frequencies. Full details of the GWAS samples and cohort sizes are provided in \u003cb\u003eSupplementary Table\u0026nbsp;2\u003c/b\u003e.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eTarget Genotype Dataset and Quality Control\u003c/h2\u003e \u003cp\u003eDue to the reduced performance of cross-ancestry prediction [\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e], analyses were restricted to a subset of participants of confirmed European ancestry. Genotype data were obtained from the ABCD Study (Release 5.1) and processed using Axiom Analysis Suite (version 2.11) from raw intensities files generated by the Affymetrix Smokescreen array. Biological samples were derived from either saliva or whole blood samples. Participants were selected based on genotyping call rate, concordance between reported and genetic sex, and identity-by-state (IBS) estimates to minimize cryptic relatedness. Initial quality control (QC) was conducted by the ABCD consortium and included call rate thresholds, signal intensity checks, and marker-level filters, followed by the RICOPILI pipeline for imputation-ready QC.\u003c/p\u003e \u003cp\u003e Ancestry-specific quality control was then applied separately to participants of European ancestry and to the remaining trans-ancestry group. SNPs were filtered based on missingness rate (\u0026lt;\u0026thinsp;5%), Hardy-Weinberg equilibrium (HWE\u0026thinsp;\u0026lt;\u0026thinsp;1\u0026times;10⁻⁹), and excess heterozygosity (\u0026gt;\u0026thinsp;3 SD from the mean). Genotype phasing and imputation were performed using the TOPMed imputation server. Post-imputation SNPs were retained only if they had a minor allele frequency (MAF)\u0026thinsp;\u0026ge;\u0026thinsp;0.005 and imputation quality INFO score\u0026thinsp;\u0026ge;\u0026thinsp;0.9. A pruned set of SNPs (clumping threshold \u003cem\u003er\u0026sup2;\u003c/em\u003e = 0.1) was used to calculate the top 10 genetic principal components (PCs) using PLINK [\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e65\u003c/span\u003e]. Due to the reduced performance of cross-ancestry prediction [\u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e63\u003c/span\u003e, \u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e], all PRS analyses were restricted to a subset of participants of confirmed European ancestry.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003ePRS Computation\u003c/h2\u003e \u003cp\u003ePRS were computed using PRSice-2 (version 2.3.5) [\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e] with clumping and thresholding parameters set as clump-kb 250kb, clump-r\u0026sup2; 0.1, and clump-p 1.0. For each participant, PRS were calculated as the sum of risk alleles (0, 1, or 2) at each locus, weighted by the corresponding log-transformed OR (logOR) from the base GWAS. Resulting scores were standardized (z-scored) prior to inclusion in downstream analyses.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eCovariates\u003c/h2\u003e \u003cp\u003eTo reduce potential confounding, all analyses were adjusted for a consistent set of covariates, including age, sex, race, study site, handedness, pubertal status [\u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e], and the first 5 genetic principal components. Categorical variables (sex, race, handedness) were dummy-coded prior to inclusion in statistical models.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eStatistical Analysis\u003c/h2\u003e \u003cp\u003eAll continuous variables were standardized (Z-score) prior to analysis to account for differences in measurement scales. The Benjamini-Hochberg false discovery rate (FDR) correction was applied to adjust for multiple comparisons, and adjusted \u003cem\u003eq\u003c/em\u003e-values were reported [\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e].\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eAssociation Analysis at Baseline\u003c/h2\u003e \u003cp\u003eThe linear mixed-effects models were employed to examine the pairwise associations among ELA latent factors, controllability measures, and behavioral assessments at baseline. In each model, controllability or behavioral measure was used as the dependent variable, ELA factor and all covariates as fixed effects, and the study site as a random effect [\u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e]. The estimated regression coefficient (\u003cem\u003eβ\u003c/em\u003e), standard error and \u003cem\u003ep\u003c/em\u003e value were reported. Additionally, we estimated pairwise Pearson correlation between controllability and behavioral measures.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eStratified Analysis\u003c/h2\u003e \u003cp\u003eTo evaluate the robustness and generalizability of the observed association, sex-stratified and ancestry-stratified analyses were performed. Models were re-estimated separately for males and females, and again for participants of European versus non-European ancestry. To formally assess the similarity of results across subgroups, we calculated Pearson correlation between stratified effect sizes across all brain networks or behavioral outcomes. When low cross-group consistency was observed, additional interaction analyses between sex or ancestry group and each key independent variable were conducted in the full sample models. These interaction models allowed us to directly assess whether the strength or direction of associations differed significantly across sex or ancestry groups.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eMediation and Moderated Mediation Analyses\u003c/h2\u003e \u003cp\u003eWhen brain controllability was significantly associated with both ELA latent factors and behavioral measures, mediation analyses were conducted to test whether the controllability measure mediated the associations between ELA factors and behavioral measures. Indirect effects were estimated using a bias-corrected bootstrap procedure with 5000 iterations, and mediation was considered statistically significant if the 95% CI did not include zero.\u003c/p\u003e \u003cp\u003eFor each significant mediation pathway, moderated mediation analyses were then conducted in the European subsample to examine whether the indirect effect varied by genetic risks for ADHD and ASD. Participants were divided into low (15%), medium (70%), and high (15%) genetic risk subgroups based on the 16th, and 84th percentiles, consistent with the default settings in the \u003cem\u003ePROCESS macro\u003c/em\u003e for R [\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e]. Conditional indirect effects were estimated for each group using the same bootstrap approach.\u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003eLongitudinal Analysis\u003c/h2\u003e \u003cp\u003eA cross-lagged panel model (CLPM), implemented with the \u003cem\u003elavaan\u003c/em\u003e package in R [\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e], was applied to evaluate associations between baseline brain controllability measures and 2-year follow-up behavioral outcomes, and vice versa. These analyses were performed in participants with follow-up behavioral assessments (5,115 for CBCL and 3,999 for crystallized intelligence). Prior to model fitting, both brain controllability metrics and behavioral scores (CBCL scales and cognition) at baseline and 2-year follow-up were residualized by regressing out all covariates using linear models. The resulting residuals were then standardized (z-scored) to ensure comparability across variables. Cross-lagged associations were estimated between baseline and follow-up measures of brain and behavior. Each CLPM included autoregressive paths (baseline to follow-up) and cross-lagged paths (brain to behavior and vice versa), with correlated residuals specified at both timepoints. All models were estimated using the \u003cem\u003esem\u003c/em\u003e function from the \u003cem\u003elavaan\u003c/em\u003e R package with full-information maximum likelihood (ML) estimation and robust (Huber-White) standard errors. To control for multiple testing, false discovery rate (FDR) correction was applied across all \u003cem\u003ep\u003c/em\u003e-values derived from cross-lagged paths.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec25\" class=\"Section2\"\u003e \u003ch2\u003eDemographics\u003c/h2\u003e \u003cp\u003eA total of 7,970 participants were included (\u003cb\u003eSupplementary Fig.\u0026nbsp;1\u003c/b\u003e), comprising 4,141 males (52.0%) and 3,829 females (48.0%). The mean age of the participant was 9.47 years at baseline (standard deviation [SD]\u0026thinsp;=\u0026thinsp;0.51; range\u0026thinsp;=\u0026thinsp;9\u0026ndash;11 years). Participants were identified as Asian (1.7%), Black (15.0%), Hispanic (20.1%), White (53.0%), and Other Race (10.2%). Detailed demographic characteristics of the participants are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eDemographic Characteristics of Participants Included in the Study.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"2\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDemographic Characteristics\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo. (%/SD)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge, mean (SD), y\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e9.47 (0.51)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4141 (52.