Predicting Adolescent Mental Health Outcomes Across Cultures: A Machine Learning Approach

Machine learning-based predictive modeling of depressive symptoms in Chinese adolescents

BACKGROUND

The aim is to develop prediction models by lifestyles indicators as well as socioeconomic status to predict the risk of depressive symptoms in adolescents, and to rank and explain these predictors.

METHOD

A cross-sectional study was conducted in 32389 school students grade 4-12. A self-rating depression scale was used to define depressive symptoms (CES-D score ≥ 16), and lifestyle survey was used to investigate risk factors of depressive symptoms. Boruta-RF algorithm was used for feature selection and to rank variable importance. Random forest model was constructed to predict the risk of depressive symptom, and partial dependence plot (PDP) was used to explain the relationship between each variable and predicted outcome.

RESULTS

Boruta-RF algorithm showed that self-rated health, sleep duration, parental support for physical exercise, breakfast intake, screen time, skipping physical education classes, egg intake, grade, milk/soy product intake, and parental exercise habits were the top ten most important factors for depressive symptoms. The AUC of the random forest model was 0.829 (95% CI: 0.820 - 0.837), suggesting good accuracy for predicting depressive symptoms. Additionally, we demonstrated the nonlinear effect of each predictor for predicting risk of depressive symptoms by PDP.

CONCLUSIONS

The prediction model, using lifestyle indicators routinely collected in schools, can effectively screen for high-risk individuals needing further mental health evaluations, and facilitate early detection of depressive symptoms in adolescents. The study is limited by its cross-sectional design implying causality, use of CES-D for depressive symptoms rather than clinical diagnosis, and omission of neuroimaging biomarkers for improved accuracy.

Predicting children's emotional and behavioral difficulties at age five using pregnancy and newborn risk factors: Evidence from the UK Household Longitudinal Study

Childhood emotional and behavioral difficulties have a profound impact on later life outcomes, making it crucial to identify early-life risk factors that predict emotional and behavioral difficulties. However, much of the existing research has concentrated on diagnosing, rather than predicting, emotional and behavioral difficulties, and has often focused on adolescents rather than younger children. This study employs machine learning (ML) techniques to construct an interpretable predictive model using data from the UK Household Longitudinal Study, aiming to identify key risk factors that influence children's emotional and behavioral difficulties during childhood. We examined maternal habits during pregnancy and parent-reported data on birth, breastfeeding and regulatory problems during the newborn stage. Our findings highlighted lack of breastfeeding, low birthweight and maternal smoking during pregnancy as the three most significant predictors of emotional behavioral difficulties. Other important factors were related to infant regulatory problems. Heterogeneity analysis revealed gender differences in predictive power, with maternal smoking during pregnancy being a stronger predictor for boys, and the amount of fussing in infancy having a greater impact on girls. This study highlights the importance of comprehensive prenatal and postnatal care, advocates for early screening of emotional and behavioral difficulties, and calls for gender-specific approaches in assessing and addressing emotional and behavioral difficulties in children.

Longitudinal analysis of adolescents at high risk of depression: Prediction models

BACKGROUND

This study aimed to develop a machine-learning-based predictive model to identify adolescents at high risk of depression using longitudinal analysis to determine changes in risk factors over time.

METHODS

This longitudinal study used 4 years of data from the Korea Child and Youth Panel Survey (2018-2021). The classification of high-risk depression was the outcome variable, with predictors categorized into general characteristics and personal, family, and school factors. The machine learning algorithms used in the analysis included logistic regression, support vector machine, decision tree, random forest, and extreme gradient boosting.

RESULTS

Among the 1833 adolescents classified as having a low risk of depression during the initial survey year, 27.8 % were identified as being at a high risk of depression over the subsequent 3 years. The extreme gradient boosting algorithm yielded the best performance with an area under the curve of 0.9302. The key predictors identified included violent tendencies, self-esteem, sleep duration, gender, and coercive parenting style.

CONCLUSION

A machine-learning-based predictive model for identifying adolescents at high risk of depression was developed. These findings provide a foundation for early screening and the development of intervention programs and policies aimed at mitigating adolescent depression risk.

