The growing interdisciplinarity of developmental psychopathology: Implications for science and training.

The field of developmental psychopathology has grown exponentially over the past decades, and has become increasingly multifaceted. The initial focus on understanding abnormal child psychology has broadened to the study of the origins of psychopathology, with the goals of preventing and alleviating disorder and promoting healthy development. In this paper, we discuss how technological advances and global events have expanded the questions that researchers in developmental psychopathology can address. We do so by describing a longitudinal study that we have been conducting for the past dozen years. We originally planned to examine the effects of early adversity on trajectories of brain development, endocrine function, and depressive symptoms across puberty; it has since become an interdisciplinary study encompassing diverse domains like inflammation, sleep, biological aging, the environment, and child functioning post-pandemic, that we believe will advance our understanding of neurobehavioral development. This increase in the breadth in our study emerged from an expansion of the field; we encourage researchers to embrace these dynamic changes. In this context, we discuss challenges, opportunities, and institutional changes related to the growing interdisciplinarity of the field with respect to training the next generation of investigators to mitigate the burden of mental illness in youth.

The cortisol/DHEA ratio mediates the association between early life stress and externalizing problems in adolescent boys.

BACKGROUND: Despite evidence that early life stress (ELS) can influence the functioning of the hypothalamic-pituitary-adrenal (HPA) axis and increase maladaptive behaviors in adolescence, less attention has been paid to the role of the coordinated effects of the two primary adrenal hormones, cortisol and dehydroepiandrosterone (DHEA), in these associations. METHODS: 138 typically developing adolescents (76 females) reported the stressful events experienced during childhood and early adolescence across 30 domains. Two years later we assessed levels of externalizing problems and obtained salivary levels of cortisol and DHEA. Using causal moderated mediation analyses, we examined whether the ratio of cortisol to DHEA (CD ratio) mediates the association between ELS and subsequent externalizing problems. RESULTS: We found that ELS is associated with both a lower CD ratio and more externalizing problems. Importantly, a lower CD ratio mediated the association between ELS and externalizing problems in boys. CONCLUSIONS: An imbalance in adrenal hormones may be a mechanism through which ELS leads to an increase in externalizing problems in adolescent boys. These findings underscore the utility of using the CD ratio to index HPA-axis functioning.

Intracranial Recordings of the Human Orbitofrontal Cortical Activity during Self-Referential Episodic and Valenced Self-Judgments.

We recorded directly from the orbital (oPFC) and ventromedial (vmPFC) subregions of the orbitofrontal cortex (OFC) in 22 (9 female, 13 male) epilepsy patients undergoing intracranial electroencephalography (iEEG) monitoring during an experimental task in which the participants judged the accuracy of self-referential autobiographical statements as well as valenced self-judgments (SJs). We found significantly increased high-frequency activity (HFA) in ∼13% of oPFC sites (10/18 subjects) and 16% of vmPFC sites (4/12 subjects) during both of these self-referential thought processes, with the HFA power being modulated by the content of self-referential stimuli. The location of these activated sites corresponded with the location of fMRI-identified limbic network. Furthermore, the onset of HFA in the vmPFC was significantly earlier than that in the oPFC in all patients with simultaneous recordings in both regions. In 11 patients with available depression scores from comprehensive neuropsychological assessments, we documented diminished HFA in the OFC during positive SJ trials among individuals with higher depression scores; responses during negative SJ trials were not related to the patients' depression scores. Our findings provide new temporal and anatomical information about the mode of engagement in two important subregions of the OFC during autobiographical memory and SJ conditions. Our findings from the OFC support the hypothesis that diminished brain activity during positive self-evaluations, rather than heightened activity during negative self-evaluations, plays a key role in the pathophysiology of depression.

Multi-site benchmark classification of major depressive disorder using machine learning on cortical and subcortical measures.