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemale\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e3 829\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePubertal Development Scale Score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e-0.008 (0.97)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHandedness\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRight-handed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e6365 (79.9)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLeft-handed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e560 (7.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMix-handed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1045 (13.1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRace\u003csup\u003ea\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAsian\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e136 (1.7)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBlack\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1193 (15.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHispanic\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e1602 (20.1)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWhite\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e4228 (53.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOther\u003csup\u003eb\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e811 (10.2)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eChild Behavior Checklist (CBCL)\u003csup\u003ec\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCBCL Internal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e48.6 (10.6)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCBCL External\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e45.7 (10.2)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCBCL Totalprob\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e46.0 (11.2)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNIH Toolbox Cognition\u003csup\u003ed\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNIH FluidComp\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e45.9 (10.8)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNIH CrystComp\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e51.4 (11.0)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNIH TotalComp\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e48.1 (10.8)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003ctfoot\u003e \u003ctr\u003e\u003ctd colspan=\"2\"\u003e\u003csup\u003ea\u003c/sup\u003e Race was reported by each participant\u0026rsquo;s caregiver.\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"2\"\u003e\u003csup\u003eb\u003c/sup\u003e Other includes American Indian or Alaska Native, Native Hawaiian, and Pacific Islander participants.\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"2\"\u003e\u003csup\u003ec\u003c/sup\u003e CBCL scores are derived from caregiver report using empirically based syndrome scales: internalizing (CBCL Internal), externalizing (CBCL External), and total problems (CBCL Totalprob).\u003c/td\u003e\u003c/tr\u003e \u003ctr\u003e\u003ctd colspan=\"2\"\u003e\u003csup\u003ed\u003c/sup\u003e Cognitive scores were obtained from the NIH Toolbox Cognition Battery, including fluid intelligence (NIH FluidComp), crystallized intelligence (NIH CrystComp), and total composite score (NIH TotalComp).\u003c/td\u003e\u003c/tr\u003e \u003c/tfoot\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cdiv id=\"Sec26\" class=\"Section3\"\u003e \u003ch2\u003eELA Dimensions\u003c/h2\u003e \u003cp\u003eTo derive latent dimensions of early life adversity (ELA), we first applied quality control to 139 candidate variables (\u003cb\u003eSupplementary Fig.\u0026nbsp;1 and\u003c/b\u003e Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), retaining 67 indicators for exploratory factor analysis (EFA) (\u003cb\u003eSupplementary Fig.\u0026nbsp;2\u003c/b\u003e). The data were suitable for factor analysis, as evidenced by Kaiser-Meyer-Olkin statistic\u0026thinsp;=\u0026thinsp;0.827 and Bartlett test of sphericity (χ\u0026sup2; = 266 144, \u003cem\u003ep\u003c/em\u003e \u0026lt; .001). To determine the optimal number of latent factors, we used Very Simple Structure (VSS), Minimum Average Partial (MAP), and model fit indices. VSS suggested four and five factors as optimal solutions under complexity levels 1 and 2 (VSS\u0026thinsp;=\u0026thinsp;0.553 and 0.614, respectively). Although the MAP statistic reached its global minimum at nine factors (MAP\u0026thinsp;=\u0026thinsp;0.0048), values were near-minimal at four (MAP\u0026thinsp;=\u0026thinsp;0.0058) and five (MAP\u0026thinsp;=\u0026thinsp;0.0054) factors, indicating that these solutions captured the most common variance. Moreover, root mean square error of approximation (RMSEA) decreased progressively and plateaued at five factors (RMSEA\u0026thinsp;=\u0026thinsp;0.027), supporting the selection of a five-factor model (\u003cb\u003eSupplementary Fig.\u0026nbsp;3\u003c/b\u003e). Confirmatory factor analysis (CFA) of the five-factor solution showed excellent model fit in 10-fold cross-validation (\u003cb\u003eSupplementary Table\u0026nbsp;3\u003c/b\u003e), with cross-validated factor scores highly correlated with full-sample scores (Spearman \u003cem\u003eρ\u003c/em\u003e range, 0.99-1.00), supporting the stability and robustness of the five-factor model.\u003c/p\u003e \u003cp\u003eThe five latent ELA dimensions represented distinct psychosocial contexts (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, \u003cb\u003eSupplementary Table\u0026nbsp;4\u003c/b\u003e). Factor 1 (F1) reflected paternal psychosocial adversity, including high loadings on father-related marital, occupational, legal, behavioral, and substance use problems. Factor 2 (F2) captured family conflict and hostile climate, characterized by frequent fights, criticism, physical aggression, and low conflict resolution. Factor 3 (F3) indicated low caregiver and emotional support, with high loadings on less frequent caregiver comfort, love expression, open communication, and emotional availability. Factor 4 (F4) represented caregiver psychopathology, with high loadings on caregiver depressive, anxiety, somatic, ADHD, antisocial, and avoidant symptoms. Factor 5 (F5) indexed socioeconomic disadvantages, with high loadings for family income (negative loading), parental education (negative loading), and neighborhood safety concerns (positive loading). Together, these dimensions offer a comprehensive, multidimensional characterization of ELA during early development.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec27\" class=\"Section3\"\u003e \u003ch2\u003eAssociations between ELA and Brain Controllability\u003c/h2\u003e \u003cp\u003eThree ELA dimensions were significantly associated with brain controllability (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, \u003cb\u003eSupplementary Table\u0026nbsp;5\u003c/b\u003e). We identified consistent positive correlations between F1 (paternal psychosocial adversity) and controllability in the medial frontal, motor, and salience networks (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.028\u0026ndash;0.036; \u003cem\u003eq\u003c/em\u003e \u0026lt; .05), F4 (caregiver psychopathology) and controllability in the salience, medial frontal, visual association, motor, frontoparietal, and cerebellar networks (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.030\u0026ndash;0.047; \u003cem\u003eq\u003c/em\u003e \u0026lt; .05), and F5 (socioeconomic disadvantage) and controllability in the cerebellar and salience networks (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.039\u0026ndash;0.057; \u003cem\u003eq\u003c/em\u003e \u0026lt; .05). These findings indicate that distinct ELA dimensions could consistently increase brain controllability, with both shared and unique effects. However, F2 and F3 were not significantly associated with controllability in any network.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec28\" class=\"Section2\"\u003e \u003ch2\u003eAssociations between ELA and Behavioral Outcomes\u003c/h2\u003e \u003cp\u003eAll five ELA factors were significantly associated with behavioral outcomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, \u003cb\u003eSupplementary Table\u0026nbsp;6\u003c/b\u003e). Higher ELA factor scores, especially F4 scores, were consistently associated with greater internalizing, externalizing, and total behavioral problems on the CBCL (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.091 to 0.564; \u003cem\u003eq\u003c/em\u003e \u0026lt; .05). All, but F4, were associated with lower cognition composite scores (\u003cem\u003eβ\u003c/em\u003e = -0.029 to -0.167; \u003cem\u003eq\u003c/em\u003e \u0026lt; .05), particularly for F5 (\u003cem\u003eβ\u003c/em\u003e = -0.120 to -0.167; \u003cem\u003eq\u003c/em\u003e \u0026lt; .05).