Application of machine learning in predicting adolescent Internet behavioral addiction

Objective

To explore the risk factors affecting adolescents' Internet addiction behavior and build a prediction model for adolescents' Internet addiction behavior based on machine learning algorithms.

Methods

A total of 4461 high school students in Chongqing were selected using stratified cluster sampling, and questionnaires were administered. Based on the presence of Internet addiction behavior, students were categorized into an Internet addiction group (n=1210) and a non-Internet addiction group (n=3115). Gender, age, residence type, and other data were compared between the groups, and independent risk factors for adolescent Internet addiction were analyzed using a logistic regression model. Six methods-multi-level perceptron, random forest, K-nearest neighbor, support vector machine, logistic regression, and extreme gradient boosting-were used to construct the model. The model's indicators under each algorithm were compared, evaluated with a confusion matrix, and the optimal model was selected.

Result

The proportion of male adolescents, urban household registration, and scores on the family function, planning, action, and cognitive subscales, along with psychoticism, introversion-extroversion, neuroticism, somatization, obsessive-compulsiveness, interpersonal sensitivity, depression, anxiety, hostility, paranoia, and psychosis, were significantly higher in the Internet addiction group than in the non-Internet addiction group (P < 0.05). No significant differences were found in age or only-child status (P > 0.05). Statistically significant variables were analyzed using a logistic regression model, revealing that gender, household registration type, and scores on planning, action, introversion-extroversion, psychoticism, neuroticism, cognitive, obsessive-compulsive, depression, and hostility scales are independent risk factors for adolescent Internet addiction. The area under the curve (AUC) for multi-level perceptron, random forest, K-nearest neighbor, support vector machine, logistic regression, and extreme gradient boosting models were 0.843, 0.817, 0.778, 0.846, 0.847, and 0.836, respectively, with extreme gradient boosting showing the best predictive performance among these models.

Conclusion

The detection rate of Internet addiction is higher in males than in females, and adolescents with impulsive, extroverted, psychotic, neurotic, obsessive, depressive, and hostile traits are more prone to developing Internet addiction. While the overall performance of the machine learning models for predicting adolescent Internet addiction is moderate, the extreme gradient boosting method outperforms others, effectively identifying risk factors and enabling targeted interventions.

Research on prediction model of adolescent suicide and self-injury behavior based on machine learning algorithm

Objective

To explore the risk factors that affect adolescents' suicidal and self-injurious behaviors and to construct a prediction model for adolescents' suicidal and self-injurious behaviors based on machine learning algorithms.

Methods

Stratified cluster sampling was used to select high school students in Chongqing, yielding 3,000 valid questionnaires. Based on whether students had engaged in suicide or self-injury, they were categorized into a suicide/self-injury group (n=78) and a non-suicide/self-injury group (n=2,922). Gender, age, insomnia, and mental illness data were compared between the two groups, and a logistic regression model was used to analyze independent risk factors for adolescent suicidal and self-injurious behavior. Six methods-multi-level perceptron, random forest, K-nearest neighbor, support vector machine, logistic regression, and extreme gradient boosting-were used to build predictive models. Various model indicators for suicidal and self-injurious behavior were compared across the six algorithms using a confusion matrix to identify the optimal model.

Result

In the self-injury and suicide groups, the proportions of male adolescents, late adolescence, insomnia, and mental illness were significantly higher than in the non-suicide and self-injury groups (p <0.05). Compared with the non-suicidal self-injury group, this group also showed significantly increased scores in cognitive subscales, impulsivity, psychoticism, introversion-extroversion, neuroticism, interpersonal sensitivity, depression, anxiety, hostility, terror, and paranoia (p <0.05). These statistically significant variables were analyzed in a logistic regression model, revealing that gender, impulsivity, psychoticism, neuroticism, interpersonal sensitivity, depression, and paranoia are independent risk factors for adolescent suicide and self-injury. The logistic regression model achieved the highest sensitivity and specificity in predicting adolescent suicide and self-injury behavior (0.9948 and 0.9981, respectively). Performance of the random forest, multi-level perceptron, and extreme gradient models was acceptable, while the K-nearest neighbor algorithm and support vector machine performed poorly.