Machine learning (ML) techniques have gained popularity in the neuroimaging field due to their potential for classifying neuropsychiatric disorders. However, the diagnostic predictive power of the existing algorithms has been limited by small sample sizes, lack of representativeness, data leakage, and/or overfitting. Here, we overcome these limitations with the largest multi-site sample size to date (N = 5365) to provide a generalizable ML classification benchmark of major depressive disorder (MDD) using shallow linear and non-linear models. Leveraging brain measures from standardized ENIGMA analysis pipelines in FreeSurfer, we were able to classify MDD versus healthy controls (HC) with a balanced accuracy of around 62%. But after harmonizing the data, e.g., using ComBat, the balanced accuracy dropped to approximately 52%. Accuracy results close to random chance levels were also observed in stratified groups according to age of onset, antidepressant use, number of episodes and sex. Future studies incorporating higher dimensional brain imaging/phenotype features, and/or using more advanced machine and deep learning methods may yield more encouraging prospects.

Sex-Specific Vulnerability to Externalizing Problems: Sensitivity to Early Stress and Nucleus Accumbens Activation Over Adolescence.

BACKGROUND: Exposure and sensitivity to early-life stress (ELS) are related to increased risk for psychopathology in adolescence. While cross-sectional studies have reported blunted nucleus accumbens (NAcc) activation in the context of these associations, researchers have not yet assessed the effects of ELS on developmental trajectories of activation. We examined whether trajectories are affected by stress and the moderating role of biological sex in predicting vulnerability to symptoms of psychopathology. METHODS: Adolescents (n = 173) completed 3 assessments at 2-year intervals across puberty (ages 9-18 years). At baseline, we assessed objective ELS and stress sensitivity using the Traumatic Events Screening Inventory for Children. At all time points, we assessed NAcc activation using the Monetary Incentive Delay task and externalizing, internalizing, and total problems using the Youth Self-Report. We examined correlations between NAcc trajectories (extracted using linear mixed-effects models) with ELS and stress sensitivity and conducted multivariate regression analysis to examine the interaction of NAcc trajectories and biological sex in predicting symptoms of psychopathology. RESULTS: Symptoms increased over adolescence. Stress sensitivity, but not objective ELS, was associated with decreasing trajectories of NAcc activation. Biological sex interacted with NAcc trajectories to predict psychopathology; boys, but not girls, with decreasing NAcc activation had more severe externalizing problems in adolescence. These findings were replicated in the putamen and caudate but not in the medial prefrontal cortex or control brain regions. CONCLUSIONS: NAcc activation may be a sex-specific marker of externalizing problems in adolescence. Efforts to reduce stress sensitivity may help to decrease symptoms of psychopathology in adolescent boys.

Large-scale proteomics in the first trimester of pregnancy predict psychopathology and temperament in preschool children: an exploratory study.

BACKGROUND: Understanding the prenatal origins of children's psychopathology is a fundamental goal in developmental and clinical science. Recent research suggests that inflammation during pregnancy can trigger a cascade of fetal programming changes that contribute to vulnerability for the emergence of psychopathology. Most studies, however, have focused on a handful of proinflammatory cytokines and have not explored a range of prenatal biological pathways that may be involved in increasing postnatal risk for emotional and behavioral difficulties. METHODS: Using extreme gradient boosted machine learning models, we explored large-scale proteomics, considering over 1,000 proteins from first trimester blood samples, to predict behavior in early childhood. Mothers reported on their 3- to 5-year-old children's (N = 89, 51% female) temperament (Child Behavior Questionnaire) and psychopathology (Child Behavior Checklist). RESULTS: We found that machine learning models of prenatal proteomics predict 5%-10% of the variance in children's sadness, perceptual sensitivity, attention problems, and emotional reactivity. Enrichment analyses identified immune function, nervous system development, and cell signaling pathways as being particularly important in predicting children's outcomes. CONCLUSIONS: Our findings, though exploratory, suggest processes in early pregnancy that are related to functioning in early childhood. Predictive features included far more proteins than have been considered in prior work. Specifically, proteins implicated in inflammation, in the development of the central nervous system, and in key cell-signaling pathways were enriched in relation to child temperament and psychopathology measures.