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec29\" class=\"Section2\"\u003e \u003ch2\u003eAssociation between Controllability and Behavioral Outcomes\u003c/h2\u003e \u003cp\u003eHigher controllability in the medial frontal, frontoparietal, default mode, motor, and salience networks was significantly associated with greater behavioral problems (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.031 to 0.052; \u003cem\u003eq\u003c/em\u003e \u0026lt; .05, Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA\u003cb\u003e/B, Supplementary Table\u0026nbsp;7\u003c/b\u003e). In particular, Higher medial frontal and motor network controllability was consistently associated with higher externalizing (MF: \u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.044; Mot: \u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.028) and internalizing (MF: \u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.043; Mot: \u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.036) symptoms (all \u003cem\u003eq\u003c/em\u003e \u0026lt; .05). However, SAL network controllability was also correlated with externalizing symptoms (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.030; \u003cem\u003eq\u003c/em\u003e \u0026lt; .05), while FP and DMN network controllability was additionally associated with internalizing symptoms (FP: \u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.031; DMN: \u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.033; \u003cem\u003eq\u003c/em\u003e \u0026lt; .05).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eWorse cognitive performance was also linked to higher controllability across seven networks, including the medial frontal, frontoparietal, default mode, motor, salience, visual association, and subcortical networks (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC\u003cb\u003e/D, Supplementary Table\u0026nbsp;8\u003c/b\u003e). Specifically, higher controllability in the medial frontal, default mode, and motor networks was linked to lower crystallized cognition composite scores (\u003cem\u003eβ\u003c/em\u003e = -0.027 to -0.047; \u003cem\u003eq\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05).\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eStratified Analyses\u003c/h3\u003e\n\u003cp\u003eIn sex-stratified analyses, ELA-brain controllability and brain controllability-behavior associations were generally consistent across males and females (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.552 to 0.993, \u003cem\u003eq\u003c/em\u003e \u0026lt; .05) (\u003cb\u003eSupplementary Fig.\u0026nbsp;6\u003c/b\u003e). However, there were several potential sex-specific effects. For ELA, F4 and F5 showed stronger associations with controllability in the cerebellar network for females (\u003cb\u003eSupplementary Fig.\u0026nbsp;7\u003c/b\u003e). F5 showed stronger associations with CBCL total problems in males (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.282 vs 0.241 in females, \u003cb\u003eSupplementary Fig.\u0026nbsp;8\u003c/b\u003e). Controllability in the subcortical, frontoparietal, and medial frontal networks were more strongly associated with cognitive total composite score in females (\u003cb\u003eSupplementary Fig.\u0026nbsp;9\u003c/b\u003e).\u003c/p\u003e \u003cp\u003eIn contrast, ancestry-stratified analyses revealed divergence in specific domains (\u003cb\u003eSupplementary Fig.\u0026nbsp;10\u003c/b\u003e). Associations between ELA factors and brain controllability showed poor concordance across European and non-European participants (\u003cem\u003er\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.073, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.614), suggesting ancestry-specific effects. Among participants of European ancestry, F5 was positively associated with controllability in all ten networks (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.042\u0026ndash;0.088, \u003cem\u003eq\u003c/em\u003e \u0026lt; .05, \u003cb\u003eSupplementary Fig.\u0026nbsp;11\u003c/b\u003e), but these effects were attenuated or absent in non-European participants. By contrast, associations between ELA and behavioral outcomes, as well as brain controllability-behavior links, were broadly consistent across ancestry groups (\u003cb\u003eSupplementary Figs.\u0026nbsp;12\u0026ndash;13\u003c/b\u003e).\u003c/p\u003e \u003cp\u003eHowever, interaction analyses revealed significant moderation by ancestry in the associations between F5 and brain controllability (\u003cem\u003eq\u003c/em\u003e \u0026lt; .05, \u003cb\u003eSupplementary Fig.\u0026nbsp;14, Supplementary Table\u0026nbsp;9\u003c/b\u003e), particularly in the frontoparietal, default mode, and medial frontal networks. All interaction coefficients were positive, indicating that the association between F5 and brain controllability was stronger among participants of European ancestry. In other words, increases in socioeconomic disadvantage were more strongly related to alterations in controllability in European than in non‑European participants. These findings underscore the importance of considering population-specific contexts in developmental neuroimaging studies.\u003c/p\u003e \u003cdiv id=\"Sec31\" class=\"Section2\"\u003e \u003ch2\u003eMediation and Moderated Mediation Analyses\u003c/h2\u003e \u003cp\u003eWe tested a fully moderated mediation framework to examine whether brain controllability mediated the effects of ELA on cognitive and behavioral outcomes, and whether these indirect effects varied by genetic risk for neurodevelopmental disorders (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eMediation analyses revealed that brain controllability in specific brain networks partially mediated the associations between ELA factors and behavioral problems and cognitive performance (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB and \u003cb\u003eSupplementary Table\u0026nbsp;10\u003c/b\u003e). For instance, the medial frontal network controllability partially mediated the associations of F4 (caregiver psychopathology) and NIH cognitive scores, including the total composite (proportion mediated\u0026thinsp;=\u0026thinsp;3.1%, indirect \u003cem\u003eβ\u003c/em\u003e = -0.002, 95% CI = -0.004 to -0.001), and crystallized composite (proportion mediated\u0026thinsp;=\u0026thinsp;3.8%, indirect \u003cem\u003eβ\u003c/em\u003e = -0.002, 95% CI = -0.004 to -0.001). Medial frontal controllability also mediated the link between F4 and CBCL total problems (proportion mediated\u0026thinsp;=\u0026thinsp;0.6%, indirect \u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.001, 95% CI\u0026thinsp;=\u0026thinsp;0.0002 to 0.002). Similarly, the frontoparietal network controllability partially mediated the association between F4 and total cognition score (proportion mediated\u0026thinsp;=\u0026thinsp;4.8%, indirect \u003cem\u003eβ\u003c/em\u003e = -0.001, 95% CI = -0.002 to -0.0003).\u003c/p\u003e \u003cp\u003eTo evaluate moderation by genetic risk, we stratified individuals based on PRS\u003csub\u003eADHD\u003c/sub\u003e or PRS\u003csub\u003eASD\u003c/sub\u003e. In individuals with higher PRS\u003csub\u003eADHD\u003c/sub\u003e or PRS\u003csub\u003eASD\u003c/sub\u003e, the indirect effects of F5 (socioeconomic disadvantage) on crystallized cognition via medial frontal network controllability were more pronounced compared to those with medium PRSs individuals. For example, among those with high PRS\u003csub\u003eASD\u003c/sub\u003e, the moderated proportion mediated was 7.5% (indirect \u003cem\u003eβ\u003c/em\u003e = -0.012, 95% CI = -0.021 to -0.0006), compared to 3.7% in the medium PRS\u003csub\u003eASD\u003c/sub\u003e group (indirect \u003cem\u003eβ\u003c/em\u003e = -0.006, 95% CI = -0.011 to -0.0003). Similarly, in high PRS\u003csub\u003eADHD\u003c/sub\u003e individuals, the moderated proportion mediated was 3.9% (indirect \u003cem\u003eβ\u003c/em\u003e = -0.008, 95% CI = -0.017 to -0.0002) versus 3.6% in the medium PRS\u003csub\u003eADHD\u003c/sub\u003e group (indirect \u003cem\u003eβ\u003c/em\u003e = -0.006, 95% CI = -0.011 to -0.0002). These results suggest that genetic vulnerability amplifies the impact of ELA on cognition through altered brain controllability, potentially in a dose-dependent relationship. Additional significant statistically moderated mediations are reported in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC\u003cb\u003e/D\u003c/b\u003e, \u003cb\u003eSupplementary Table\u0026nbsp;11\u003c/b\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec32\" class=\"Section2\"\u003e \u003ch2\u003eLongitudinal Associations Between Controllability and Behavioral Outcomes\u003c/h2\u003e \u003cp\u003eWe performed cross-lagged panel modeling (CLPM) to examine bidirectional relationships between brain controllability and behavioral outcomes over time. Analyses included 5,115 participants with 2-year follow-up data for behavioral problem and in 3,999 for crystallized cognition. All combinations of 10 brain networks and 4 outcomes (3 behavioral and 1 cognitive measures) were tested, but only a subset demonstrated significant longitudinal associations.\u003c/p\u003e \u003cp\u003eAutoregressive paths revealed strong within-domain stability over the 2-year interval, with all brain controllability and behavioral/cognitive variables showing significant self-prediction (e.g., medial frontal controllability: \u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.531, \u003cem\u003eq\u003c/em\u003e \u0026lt; .001; crystallized cognition: \u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.665, \u003cem\u003eq\u003c/em\u003e \u0026lt; .001; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE, F; \u003cb\u003eSupplementary Tables\u0026nbsp;12, 13\u003c/b\u003e).