Conclusion

The detection rate of suicidal and self-injurious behaviors is higher in women than in men. Adolescents displaying impulsiveness, psychoticism, neuroticism, interpersonal sensitivity, depression, and paranoia have a greater likelihood of engaging in such behaviors. The machine learning model for classifying and predicting adolescent suicide and self-injury risk effectively identifies these behaviors, enabling targeted interventions.

Subjective well-being of children with special educational needs: Longitudinal predictors using machine learning

Children with special educational needs (SEN) are a diverse group facing numerous challenges related to well-being and mental health. Understanding the predictors of well-being in this population requires the incorporation of diverse factors along with approaches that can uncover complexity in how these factors work together to influence well-being. We longitudinally predicted subjective well-being in a group of children with diverse special educational needs (N = 499; M = 8.4 ± 0.9 years). Thirty-two variables - ranging from demographics to various categories of life experiences - were used as predictors for both nonlinear machine learning and classical linear classifiers. Nonlinear machine learning classifiers exhibited much performance in predicting subjective well-being (F1 score = 0.72 to 0.84) compared to traditional linear classifiers. Overall, across all children, prior subjective well-being, numeracy, literacy skills, and interpersonal dimensions played important roles. However, clustering further identified four distinct clusters sharing important predictors: a 'socializer' cluster dominated by interpersonal functioning predictors, an 'analyzer' cluster emphasizing academic skills predictors, and two clusters with more diverse sets of important predictors. Our research highlights the multiple pathways toward well-being in children with SEN as uncovered by machine learning, with implications for understanding and supporting their well-being.

Exploring the Validity of Adolescent Responses to a Measure of Psychological Flexibility and Inflexibility

Validating measures of psychological flexibility (PF) and psychological inflexibility (PI) has occurred in multiple adult samples, but little research has validated PF and PI measures with adolescents. This manuscript describes two studies exploring the validity of responses to the Multidimensional Psychological Flexibility Inventory (MPFI) with two samples of adolescents. The first study used exploratory factor analyses on responses to the MPFI with a sample of 16-17-year-olds (N = 249). The results yielded a reduced and simplified measurement model that consisted of two general factors: one for PF and the other for PI. These exploratory findings were further investigated with confirmatory factor analyses in the second study, with a larger sample of 14-17-year-olds (N = 503). The results from the second study generally confirmed the factor model from the first study. Findings from both studies showed that scores derived from the reduced MPFI measurement model evidenced convergent and divergent validity with a variety of mental health criterion measures. Moreover, findings from the second study showed that PF and PI scores had differential predictive power on different concurrent mental health outcomes. This discussion highlights the implications of measuring PF and PI in adolescents, considers limitations of the present studies, and recommends next steps for research.

Exploring occupational well-being profiles, outcomes, and predictors among Chinese teachers: A mixed-methods approach using latent profile and decision tree analysis

Understanding the varied profiles of occupational well-being, their outcomes, and predictors is key to formulating effective strategies for enhancing teachers' occupational health and well-being. This study employed latent profile analysis (LPA) to identify distinct occupational well-being profiles and their outcomes among 366 Chinese teachers, and decision tree analysis to explore the factors predicting each profile. The results showed three occupational well-being profiles: burnout, engaged, and burnout-engaged. The "engaged" group exhibited normal ranges for depression and stress, along with mild anxiety. The "burnout" group demonstrated moderate depression and stress, coupled with severe anxiety. The "burnout-engaged" group was near the threshold of mild depression and moderate anxiety. The result of the decision tree model revealed that marital status, teaching experience, income, role as a class teacher, school type, and working hours significantly influenced these occupational well-being profiles. Specific combinations of variables were associated with each occupational well-being profile, offering a nuanced understanding of the risky and protective factors for teacher occupational well-being. By identifying distinct occupational well-being profiles among Chinese teachers and their outcomes, and elucidating the key predictors and their interrelations, this study provides insights into how to quickly screen for teachers in need of help at work, and how to design targeted interventions for different teachers.