A network analysis of psychopathology in young Black children: Implications for predicting outcomes in adolescence.

OBJECTIVE: Network analysis may identify specific symptoms involved in the maintenance and development of psychopathology. This approach, however, has not been applied to the study of young Black children, a population facing unique challenges and developmental risks. It is also unclear whether network analysis identifies early symptoms in Black children that are linked to their longer-term difficulties and strengths in adolescence. METHODS: We conducted a network analysis of emotional and behavioral difficulties in 1238 Black (non-Hispanic) children from the age-3 assessment in the Future of Families and Child Wellbeing Study (47 % female). We also explored whether early childhood symptoms predict subsequent caregiver-reported internalizing and externalizing problems, and youth-reported social competencies and extracurricular and community involvement, at the age-15 assessment. RESULTS: We identified specific symptoms of externalizing and emotional reactivity as central in the network. Symptoms of emotional reactivity were also involved in comorbidity, bridging different communities of symptoms. Using elastic net models, we identified specific central and bridge symptoms, but also peripheral network symptoms, that contributed uniquely to the prediction of internalizing and externalizing problems in adolescence. Early childhood symptoms were less predictive of positive outcomes in adolescence. CONCLUSIONS: This study identified central and bridge symptoms in young Black children, an underrepresented population in network analysis research. Some of these central and bridge symptoms, but also peripheral network symptoms, may be useful targets in early interventions to prevent long-term difficulties. Conversely, network approaches to understanding early psychopathology may have less utility for predicting Black children's subsequent strengths in adolescence.

Exploring sex differences in trajectories of pubertal development and mental health following early adversity.

Despite evidence that early life adversity (ELA) affects mental health in adolescence, we know little about sex differences in how distinct dimensions of adversity affect development and their corresponding effects on mental health. In this three-wave longitudinal study, 209 participants (118 females; ages 9-13 years at baseline) provided objective (salivary hormones, BMI, age of menarche) and subjective (perceived gonadal and adrenal status) measures of puberty and physical development, and reported on levels of internalizing and externalizing symptoms at all timepoints. Participants also reported lifetime exposure to three distinct types of ELA: deprivation, threat, and unpredictability. Using generalized additive mixed models, we tested within each sex whether dimensions of adversity were associated with longitudinal changes in measures of pubertal and physical development, and whether these indices of development were associated with trajectories of internalizing and externalizing symptoms. In females, experiences of threat and unpredictability were significantly associated with earlier pubertal timing (e.g., age of menarche) whereas experiences of deprivation were associated with steeper increases in BMI; further, faster pubertal tempo (i.e., steeper increases in pubertal stage) was associated with increases in internalizing and externalizing symptoms. In males, however, ELA was not associated with any measures of pubertal or physical development or with symptoms. Together, our results suggest that adverse experiences during early life have sex-selective consequences for pubertal and physical maturation and mental health trajectories in ways that may elucidate why females are at higher risk for mental health difficulties during puberty, particularly following exposure to unpredictable and threatening experiences of adversity.

Early life stress moderates the relation between systemic inflammation and neural activation to reward in adolescents both cross-sectionally and longitudinally.

Elevated levels of systemic inflammation are associated with altered reward-related brain function in ventral striatal areas of the brain like the nucleus accumbens (NAcc). In adolescents, cross-sectional research indicates that exposure to early life stress (ELS) can moderate the relation between inflammation and neural activation, which may contribute to atypical reward function; however, no studies have tested whether this moderation by ELS of neuroimmune associations persists over time. Here, we conducted a cross-sectional analysis and the first exploratory longitudinal analysis testing whether cumulative severity of ELS moderates the association of systemic inflammation with reward-related processing in the NAcc in adolescents (n = 104; 58F/46M; M[SD] age = 16.00[1.45] years; range = 13.07-19.86 years). For the cross-sectional analysis, we modeled a statistical interaction between ELS and levels of C-reactive protein (CRP) predicting NAcc activation during the anticipation and outcome phases of a monetary reward task. We found that higher CRP was associated with blunted NAcc activation during the outcome of reward in youth who experienced higher levels of ELS (ß = -0.31; p = 0.006). For the longitudinal analysis, we modeled an interaction between ELS and change in CRP predicting change in NAcc activation across 2 years. This analysis similarly showed that increasing CRP over time was associated with decreasing NAcc during reward outcomes in youth who experienced higher levels of ELS (ß = -0.47; p = 0.022). Both findings support contemporary theoretical frameworks involving associations among inflammation, reward-related brain function, and ELS exposure, and suggest that experiencing ELS can have significant and enduring effects on neuroimmune function and adolescent neurodevelopment.