\u003c/p\u003e \u003cp\u003eCross-domain associations were selective and often asymmetric (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE, F; \u003cb\u003eSupplementary Tables\u0026nbsp;12, 13\u003c/b\u003e). For example, higher medial frontal controllability at baseline predicted lower crystallized cognition at follow-up (\u003cem\u003eβ\u003c/em\u003e = -0.037; \u003cem\u003eq\u003c/em\u003e = .003), and lower baseline cognition also predicted increased medial frontal controllability at follow-up (\u003cem\u003eβ\u003c/em\u003e = -0.033; \u003cem\u003eq\u003c/em\u003e = .020), suggesting a bidirectional relationship between control dynamics and cognition development (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE). In contrast, total behavioral problems at baseline predicted increased motor network controllability two years later (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.028; \u003cem\u003eq\u003c/em\u003e = .041), but not vice versa (\u003cem\u003eβ\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.013; \u003cem\u003eq\u003c/em\u003e = .296), indicating a unidirectional influence of behavior on brain development (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eF).\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this large, demographically diverse pediatric sample, we identified associations between specific ELA dimensions to distinct brain networks and, in turn, to cognitive and behavioral outcomes. Adversity related to caregiver psychopathology and paternal psychosocial disruption displayed the strongest associations with brain controllability in the medial frontal, frontoparietal, salience, and motor networks. These same networks were also the strongest mediators linking ELA to overall behavioral problems and crystallized cognition, suggesting that behavioral regulation and cognitive control circuits may be particularly sensitive to early psychosocial stress. Notably, the effects of ELA on behavior and cognition via controllability were moderated by polygenic risk for ADHD and ASD, highlighting a gene-environment interaction in shaping neurodevelopmental trajectories. Taken together, these findings provide compelling evidence for a structured neurodevelopmental cascade, from adversity to altered brain connectivity to behavioral impairment, refined further by inherited genetic risk. This work provides a systems-level framework for delineating how early adversity shapes neurodevelopmental pathways, offering new directions for predictive models and precision-oriented early interventions integrating neurobiology and genomics.\u003c/p\u003e \u003cp\u003eHistorically, it has been difficult to disentangle the many early life experiences that drive neurodevelopment [\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e, \u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e]. A key strength of this study lies in the identification of five distinct and interpretable latent factors of ELA. This multidimensional approach incorporates the co-occurrence of ELA and supports growing evidence that ELA is not a unitary construct [\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e]. These results emphasize the importance of parsing the heterogeneity in ELA to characterize its downstream impact on the brain and behavior. For example, these latent factors exhibited unique associations with controllability and later behavior, which may have been occluded if ELA was treated as a unitary construct.\u003c/p\u003e \u003cp\u003eAlthough no prior studies have examined controllability in the context of ELA, our findings are broadly consistent with previous reports linking parental psychosocial adversity, psychopathology, and socioeconomic disadvantages to altered white matter integrity and functional connectivity in key regulatory networks, including the salience, medial frontal, frontoparietal, and cerebellar networks [\u003cspan additionalcitationids=\"CR75 CR76 CR77\" citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e]. We found that paternal psychosocial adversity (F1) and caregiver psychopathology (F4) were associated with the controllability in the salience, frontoparietal, visual, and motor networks. These networks are involved in affect regulation and executive function [\u003cspan additionalcitationids=\"CR80 CR81\" citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e]. These brain network alterations may reflect heightened allostatic demands and adaptive neurodevelopmental tuning in response to chronic caregiver-related stress. [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e] Interestingly, socioeconomic disadvantage and neighborhood insecurity (F5) were associated with controllability in the salience and cerebellar networks, suggesting partially distinct neural pathways for deprivation-based versus relational adversity. These findings are consistent with prior work demonstrating that different types of ELA, such as threat versus deprivation, may exert dissociable effects on brain organization and function [\u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e83\u003c/span\u003e, \u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e]. The recurrent involvement of the salience network across multiple ELA dimensions in our findings underscores its potential role as a common transdiagnostic hub affected by early adversity. The well-established functions of salience network in detecting salient internal and external stimuli, mediating switching between major brain networks, and coordinating allostatic responses may account for its vulnerability to early environmental stressors [\u003cspan citationid=\"CR85\" class=\"CitationRef\"\u003e85\u003c/span\u003e]. From a neurodevelopmental perspective, repeated stress exposure may recalibrate the controllability of salience network, potentially enhancing their influence over dynamic brain states in an effort to maintain functional homeostasis. This hypothesis is supported by emerging evidence that early life adversity may accelerate the maturation or functional coupling of salience-related circuits [\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e, \u003cspan additionalcitationids=\"CR87 CR88\" citationid=\"CR86\" class=\"CitationRef\"\u003e86\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR89\" class=\"CitationRef\"\u003e89\u003c/span\u003e]. Consistent with this view, a study reported that young adults with a history of childhood trauma exhibited increased network switching rates, reflecting reduced temporal stability of functional connectivity, especially within the frontoparietal and cingulo-opercular networks [\u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e90\u003c/span\u003e]. These results suggest that early adversity may lead to more dysregulated network dynamics, potentially increasing the allostatic load required to maintain cognitive and emotional homeostasis.\u003c/p\u003e \u003cp\u003eNotably, the ancestry‑specific moderation was observed for the association between the socioeconomic adversity (F5) and brain controllability. Participants of European ancestry showed steeper positive slopes between F5 and controllability in the medial frontal, frontoparietal, default mode, and salience networks. One possible interpretation is that, in relatively resource-advantaged contexts, socioeconomic adversity may represent a greater deviation from normative expectations, thereby exerting a stronger physiological and neurodevelopmental impact. This interpretation aligns with prior research suggesting that the magnitude of brain-environment associations depend not only on the severity of exposure but also on the distributional range and contextual meaning of socioeconomic risk [\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIn several brain networks, higher controllability associated with greater behavioral problems and worse cognitive abilities. Controllability reflects the theoretical ability of a network to drive the brain into difficult-to-reach states [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Thus, higher controllability may indicate greater neural instability or inefficient regulation, especially under higher allostatic loads. Further, controllability typically increases in adolescents [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. Greater controllability in the context of ELA may also reflect accelerated development [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e91\u003c/span\u003e], which is commonly observed in response to ELAs. These associations emphasize that higher controllability should not be interpreted as uniformly advantageous or disadvantageous. Together, the findings highlight the importance of contextualizing controllability and other neuroimaging measures within developmental histories when evaluating their functional significance.