Facial recognition and analysis: A machine learning-based pathway to corporate mental health management

Background

In modern workplaces, emotional well-being significantly impacts productivity, interpersonal relationships, and organizational stability. This study introduced an innovative facial-based emotion recognition system aimed at the real-time monitoring and management of employee emotional states.

Methods

Utilizing the RetinaFace model for facial detection, the Dlib algorithm for feature extraction, and VGG16 for micro-expression classification, the system constructed a 10-dimensional emotion feature vector. Emotional anomalies were identified using the K-Nearest Neighbors algorithm and assessed with a 3σ-based risk evaluation method.

Results

The system achieved high accuracy in emotion classification, as demonstrated by an empirical analysis, where VGG16 outperformed MobileNet and ResNet50 in key metrics such as accuracy, precision, and recall. Data augmentation techniques were employed to enhance the performance of the micro-expression classification model.

Conclusion

These techniques improved the across diverse emotional expressions, resulting in more accurate and robust emotion recognition. When deployed in a corporate environment, the system successfully monitored employees' emotional trends, identified potential risks, and provided actionable insights into early intervention. This study contributes to advancing corporate mental health management and lays the foundation for scalable emotion-based support systems in organizational settings.

Predicting adolescent psychopathology from early life factors: A machine learning tutorial

Objective

The successful implementation and interpretation of machine learning (ML) models in epidemiological studies can be challenging without an extensive programming background. We provide a didactic example of machine learning for risk prediction in this study by determining whether early life factors could be useful for predicting adolescent psychopathology.

Methods

In total, 9643 adolescents ages 9-10 from the Adolescent Brain and Cognitive Development (ABCD) Study were included in ML analysis to predict high Child Behavior Checklist (CBCL) scores (i.e., t-scores ≥ 60). ML models were constructed using a series of predictor combinations (prenatal, family history, sociodemographic) across 5 different algorithms. We assessed ML performance through sensitivity, specificity, F1-score, and area under the curve (AUC) metrics.

Results

A total of 1267 adolescents (13.1 %) were found to have high CBCL scores. The best performing algorithms were elastic net and gradient boosted trees. The best performing elastic net models included prenatal and family history factors (Sensitivity 0.654, Specificity 0.713; AUC 0.742, F1-score 0.401) and prenatal, family, history, and sociodemographic factors (Sensitivity 0.668, Specificity 0.704; AUC 0.745, F1-score 0.402). Across all 5 ML algorithms, family history factors (e.g., either parent had nervous breakdowns, trouble holding jobs/fights/police encounters, and counseling for mental issues) and sociodemographic covariates (e.g., maternal age, child's sex, caregiver income and caregiver education) tended to be better predictors of adolescent psychopathology. The most important prenatal predictors were unplanned pregnancy, birth complications, and pregnancy complications.

Conclusion

Our results suggest that inclusion of prenatal, family history, and sociodemographic factors in ML models can generate moderately accurate predictions of adolescent psychopathology. Issues associated with model overfitting, hyperparameter tuning, and system seed setting should be considered throughout model training, testing, and validation. Future early risk predictions models may improve with the inclusion of additional relevant covariates.

Puberty and NeuroDevelopment in adolescents (PANDA): a study protocol

BACKGROUND

Biopsychosocial changes during adolescence are thought to confer risk for emotion dysregulation, and in particular, anxiety disorders. However, there are substantial gaps in our knowledge about the biological mechanisms underlying anxiety during adolescence, and whether this contributes to the higher prevalence in females. The Puberty and NeuroDevelopment in Adolescents (PANDA) study aims to examine links between biological (sex hormones, cortisol) and social environmental factors and brain function during adolescence, with a focus on key processes (emotion regulation, fear learning) identified as relevant for the development of anxiety disorders.

METHODS

PANDA is a cross-sectional study with an observational design that aims to recruit a total of 175 adolescents aged 11-16 (majority female) and their parents/guardians, from the community. Brain function will be examined using magnetic resonance imaging (MRI), including functional MRI tasks of emotion regulation and fear learning. Hormones will be measured from hair (i.e., cortisol) and weekly saliva samples (i.e., oestradiol, progesterone, five across a month in females). Questionnaires and semi-structured interviews will be used to assess mental health and social environmental factors such as parenting and adverse childhood experiences. An online study of 113 adolescents was also incorporated during the COVID-19 pandemic as a questionnaire-only sub-study.