Neighborhood Socioeconomic Disadvantage and White Matter Microstructure of the Arcuate Fasciculus and Uncinate Fasciculus in Adolescents.

Background: Neighborhood- or area-level socioeconomic disadvantage is associated with neural alterations across the life span. However, few studies have examined the effects of neighborhood disadvantage on white matter microstructure during adolescence, an important period of development that coincides with increased risk for psychopathology. Methods: In 200 adolescents (ages 13-20 years; 54.5% female, 4% nonbinary) recruited from 2 studies enriched for early adversity and depression, we examined whether neighborhood socioeconomic disadvantage derived from census tract data was related to white matter microstructure in several major white matter tracts. We also examined whether depressive symptoms and sex moderated these associations. Results: Greater neighborhood socioeconomic disadvantage was associated with lower fractional anisotropy (FA) in the left arcuate fasciculus (ß = -0.24, false discovery rate [FDR]-corrected p = .035) and right uncinate fasciculus (ß = -0.32, FDR-corrected p = .002) above and beyond the effects of family-level socioeconomic status. Depressive symptoms significantly moderated the association between left arcuate fasciculus FA and both neighborhood (ß = 0.17, FDR-corrected p = .026) and unemployment (ß = 0.22, FDR-corrected p = .004) disadvantage such that these associations were only significant in adolescents who reported less severe depression. Sex did not moderate the association between socioeconomic disadvantage and FA in these tracts. Conclusions: Greater neighborhood socioeconomic disadvantage, particularly poverty and educational attainment levels, was associated with lower FA in the arcuate fasciculus and uncinate fasciculus above and beyond the effects of family-level measures of socioeconomic status. These patterns were only observed in adolescents with low levels of depression, suggesting that we must be cautious about generalizing these findings to youths who struggle with mental health difficulties.

Genetics, epigenetics, and neurobiology of childhood-onset depression: an umbrella review.

Depression is a serious and persistent psychiatric disorder that commonly first manifests during childhood. Depression that starts in childhood is increasing in frequency, likely due both to evolutionary trends and to increased recognition of the disorder. In this umbrella review, we systematically searched the extant literature for genetic, epigenetic, and neurobiological factors that contribute to a childhood onset of depression. We searched PubMed, EMBASE, OVID/PsychInfo, and Google Scholar with the following inclusion criteria: (1) systematic review or meta-analysis from a peer-reviewed journal; (2) inclusion of a measure assessing early age of onset of depression; and (3) assessment of neurobiological, genetic, environmental, and epigenetic predictors of early onset depression. Findings from 89 systematic reviews of moderate to high quality suggest that childhood-onset depressive disorders have neurobiological, genetic, environmental, and epigenetic roots consistent with a diathesis-stress theory of depression. This review identified key putative markers that may be targeted for personalized clinical decision-making and provide important insights concerning candidate mechanisms that might underpin the early onset of depression.

Early life stress predicts trajectories of emotional problems and hippocampal volume in adolescence.