\u003c/p\u003e \u003cp\u003eMediation analyses revealed that brain controllability served as a partial mechanism linking ELA to both cognitive deficits and behavioral problems. These indirect effects were most prominent through the medial frontal and frontoparietal networks, regions central to cognitive control and affective processing [\u003cspan additionalcitationids=\"CR93\" citationid=\"CR92\" class=\"CitationRef\"\u003e92\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR94\" class=\"CitationRef\"\u003e94\u003c/span\u003e]. Conceptually, these results are consistent with a prior network control theory study showing that lower control energy needed to engage frontoparietal networks associates with better executive performance and partly mediates age-related gains in cognition [\u003cspan citationid=\"CR95\" class=\"CitationRef\"\u003e95\u003c/span\u003e]. Further, our results extend previous work by demonstrating that controllability can also act as a partial mediator of adversity-related outcomes. Notably, these mediation effects were further moderated by polygenic risk for ADHD and ASD. The inclusion of these mediation and moderated mediation models is a strength and nuances the understanding of individual variability in neurodevelopment. Overall, they support a \"double-hit\" model, where ELA and genetic vulnerability jointly influence neurodevelopment through alterations in brain controllability [\u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e96\u003c/span\u003e, \u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e97\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe longitudinal findings indicate that bidirectional associations between brain controllability and behavior exist, indicating that cognitive abilities and controllability mutually shape each other during adolescence. This dynamic aligns with the sensorimotor-association axis of cortical maturation, wherein higher order association regions involved in cognitive functions undergo prolonged development and plasticity [\u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e98\u003c/span\u003e]. In contrast, the unidirectional effect of early behavioral problems on later motor network controllability indicates that dysregulation may shape the motor network development. This finding is partially supported by structural imaging studies showing that persistent externalizing behaviors predict altered subcortical maturation and disrupted prefrontal-limbic coupling [\u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e99\u003c/span\u003e]. However, cross-lagged panel models account for temporal stability using autoregressive paths and do not distinguish between-person differences and within-person fluctuations. These models reflect group averages, rather than casual individual differences [\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThis study has several limitations. First, the ELA factors were based on caregiver and child reports of past experiences, which could be influenced by recall or reporting bias. Second, PRS analyses were limited to participants of European ancestry; findings may not extend to more diverse populations. Finally, our models examined only linear associations and did not account for possible nonlinear effects.\u003c/p\u003e \u003cp\u003eUsing data from the ABCD study, we revealed that ELA dimension-specific influences on brain controllability, which in turn affected behavioral outcomes in children. Genetic vulnerability further moderated these relationships, underscoring the interplay between genes and environment in shaping neurodevelopment. Ultimately, identifying individual-level brain markers (e.g., controllability) could help improve early risk detection and guide personalized intervention strategies.\u003c/p\u003e"},{"header":"Declarations","content":"\u003ch2\u003eAcknowledgements\u003c/h2\u003e \u003cp\u003eThis study was supported by funding from the Tianjin Natural Science Foundation (24ZXGQSY00040) and Tianjin Key Medical Discipline Construction Project (TJYXZDXK-3-008C) for Huaigui Liu. The ABCD study is supported by the National Institutes of Health and additional federal partners under award numbers U01DA041048, U01DA050989, U01DA051016, U01DA041022, U01DA051018, U01DA051037, U01DA050987, U01DA041174, U01DA041106, U01DA041117, U01DA041028, U01DA041134, U01DA050988, U01DA051039, U01DA041156, U01DA041025, U01DA041120, U01DA051038, U01DA041148, U01DA041093, U01DA041089, U24DA041123, and U24DA041147. A full list of federal partners is available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://abcdstudy.org/federal-partners.html\u003c/span\u003e\u003cspan address=\"https://abcdstudy.org/federal-partners.html\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cb\u003eConflict of Interest\u003c/b\u003e: None reported.\u003c/p\u003e\u003ch2\u003eData availability\u003c/h2\u003e \u003cp\u003eData used in the preparation of this article were obtained from the Adolescent Brain Cognitive Development Study\u0026reg; (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://abcdstudy\u003c/span\u003e\u003cspan address=\"https://abcdstudy\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. org), held in the NIMH Data Archive (NDA). Only researchers with an approved NDA Data Use Certification (DUC) may obtain ABCD Study data.\u003c/p\u003e\u003ch2\u003eCode availability\u003c/h2\u003e \u003cp\u003eWe made use of publicly available software and tools. The publicly available tools used in our analyses are described in Methods. The core codes are available at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/haigui0913/ELA_Controllability_\u003c/span\u003e\u003cspan address=\"https://github.com/haigui0913/ELA_Controllability_\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e Behavior_ABCD.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eKim B, Royle M. Annual Research Review: Mapping the multifaceted approaches and impacts of adverse childhood experiences - an umbrella review of meta-analyses. \u003cem\u003eJ Child Psychol Psychiatry\u003c/em\u003e 2025; 66(4): 399\u0026ndash;416.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWade M, Wright L, Finegold KE. The effects of early life adversity on children's mental health and cognitive functioning. \u003cem\u003eTransl Psychiatry\u003c/em\u003e 2022; 12(1): 244.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuby JL, Barch D, Whalen D, Tillman R, Belden A. Association Between Early Life Adversity and Risk for Poor Emotional and Physical Health in Adolescence: A Putative Mechanistic Neurodevelopmental Pathway. \u003cem\u003eJAMA Pediatr\u003c/em\u003e 2017; 171(12): 1168\u0026ndash;1175.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHughes K, Bellis MA, Hardcastle KA, Sethi D, Butchart A, Mikton C \u003cem\u003eet al.\u003c/em\u003e The effect of multiple adverse childhood experiences on health: a systematic review and meta-analysis. \u003cem\u003eLancet Public Health\u003c/em\u003e 2017; 2(8): e356-e366.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarr CP, Martins CM, Stingel AM, Lemgruber VB, Juruena MF. The role of early life stress in adult psychiatric disorders: a systematic review according to childhood trauma subtypes. \u003cem\u003eJ Nerv Ment Dis\u003c/em\u003e 2013; 201(12): 1007\u0026ndash;1020.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKessler RC, Davis CG, Kendler KS. Childhood adversity and adult psychiatric disorder in the US National Comorbidity Survey. \u003cem\u003ePsychol Med\u003c/em\u003e 1997; 27(5): 1101\u0026ndash;1119.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBrieant A, Vannucci A, Nakua H, Harris J, Lovell J, Brundavanam D \u003cem\u003eet al.\u003c/em\u003e Characterizing the dimensional structure of early-life adversity in the Adolescent Brain Cognitive Development (ABCD) Study. \u003cem\u003eDev Cogn Neurosci\u003c/em\u003e 2023; 61: 101256.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcLaughlin KA, Sheridan MA, Humphreys KL, Belsky J, Ellis BJ. The Value of Dimensional Models of Early Experience: Thinking Clearly About Concepts and Categories. \u003cem\u003ePerspect Psychol Sci\u003c/em\u003e 2021; 16(6): 1463\u0026ndash;1472.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEllis BJ, Sheridan MA, Belsky J, McLaughlin KA. Why and how does early adversity influence development? Toward an integrated model of dimensions of environmental experience. \u003cem\u003eDev Psychopathol\u003c/em\u003e 2022; 34(2): 447\u0026ndash;471.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchafer JL, McLaughlin KA, Manfro GG, Pan P, Rohde LA, Miguel EC \u003cem\u003eet al.\u003c/em\u003e Threat and deprivation are associated with distinct aspects of cognition, emotional processing, and psychopathology in children and adolescents. \u003cem\u003eDev Sci\u003c/em\u003e 2023; 26(1): e13267.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeanah CH, Egger HL, Smyke AT, Nelson CA, Fox NA, Marshall PJ \u003cem\u003eet al.\u003c/em\u003e Institutional rearing and psychiatric disorders in Romanian preschool children. \u003cem\u003eAm J Psychiatry\u003c/em\u003e 2009; 166(7): 777\u0026ndash;785.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDeater-Deckard K, Mullineaux PY, Beekman C, Petrill SA, Schatschneider C, Thompson LA. Conduct problems, IQ, and household chaos: a longitudinal multi-informant study. \u003cem\u003eJ Child Psychol Psychiatry\u003c/em\u003e 2009; 50(10): 1301\u0026ndash;1308.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTomassi S, Tosato S, Mondelli V, Faravelli C, Lasalvia A, Fioravanti G \u003cem\u003eet al.\u003c/em\u003e Influence of childhood trauma on diagnosis and substance use in first-episode psychosis. \u003cem\u003eBr J Psychiatry\u003c/em\u003e 2017; 211(3): 151\u0026ndash;156.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMay A. Experience-dependent structural plasticity in the adult human brain. \u003cem\u003eTrends Cogn Sci\u003c/em\u003e 2011; 15(10): 475\u0026ndash;482.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNelson CA, Sullivan EF, Valdes V. Early adversity alters brain architecture and increases susceptibility to mental health disorders. \u003cem\u003eNat Rev Neurosci\u003c/em\u003e 2025; 26(10): 642\u0026ndash;656.