Exposure to early life stress (ELS) has been consistently associated with adverse emotional and neural consequences in youth. The development of brain structures such as the hippocampus, which plays a significant role in stress and emotion regulation, may be particularly salient in the development of psychopathology. Prior work has documented smaller hippocampal volume (HCV) in relation to both ELS exposure and risk for psychopathology. We used longitudinal k-means clustering to identify simultaneous trajectories of HCV and emotional problems in 155 youth across three assessments conducted approximately two years apart (mean baseline age = 11.33 years, 57% female). We also examined depressive symptoms and resilience approximately two years after the third timepoint. We identified three clusters of participants: a cluster with high HCV and low emotional problems; a cluster with low HCV and high emotional problems; and a cluster with low HCV and low emotional problems. Importantly, severity of ELS was associated with greater likelihood of belonging to the low HCV/high symptom cluster than to the low HCV/low symptom cluster. Further, low HCV/high symptom participants had more depressive symptoms and lower resilience scores than did participants in the low HCV/low symptom, but not than in the high HCV/low symptom cluster. Our findings suggest that smaller HCV indexes biological sensitivity to stress. This adds to our understanding of the ways in which ELS can affect hippocampal and emotional development in young people and points to hippocampal volume as a marker of susceptibility to context.

Associations among early life adversity, sleep disturbances, and depressive symptoms in adolescent females and males: a longitudinal investigation.

BACKGROUND: Exposure to adversity early in life (ELA) has been associated with elevated risk for depression during adolescence, particularly for females; the mechanisms underlying this association, however, are poorly understood. One potential mechanism linking ELA and sex differences in depressive symptoms is sleep disturbances, which increase during adolescence and are more common in females. Here, we examined whether sleep disturbances mediate the association between ELA and increases in depressive symptoms during adolescence and whether this mediation differs by sex. METHODS: 224 (N = 132 females) youth were recruited at age 9-13 years and assessed every 2 years across three timepoints. At the first timepoint, we conducted extensive interviews about stressful events participants experienced; participants provided subjective severity ratings of events and we objectively scored the severity of each event. Self-reported sleep disturbances and depressive symptoms were assessed at all timepoints. We conducted linear mixed models to estimate both initial levels and changes in sleep disturbances and depressive symptoms, and moderated mediation analyses to test whether initial levels and/or changes in sleep disturbances mediated the association of ELA (objective and subjective) with increases in depressive symptoms across adolescence and whether the mediations differed by sex. RESULTS: While higher initial levels and increases in sleep problems were uniquely associated with increases in depressive symptoms for males and females, they were related to ELA differently by sex. For females, greater ELA (both objectively and subjectively rated) was associated with higher initial levels of sleep problems, which in turn were associated with increases in depressive symptoms from early to late adolescence. In contrast, for males, ELA exposure was not associated with either initial levels of, or increases in, sleep problems. CONCLUSIONS: These findings highlight the role of sleep disturbances during the transition to adolescence in mediating sex differences in the effects of ELA on depressive symptoms.

Preparation and processing of dried blood spots for microRNA sequencing.

Dried blood spots (DBS) are biological samples commonly collected from newborns and in geographic areas distanced from laboratory settings for the purposes of disease testing and identification. MicroRNAs (miRNAs)-small non-coding RNAs that regulate gene activity at the post-transcriptional level-are emerging as critical markers and mediators of disease, including cancer, infectious diseases, and mental disorders. This protocol describes optimized procedural steps for utilizing DBS as a reliable source of biological material for obtaining peripheral miRNA expression profiles. We outline key practices, such as the method of DBS rehydration that maximizes RNA extraction yield, and the use of degenerate oligonucleotide adapters to mitigate ligase-dependent biases that are associated with small RNA sequencing. The standardization of miRNA readout from DBS offers numerous benefits: cost-effectiveness in sample collection and processing, enhanced reliability and consistency of miRNA profiling, and minimal invasiveness that facilitates repeated testing and retention of participants. The use of DBS-based miRNA sequencing is a promising method to investigate disease mechanisms and to advance personalized medicine.

Biological sensitivity to adolescent-parent discrepancies in perceived parental warmth.