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVaidya N, Marquand AF, Nees F, Siehl S, Schumann G. The impact of psychosocial adversity on brain and behaviour: an overview of existing knowledge and directions for future research. \u003cem\u003eMol Psychiatry\u003c/em\u003e 2024; 29(10): 3245\u0026ndash;3267.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSheridan MA, Mukerji CE, Wade M, Humphreys KL, Garrisi K, Goel S \u003cem\u003eet al.\u003c/em\u003e Early deprivation alters structural brain development from middle childhood to adolescence. \u003cem\u003eSci Adv\u003c/em\u003e 2022; 8(40): eabn4316.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSheridan MA, Fox NA, Zeanah CH, McLaughlin KA, Nelson CA, 3rd. Variation in neural development as a result of exposure to institutionalization early in childhood. \u003cem\u003eProc Natl Acad Sci U S A\u003c/em\u003e 2012; 109(32): 12927\u0026ndash;12932.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMehta MA, Golembo NI, Nosarti C, Colvert E, Mota A, Williams SC \u003cem\u003eet al.\u003c/em\u003e Amygdala, hippocampal and corpus callosum size following severe early institutional deprivation: the English and Romanian Adoptees study pilot. \u003cem\u003eJ Child Psychol Psychiatry\u003c/em\u003e 2009; 50(8): 943\u0026ndash;951.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGee DG, Gabard-Durnam LJ, Flannery J, Goff B, Humphreys KL, Telzer EH \u003cem\u003eet al.\u003c/em\u003e Early developmental emergence of human amygdala-prefrontal connectivity after maternal deprivation. \u003cem\u003eProc Natl Acad Sci U S A\u003c/em\u003e 2013; 110(39): 15638\u0026ndash;15643.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLim L, Radua J, Rubia K. Gray matter abnormalities in childhood maltreatment: a voxel-wise meta-analysis. \u003cem\u003eAm J Psychiatry\u003c/em\u003e 2014; 171(8): 854\u0026ndash;863.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBusso DS, McLaughlin KA, Brueck S, Peverill M, Gold AL, Sheridan MA. Child Abuse, Neural Structure, and Adolescent Psychopathology: A Longitudinal Study. \u003cem\u003eJ Am Acad Child Adolesc Psychiatry\u003c/em\u003e 2017; 56(4): 321\u0026ndash;328 e321.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGold AL, Sheridan MA, Peverill M, Busso DS, Lambert HK, Alves S \u003cem\u003eet al.\u003c/em\u003e Childhood abuse and reduced cortical thickness in brain regions involved in emotional processing. \u003cem\u003eJ Child Psychol Psychiatry\u003c/em\u003e 2016; 57(10): 1154\u0026ndash;1164.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGerin MI, Viding E, Herringa RJ, Russell JD, McCrory EJ. A systematic review of childhood maltreatment and resting state functional connectivity. \u003cem\u003eDev Cogn Neurosci\u003c/em\u003e 2023; 64: 101322.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBick J, Zhu T, Stamoulis C, Fox NA, Zeanah C, Nelson CA. Effect of early institutionalization and foster care on long-term white matter development: a randomized clinical trial. \u003cem\u003eJAMA Pediatr\u003c/em\u003e 2015; 169(3): 211\u0026ndash;219.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHanson JL, Adluru N, Chung MK, Alexander AL, Davidson RJ, Pollak SD. Early neglect is associated with alterations in white matter integrity and cognitive functioning. \u003cem\u003eChild Dev\u003c/em\u003e 2013; 84(5): 1566\u0026ndash;1578.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDennison MJ, Rosen ML, Sambrook KA, Jenness JL, Sheridan MA, McLaughlin KA. Differential Associations of Distinct Forms of Childhood Adversity With Neurobehavioral Measures of Reward Processing: A Developmental Pathway to Depression. \u003cem\u003eChild Dev\u003c/em\u003e 2019; 90(1): e96-e113.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDufford AJ, Kim P. Family Income, Cumulative Risk Exposure, and White Matter Structure in Middle Childhood. \u003cem\u003eFront Hum Neurosci\u003c/em\u003e 2017; 11: 547.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTendolkar I, Martensson J, Kuhn S, Klumpers F, Fernandez G. Physical neglect during childhood alters white matter connectivity in healthy young males. \u003cem\u003eHum Brain Mapp\u003c/em\u003e 2018; 39(3): 1283\u0026ndash;1290.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChoi J, Jeong B, Polcari A, Rohan ML, Teicher MH. Reduced fractional anisotropy in the visual limbic pathway of young adults witnessing domestic violence in childhood. \u003cem\u003eNeuroimage\u003c/em\u003e 2012; 59(2): 1071\u0026ndash;1079.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGu S, Pasqualetti F, Cieslak M, Telesford QK, Yu AB, Kahn AE \u003cem\u003eet al.\u003c/em\u003e Controllability of structural brain networks. \u003cem\u003eNat Commun\u003c/em\u003e 2015; 6: 8414.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTang E, Giusti C, Baum GL, Gu S, Pollock E, Kahn AE \u003cem\u003eet al.\u003c/em\u003e Developmental increases in white matter network controllability support a growing diversity of brain dynamics. \u003cem\u003eNat Commun\u003c/em\u003e 2017; 8(1): 1252.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSun H, Jiang R, Dai W, Dufford AJ, Noble S, Spann MN \u003cem\u003eet al.\u003c/em\u003e Network controllability of structural connectomes in the neonatal brain. \u003cem\u003eNat Commun\u003c/em\u003e 2023; 14(1): 5820.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJohnson D, Policelli J, Li M, Dharamsi A, Hu Q, Sheridan MA \u003cem\u003eet al.\u003c/em\u003e Associations of Early-Life Threat and Deprivation With Executive Functioning in Childhood and Adolescence: A Systematic Review and Meta-analysis. \u003cem\u003eJAMA Pediatr\u003c/em\u003e 2021; 175(11): e212511.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePerevoznikova A, Berger P, Grosse Wiesmann C, White LO, Keil J. Maltreatment and executive functioning in childhood and adolescence - A multilevel meta-analysis. \u003cem\u003eChild Abuse Negl\u003c/em\u003e 2025; 169(Pt 2): 107663.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHahn T, Winter NR, Ernsting J, Gruber M, Mauritz MJ, Fisch L \u003cem\u003eet al.\u003c/em\u003e Genetic, individual, and familial risk correlates of brain network controllability in major depressive disorder. \u003cem\u003eMol Psychiatry\u003c/em\u003e 2023; 28(3): 1057\u0026ndash;1063.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi Q, Yao L, You W, Liu J, Deng S, Li B \u003cem\u003eet al.\u003c/em\u003e Controllability of Functional Brain Networks and Its Clinical Significance in First-Episode Schizophrenia. \u003cem\u003eSchizophr Bull\u003c/em\u003e 2023; 49(3): 659\u0026ndash;668.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTian Y, Shi W, Tao Q, Yang H, Guo H, Wen B \u003cem\u003eet al.\u003c/em\u003e Altered controllability of functional brain networks in obsessive-compulsive disorder. \u003cem\u003eJ Psychiatr Res\u003c/em\u003e 2025; 182: 522\u0026ndash;529.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCaspi A, Moffitt TE. Gene-environment interactions in psychiatry: joining forces with neuroscience. \u003cem\u003eNat Rev Neurosci\u003c/em\u003e 2006; 7(7): 583\u0026ndash;590.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBelsky J, Jonassaint C, Pluess M, Stanton M, Brummett B, Williams R. Vulnerability genes or plasticity genes? \u003cem\u003eMol Psychiatry\u003c/em\u003e 2009; 14(8): 746\u0026ndash;754.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSullivan PF, Daly MJ, O'Donovan M. Genetic architectures of psychiatric disorders: the emerging picture and its implications. \u003cem\u003eNat Rev Genet\u003c/em\u003e 2012; 13(8): 537\u0026ndash;551.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLewis CM, Vassos E. Polygenic risk scores: from research tools to clinical instruments. \u003cem\u003eGenome Med\u003c/em\u003e 2020; 12(1): 44.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRanlund S, Rosa MJ, de Jong S, Cole JH, Kyriakopoulos M, Fu CHY \u003cem\u003eet al.\u003c/em\u003e Associations between polygenic risk scores for four psychiatric illnesses and brain structure using multivariate pattern recognition. \u003cem\u003eNeuroimage Clin\u003c/em\u003e 2018; 20: 1026\u0026ndash;1036.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGriffiths KR, Braund TA, Kohn MR, Clarke S, Williams LM, Korgaonkar MS. Structural brain network topology underpinning ADHD and response to methylphenidate treatment. \u003cem\u003eTransl Psychiatry\u003c/em\u003e 2021; 11(1): 150.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAdisetiyo V. Revealing the Structural Neural Circuitry of Attention Deficit Hyperactivity Disorder With Diffusion MRI: Implications for Future Diagnosis and Treatment. \u003cem\u003eAJR Am J Roentgenol\u003c/em\u003e 2018; 210(4): 731\u0026ndash;733.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi SJ, Wang Y, Qian L, Liu G, Liu SF, Zou LP \u003cem\u003eet al.\u003c/em\u003e Alterations of White Matter Connectivity in Preschool Children with Autism Spectrum Disorder. \u003cem\u003eRadiology\u003c/em\u003e 2018; 288(1): 209\u0026ndash;217.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRudie JD, Brown JA, Beck-Pancer D, Hernandez LM, Dennis EL, Thompson PM \u003cem\u003eet al.\u003c/em\u003e Altered functional and structural brain network organization in autism. \u003cem\u003eNeuroimage Clin\u003c/em\u003e 2012; 2: 79\u0026ndash;94.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHagler DJ, Jr., Hatton S, Cornejo MD, Makowski C, Fair DA, Dick AS \u003cem\u003eet al.\u003c/em\u003e Image processing and analysis methods for the Adolescent Brain Cognitive Development Study. \u003cem\u003eNeuroimage\u003c/em\u003e 2019; 202: 116091.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBarch DM, Albaugh MD, Baskin-Sommers A, Bryant BE, Clark DB, Dick AS \u003cem\u003eet al.\u003c/em\u003e Demographic and mental health assessments in the adolescent brain and cognitive development study: Updates and age-related trajectories. \u003cem\u003eDev Cogn Neurosci\u003c/em\u003e 2021; 52: 101031.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHoffman EA, Clark DB, Orendain N, Hudziak J, Squeglia LM, Dowling GJ. Stress exposures, neurodevelopment and health measures in the ABCD study. \u003cem\u003eNeurobiol Stress\u003c/em\u003e 2019; 10: 100157.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKuhn M, Wing J, Weston S, Williams A, Keefer C, Engelhardt A \u003cem\u003eet al.\u003c/em\u003e Package \u0026lsquo;caret\u0026rsquo;. 2020; 223(7): 48.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu J, Liu N, Polemiti E, Garcia-Mondragon L, Tang J, Liu X \u003cem\u003eet al.\u003c/em\u003e Effects of urban living environments on mental health in adults. \u003cem\u003eNat Med\u003c/em\u003e 2023; 29(6): 1456\u0026ndash;1467.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Buuren S, Groothuis-Oudshoorn KJJoss. mice: Multivariate imputation by chained equations in R. 2011; 45: 1\u0026ndash;67.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLuciana M, Bjork JM, Nagel BJ, Barch DM, Gonzalez R, Nixon SJ \u003cem\u003eet al.\u003c/em\u003e Adolescent neurocognitive development and impacts of substance use: Overview of the adolescent brain cognitive development (ABCD) baseline neurocognition battery. \u003cem\u003eDev Cogn Neurosci\u003c/em\u003e 2018; 32: 67\u0026ndash;79.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGershon RC, Wagster MV, Hendrie HC, Fox NA, Cook KF, Nowinski CJ. NIH toolbox for assessment of neurological and behavioral function. \u003cem\u003eNeurology\u003c/em\u003e 2013; 80(11 Suppl 3): S2-6.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYeh FC, Liu L, Hitchens TK, Wu YL. Mapping immune cell infiltration using restricted diffusion MRI. \u003cem\u003eMagn Reson Med\u003c/em\u003e 2017; 77(2): 603\u0026ndash;612.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYeh FC, Wedeen VJ, Tseng WY. Generalized q-sampling imaging. \u003cem\u003eIEEE Trans Med Imaging\u003c/em\u003e 2010; 29(9): 1626\u0026ndash;1635.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eShen X, Tokoglu F, Papademetris X, Constable RT. Groupwise whole-brain parcellation from resting-state fMRI data for network node identification. \u003cem\u003eNeuroimage\u003c/em\u003e 2013; 82: 403\u0026ndash;415.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNoble S, Spann MN, Tokoglu F, Shen X, Constable RT, Scheinost D. Influences on the Test-Retest Reliability of Functional Connectivity MRI and its Relationship with Behavioral Utility. \u003cem\u003eCereb Cortex\u003c/em\u003e 2017; 27(11): 5415\u0026ndash;5429.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYan G, Vertes PE, Towlson EK, Chew YL, Walker DS, Schafer WR \u003cem\u003eet al.\u003c/em\u003e Network control principles predict neuron function in the Caenorhabditis elegans connectome. \u003cem\u003eNature\u003c/em\u003e 2017; 550(7677): 519\u0026ndash;523.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDemontis D, Walters GB, Athanasiadis G, Walters R, Therrien K, Nielsen TT \u003cem\u003eet al.\u003c/em\u003e Genome-wide analyses of ADHD identify 27 risk loci, refine the genetic architecture and implicate several cognitive domains. \u003cem\u003eNature genetics\u003c/em\u003e 2023; 55(2): 198\u0026ndash;208.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrove J, Ripke S, Als TD, Mattheisen M, Walters RK, Won H \u003cem\u003eet al.\u003c/em\u003e Identification of common genetic risk variants for autism spectrum disorder. \u003cem\u003eNature genetics\u003c/em\u003e 2019; 51(3): 431\u0026ndash;444.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCardon LR, Palmer LJ. Population stratification and spurious allelic association. \u003cem\u003eLancet\u003c/em\u003e 2003; 361(9357): 598\u0026ndash;604.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGenomes Project C, Auton A, Brooks LD, Durbin RM, Garrison EP, Kang HM \u003cem\u003eet al.\u003c/em\u003e A global reference for human genetic variation. \u003cem\u003eNature\u003c/em\u003e 2015; 526(7571): 68\u0026ndash;74.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePurcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D \u003cem\u003eet al.\u003c/em\u003e PLINK: a tool set for whole-genome association and population-based linkage analyses. \u003cem\u003eAm J Hum Genet\u003c/em\u003e 2007; 81(3): 559\u0026ndash;575.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChoi SW, O'Reilly PF. PRSice-2: Polygenic Risk Score software for biobank-scale data. \u003cem\u003eGigascience\u003c/em\u003e 2019; 8(7).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePetersen AC, Crockett L, Richards M, Boxer A. A self-report measure of pubertal status: Reliability, validity, and initial norms. \u003cem\u003eJ Youth Adolesc\u003c/em\u003e 1988; 17(2): 117\u0026ndash;133.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKorthauer K, Kimes PK, Duvallet C, Reyes A, Subramanian A, Teng M \u003cem\u003eet al.\u003c/em\u003e A practical guide to methods controlling false discoveries in computational biology. \u003cem\u003eGenome Biol\u003c/em\u003e 2019; 20(1): 118.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBates D, M\u0026auml;chler M, Bolker B, Walker S. Fitting Linear Mixed-Effects Models Using lme4. \u003cem\u003eJournal of Statistical Software\u003c/em\u003e 2015; 67(1): 1\u0026ndash;48.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAF H. \u003cem\u003eIntroduction to mediation, moderation, and conditional process analysis: A regression-based approach\u003c/em\u003e. Guilford Press2017.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHamaker EL, Kuiper RM, Grasman RP. A critique of the cross-lagged panel model. \u003cem\u003ePsychol Methods\u003c/em\u003e 2015; 20(1): 102\u0026ndash;116.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSmith KE, Pollak SD. Rethinking Concepts and Categories for Understanding the Neurodevelopmental Effects of Childhood Adversity. \u003cem\u003ePerspect Psychol Sci\u003c/em\u003e 2021; 16(1): 67\u0026ndash;93.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGee DG. Early Adversity and Development: Parsing Heterogeneity and Identifying Pathways of Risk and Resilience. \u003cem\u003eAm J Psychiatry\u003c/em\u003e 2021; 178(11): 998\u0026ndash;1013.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKemp JVA, Bernier E, Lebel C, Kopala-Sibley DC. Associations Between Parental Mood and Anxiety Psychopathology and Offspring Brain Structure: A Scoping Review. \u003cem\u003eClin Child Fam Psychol Rev\u003c/em\u003e 2022; 25(1): 222\u0026ndash;247.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFarah MJ. The Neuroscience of Socioeconomic Status: Correlates, Causes, and Consequences. \u003cem\u003eNeuron\u003c/em\u003e 2017; 96(1): 56\u0026ndash;71.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNoble KG, Houston SM, Brito NH, Bartsch H, Kan E, Kuperman JM \u003cem\u003eet al.\u003c/em\u003e Family income, parental education and brain structure in children and adolescents. \u003cem\u003eNat Neurosci\u003c/em\u003e 2015; 18(5): 773\u0026ndash;778.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarozza S, Kletenik I, Astle D, Schwamm L, Dhand A. Whole-brain white matter variation across childhood environments. \u003cem\u003eProc Natl Acad Sci U S A\u003c/em\u003e 2025; 122(15): e2409985122.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHolz NE, Berhe O, Sacu S, Schwarz E, Tesarz J, Heim CM \u003cem\u003eet al.\u003c/em\u003e Early Social Adversity, Altered Brain Functional Connectivity, and Mental Health. \u003cem\u003eBiol Psychiatry\u003c/em\u003e 2023; 93(5): 430\u0026ndash;441.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSchimmelpfennig J, Topczewski J, Zajkowski W, Jankowiak-Siuda K. The role of the salience network in cognitive and affective deficits. \u003cem\u003eFront Hum Neurosci\u003c/em\u003e 2023; 17: 1133367.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarek S, Dosenbach NUF. The frontoparietal network: function, electrophysiology, and importance of individual precision mapping. \u003cem\u003eDialogues Clin Neurosci\u003c/em\u003e 2018; 20(2): 133\u0026ndash;140.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi L, Wang L. Linking visual-frontoparietal network neural dynamics to spontaneous cognitive processing. \u003cem\u003eNeuroimage\u003c/em\u003e 2025; 312: 121229.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMendoza G, Merchant H. Motor system evolution and the emergence of high cognitive functions. \u003cem\u003eProg Neurobiol\u003c/em\u003e 2014; 122: 73\u0026ndash;93.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcLaughlin KA, Weissman D, Bitran D. Childhood Adversity and Neural Development: A Systematic Review. \u003cem\u003eAnnu Rev Dev Psychol\u003c/em\u003e 2019; 1: 277\u0026ndash;312.