Introduction: Parenting behaviors are formative to the psychological development of young people; however, parent and adolescent perceptions of parenting are only moderately correlated with each other. Whereas discrepant perceptions may represent a normative process of deindividuation from caregivers in some adolescents, in others a discrepancy might predict psychological maladjustment. The biological sensitivity to context model provides a framework from which individual differences in development can be estimated in adolescents whose perceptions of parenting diverge from those of their parents. Methods: At baseline we obtained diurnal cortisol samples from US adolescents (M = 13.37 years of age, SD = 1.06) as well as parents' and adolescents' ratings of parental warmth; we obtained adolescent-reported symptoms of psychopathology at baseline and again at follow-up two years later (N = 108, 57.5% female). We estimated waking cortisol, cortisol awakening response, and daytime cortisol slopes using piecewise regression models. Results: Lower adolescent than parent ratings of parental warmth predicted increased externalizing symptoms at follow-up. Higher waking cortisol and steeper cortisol awakening response and daytime slopes predicted increased internalizing symptoms at follow-up. Further, discrepant ratings of parental warmth interacted with cortisol awakening response and daytime slopes such that greater discrepancies predicted greater increases in externalizing symptoms in adolescents with steeper cortisol slopes. Conclusions: These findings indicate that steeper changes in cortisol production throughout the day index a greater sensitivity to perceived parental warmth. Lower adolescent than parent ratings of parental warmth may represent dysfunction in the parental relationship rather than a normative process of deindividuation in adolescents with steeper diurnal cortisol slopes.

Early life stress, sleep disturbances, and depressive symptoms during adolescence: The role of the cingulum bundle.

Adolescence is often characterized by sleep disturbances that can affect the development of white matter tracts implicated in affective and cognitive regulation, including the cingulate portion of the cingulum bundle (CGC) and the uncinate fasciculus (UF). These effects may be exacerbated in adolescents exposed to early life adversity (ELA). We examined the longitudinal relations between sleep problems and CGC and UF microstructure during adolescence and their relation to depressive symptoms as a function of exposure to ELA. We assessed self-reported sleep disturbances and depressive symptoms and acquired diffusion-weighted MRI scans twice: in early adolescence (9-13 years) and four years later (13-17 years) (N = 72 complete cases). Independent of ELA, higher initial levels and increases in sleep problems were related to increases in depressive symptoms. Further, increases in right CGC fractional anisotropy (FA) mediated the association between sleep problems and depressive symptoms for youth who experienced lower, but not higher, levels of ELA. In youth with higher ELA, higher initial levels of and steeper decreases in sleep problems were associated with greater decreases in right UF FA, but were unrelated to depressive symptoms. Our findings highlight the importance of sleep quality in shaping fronto-cingulate-limbic tract development and depressive symptoms during adolescence.

Teaching or learning from baby: Inducing explicit parenting goals influences caregiver intrusiveness.

Caregivers' goals influence their interactions with their children. In this preregistered study, we examined whether directing parents to teach their baby versus learn from their baby influenced the extent to which they engaged in intrusive (e.g., controlling, adult-centered rather than child-centered), sensitive, warm, or cognitively stimulating caregiving behaviors. Mothers and their 6-month-old infants (N = 66; 32 female infants) from the San Francisco Bay Area participated in a 10-min "free-play" interaction, coded in 2-min epochs for degree of parental intrusiveness. Prior to the final epoch, mothers were randomly assigned to receive instructions to focus on (a) teaching something to their infant or (b) learning something from their infant. A control group of mothers received no instructions. Analyses of within-person changes in intrusive behavior from before to after receiving these instructions indicated that mothers assigned to teach their infant increased in intrusiveness whereas mothers assigned to learn from their infant and mothers in the control group did not significantly change in intrusiveness. The study provides experimental evidence that caregivers' explicit goals to teach infants result, on average, in more controlling and adult-centered caregiving behavior. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Concurrent validity and reliability of suicide risk assessment instruments: A meta-analysis of 20 instruments across 27 international cohorts.