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChahal R, Miller JG, Yuan JP, Buthmann JL, Gotlib IH. An exploration of dimensions of early adversity and the development of functional brain network connectivity during adolescence: Implications for trajectories of internalizing symptoms. \u003cem\u003eDev Psychopathol\u003c/em\u003e 2022; 34(2): 557\u0026ndash;571.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMenon V, Uddin LQ. Saliency, switching, attention and control: a network model of insula function. \u003cem\u003eBrain Struct Funct\u003c/em\u003e 2010; 214(5\u0026ndash;6): 655\u0026ndash;667.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLammertink F, van den Heuvel MP, Hermans EJ, Dudink J, Tataranno ML, Benders M \u003cem\u003eet al.\u003c/em\u003e Early-life stress exposure and large-scale covariance brain networks in extremely preterm-born infants. \u003cem\u003eTransl Psychiatry\u003c/em\u003e 2022; 12(1): 256.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHerzberg MP, McKenzie KJ, Hodel AS, Hunt RH, Mueller BA, Gunnar MR \u003cem\u003eet al.\u003c/em\u003e Accelerated maturation in functional connectivity following early life stress: Circuit specific or broadly distributed? \u003cem\u003eDev Cogn Neurosci\u003c/em\u003e 2021; 48: 100922.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCallaghan BL, Tottenham N. The Stress Acceleration Hypothesis: Effects of early-life adversity on emotion circuits and behavior. \u003cem\u003eCurr Opin Behav Sci\u003c/em\u003e 2016; 7: 76\u0026ndash;81.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChan SY, Ngoh ZM, Ong ZY, Teh A, Kee MZL, Zhou JH \u003cem\u003eet al.\u003c/em\u003e The influence of early-life adversity on the coupling of structural and functional brain connectivity across childhood. \u003cem\u003eNat Ment Health\u003c/em\u003e 2024; 2(1).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHuang D, Liu Z, Cao H, Yang J, Wu Z, Long Y. Childhood trauma is linked to decreased temporal stability of functional brain networks in young adults. \u003cem\u003eJ Affect Disord\u003c/em\u003e 2021; 290: 23\u0026ndash;30.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMcEwen BS, Gianaros PJ. Central role of the brain in stress and adaptation: links to socioeconomic status, health, and disease. \u003cem\u003eAnn N Y Acad Sci\u003c/em\u003e 2010; 1186: 190\u0026ndash;222.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVincent JL, Kahn I, Snyder AZ, Raichle ME, Buckner RL. Evidence for a frontoparietal control system revealed by intrinsic functional connectivity. \u003cem\u003eJ Neurophysiol\u003c/em\u003e 2008; 100(6): 3328\u0026ndash;3342.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMenon V, D'Esposito M. The role of PFC networks in cognitive control and executive function. \u003cem\u003eNeuropsychopharmacology\u003c/em\u003e 2022; 47(1): 90\u0026ndash;103.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede la Vega A, Chang LJ, Banich MT, Wager TD, Yarkoni T. Large-Scale Meta-Analysis of Human Medial Frontal Cortex Reveals Tripartite Functional Organization. \u003cem\u003eJ Neurosci\u003c/em\u003e 2016; 36(24): 6553\u0026ndash;6562.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCui Z, Stiso J, Baum GL, Kim JZ, Roalf DR, Betzel RF \u003cem\u003eet al.\u003c/em\u003e Optimization of energy state transition trajectory supports the development of executive function during youth. \u003cem\u003eElife\u003c/em\u003e 2020; 9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNigg J, Nikolas M, Burt SA. Measured gene-by-environment interaction in relation to attention-deficit/hyperactivity disorder. \u003cem\u003eJ Am Acad Child Adolesc Psychiatry\u003c/em\u003e 2010; 49(9): 863\u0026ndash;873.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChaste P, Leboyer M. Autism risk factors: genes, environment, and gene-environment interactions. \u003cem\u003eDialogues Clin Neurosci\u003c/em\u003e 2012; 14(3): 281\u0026ndash;292.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSydnor VJ, Larsen B, Bassett DS, Alexander-Bloch A, Fair DA, Liston C \u003cem\u003eet al.\u003c/em\u003e Neurodevelopment of the association cortices: Patterns, mechanisms, and implications for psychopathology. \u003cem\u003eNeuron\u003c/em\u003e 2021; 109(18): 2820\u0026ndash;2846.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBos MGN, Wierenga LM, Blankenstein NE, Schreuders E, Tamnes CK, Crone EA. Longitudinal structural brain development and externalizing behavior in adolescence. \u003cem\u003eJ Child Psychol Psychiatry\u003c/em\u003e 2018; 59(10): 1061\u0026ndash;1072.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"molecular-psychiatry","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"mp","sideBox":"Learn more about [Molecular Psychiatry](http://www.nature.com/mp/)","snPcode":"41380","submissionUrl":"https://mts-mp.nature.com/cgi-bin/main.plex","title":"Molecular Psychiatry","twitterHandle":"@molpsychiatry","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8777490/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8777490/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eEarly life adversity (ELA) is a robust transdiagnostic risk factor for mental health disorders, yet the neurobiological mechanisms mediating its long-term impact remain poorly understood. Network control theory offers a novel framework for capturing how structural brain networks constrain and support brain dynamics. Controllability increases over development, associates with executive function and mental health, and appears sensitive to environmental insults. Thus, it may reflect a neurobiological mediator between ELA and behavioral outcomes. We tested whether alterations in modal controllability mediate the impact of multidimensional ELA on cognitive and behavioral outcomes in youth, and whether these pathways are shaped by genetic risk for neurodevelopmental conditions. Using data from 7,970 children aged 9\u0026ndash;11 years in the Adolescent Brain Cognitive Development (ABCD) Study, we derived five latent ELA dimensions from 67 indicators, and computed polygenic risk scores (PRS) for attention-deficit/hyperactivity disorder (ADHD\u003csub\u003ePRS\u003c/sub\u003e) and autism spectrum disorder (ASD\u003csub\u003ePRS\u003c/sub\u003e). Distinct ELA dimensions were associated with increased controllability in medial frontal, frontoparietal, default mode, and motor networks, as well as with externalizing symptoms and impaired crystallized cognition. Controllability partially mediated these associations, and indirect effects were significantly moderated ADHD\u003csub\u003ePRS\u003c/sub\u003e and ASD\u003csub\u003ePRS\u003c/sub\u003e. Longitudinal analyses further demonstrated that baseline controllability predicted cognitive performance two years later. These findings delineate a neurodevelopmental cascade linking early adversity and genetic vulnerability to transdiagnostic mental health risk, positioning brain controllability as a promising mechanistic marker and potential target for early intervention.\u003c/p\u003e","manuscriptTitle":"Brain controllability mediates the effects of early life adversity on adolescent behavior and cognition moderated by genetic risks","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-19 16:15:25","doi":"10.21203/rs.3.rs-8777490/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"This content is not available.","date":"2026-04-01T17:06:17+00:00","index":2,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2026-03-30T13:53:29+00:00","index":3,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2026-03-26T19:59:14+00:00","index":2,"fulltext":"This content is not available."},{"type":"reviewerAgreed","content":"This content is not available.","date":"2026-03-26T19:51:08+00:00","index":1,"fulltext":"This content is not available."},{"type":"reviewersInvited","content":"","date":"2026-03-16T23:37:41+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-02-04T15:19:32+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-02-04T15:14:58+00:00","index":"","fulltext":""},{"type":"submitted","content":"Molecular Psychiatry","date":"2026-02-04T15:09:29+00:00","index":"","fulltext":""},{"type":"checksFailed","content":"","date":"2026-02-04T14:27:16+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"molecular-psychiatry","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"mp","sideBox":"Learn more about [Molecular Psychiatry](http://www.nature.com/mp/)","snPcode":"41380","submissionUrl":"https://mts-mp.nature.com/cgi-bin/main.plex","title":"Molecular Psychiatry","twitterHandle":"@molpsychiatry","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"ec017d90-a923-49e4-88ef-97fea4a96581","owner":[],"postedDate":"March 19th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":64611693,"name":"Biological sciences/Neuroscience"},{"id":64611694,"name":"Health sciences/Biomarkers/Predictive markers"},{"id":64611695,"name":"Health sciences/Diseases/Psychiatric disorders/ADHD"},{"id":64611696,"name":"Health sciences/Diseases/Psychiatric disorders/Autism spectrum disorders"},{"id":64611697,"name":"Biological sciences/Psychology"}],"tags":[],"updatedAt":"2026-03-19T16:15:25+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-19 16:15:25","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8777490","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8777490","identity":"rs-8777490","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}