OBJECTIVE: A major limitation of current suicide research is the lack of power to identify robust correlates of suicidal thoughts or behavior. Variation in suicide risk assessment instruments used across cohorts may represent a limitation to pooling data in international consortia. METHOD: Here, we examine this issue through two approaches: (a) an extensive literature search on the reliability and concurrent validity of the most commonly used instruments and (b) by pooling data (N ∼ 6,000 participants) from cohorts from the Enhancing NeuroImaging Genetics Through Meta-Analysis (ENIGMA) Major Depressive Disorder and ENIGMA-Suicidal Thoughts and Behaviour working groups, to assess the concurrent validity of instruments currently used for assessing suicidal thoughts or behavior. RESULTS: We observed moderate-to-high correlations between measures, consistent with the wide range (κ range: 0.15-0.97; r range: 0.21-0.94) reported in the literature. Two common multi-item instruments, the Columbia Suicide Severity Rating Scale and the Beck Scale for Suicidal Ideation were highly correlated with each other (r = 0.83). Sensitivity analyses identified sources of heterogeneity such as the time frame of the instrument and whether it relies on self-report or a clinical interview. Finally, construct-specific analyses suggest that suicide ideation items from common psychiatric questionnaires are most concordant with the suicide ideation construct of multi-item instruments. CONCLUSIONS: Our findings suggest that multi-item instruments provide valuable information on different aspects of suicidal thoughts or behavior but share a modest core factor with single suicidal ideation items. Retrospective, multisite collaborations including distinct instruments should be feasible provided they harmonize across instruments or focus on specific constructs of suicidality. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

The default mode network is associated with changes in internalizing and externalizing problems differently in adolescent boys and girls.

Internalizing and externalizing problems that emerge during adolescence differentially increase boys' and girls' risk for developing psychiatric disorders. It is not clear, however, whether there are sex differences in the intrinsic functional architecture of the brain that underlie changes in the severity of internalizing and externalizing problems in adolescents. Using resting-state fMRI data and self-reports of behavioral problems obtained from 128 adolescents (73 females; 9-14 years old) at two timepoints, we conducted multivoxel pattern analysis to identify resting-state functional connectivity markers at baseline that predict changes in the severity of internalizing and externalizing problems in boys and girls 2 years later. We found sex-differentiated involvement of the default mode network in changes in internalizing and externalizing problems. Whereas changes in internalizing problems were associated with the dorsal medial subsystem in boys and with the medial temporal subsystem in girls, changes in externalizing problems were predicted by hyperconnectivity between core nodes of the DMN and frontoparietal network in boys and hypoconnectivity between the DMN and affective networks in girls. Our results suggest that different neural mechanisms predict changes in internalizing and externalizing problems in adolescent boys and girls and offer insights concerning mechanisms that underlie sex differences in the expression of psychopathology in adolescence.

Functional connectivity signatures of major depressive disorder: machine learning analysis of two multicenter neuroimaging studies.

The promise of machine learning has fueled the hope for developing diagnostic tools for psychiatry. Initial studies showed high accuracy for the identification of major depressive disorder (MDD) with resting-state connectivity, but progress has been hampered by the absence of large datasets. Here we used regular machine learning and advanced deep learning algorithms to differentiate patients with MDD from healthy controls and identify neurophysiological signatures of depression in two of the largest resting-state datasets for MDD. We obtained resting-state functional magnetic resonance imaging data from the REST-meta-MDD (N = 2338) and PsyMRI (N = 1039) consortia. Classification of functional connectivity matrices was done using support vector machines (SVM) and graph convolutional neural networks (GCN), and performance was evaluated using 5-fold cross-validation. Features were visualized using GCN-Explainer, an ablation study and univariate t-testing. The results showed a mean classification accuracy of 61% for MDD versus controls. Mean accuracy for classifying (non-)medicated subgroups was 62%. Sex classification accuracy was substantially better across datasets (73-81%). Visualization of the results showed that classifications were driven by stronger thalamic connections in both datasets, while nearly all other connections were weaker with small univariate effect sizes. These results suggest that whole brain resting-state connectivity is a reliable though poor biomarker for MDD, presumably due to disease heterogeneity as further supported by the higher accuracy for sex classification using the same methods. Deep learning revealed thalamic hyperconnectivity as a prominent neurophysiological signature of depression in both multicenter studies, which may guide the development of biomarkers in future studies.