DISCUSSION

Strengths of this study include the collection of multiple saliva samples to assess variability in hormone levels, examination of the timing of adverse childhood experiences, inclusion of both maternal and paternal parental factors, exploration of mechanisms through the examination of brain structure and function, and multi-method, multi-informant collection of mental health symptoms. This study addresses important gaps in the literature and will enhance knowledge of the biological and environmental contributors to emotion dysregulation and anxiety in adolescents.

A panel network approach of internalizing and externalizing problems in early childhood: Evidence from American and Chinese preschoolers

Internalizing and externalizing problems tend to co-occur beginning in early childhood. However, the dynamic interplay of symptom-level internalizing and externalizing problems that may drive their co-occurrence is poorly understood. Within the frameworks of the Network Approaches to Psychopathology and the Developmental Cascade Perspective, this study used a panel network approach to examine how symptoms of internalizing and externalizing problems are related in early childhood both concurrently and longitudinally and whether the pattern may differ in American (N = 1,202) and Chinese (N = 180) preschoolers. Internalizing and externalizing problems were rated by mothers in two waves. Results from cross-sectional networks showed that the bridge symptoms underlying the co-occurrence of internalizing and externalizing problems were largely consistent in American and Chinese preschoolers (e.g., withdrawal, aggressive behavior, anxiety and depressive moods). Results from cross-lagged panel networks further showed that the co-occurrence was manifested by unidirectional relations from internalizing to subsequent externalizing symptoms in both American and Chinese preschoolers. The findings contribute needed cross-cultural evidence to better understand the co-occurrence of internalizing and externalizing problems and highlight the temporal heterogeneity of the symptom networks of internalizing and externalizing problems in early childhood.

Environmental and neurodevelopmental contributors to youth mental illness

While a myriad of factors likely contribute to the development of mental illness in young people, the social environment (including early adverse experiences) in concert with neurodevelopmental alterations is undeniably important. A number of influential theories make predictions about how and why neurodevelopmental alterations may mediate or moderate the effects of the social environment on the emergence of mental illness. Here, we discuss current evidence supporting each of these theories. Although this area of research is rapidly growing, the body of evidence is still relatively limited. However, there exist some consistent findings, including increased striatal reactivity during positive affective processing and larger hippocampal volumes being associated with increased vulnerability or susceptibility to the effects of social environments on internalizing symptoms. Limited longitudinal work has investigated neurodevelopmental mechanisms linking the social environment with mental health. Drawing from human research and insights from animal studies, we propose an integrated mediation-moderation model and outline future research directions to advance the field.

The Dynamic Interplay Between Puberty and Structural Brain Development as a Predictor of Mental Health Difficulties in Adolescence: A Systematic Review

Puberty is a time of intense reorganization of brain structure and a high-risk period for the onset of mental health problems, with variations in pubertal timing and tempo intensifying this risk. We conducted 2 systematic reviews of articles published up to February 1, 2024, focusing on 1) the role of brain structure in the relationship between puberty and mental health, and 2) precision psychiatry research evaluating the utility of puberty in making individualized predictions of mental health outcomes in young people. The first review provides inconsistent evidence about whether and how pubertal and psychopathological processes may interact in relation to brain development. While most studies found an association between early puberty and mental health difficulties in adolescents, evidence on whether brain structure mediates this relationship is mixed. The pituitary gland was found to be associated with mental health status during this time, possibly through its central role in regulating puberty and its function in the hypothalamic-pituitary-gonadal and hypothalamic-pituitary-adrenal axes. In the second review, the design of studies that have explored puberty in predictive models did not allow for a quantification of its predictive power. However, when puberty was evaluated through physically observable characteristics rather than hormonal measures, it was more commonly identified as a predictor of depression, anxiety, and suicidality in adolescence. Social processes may be more relevant than biological ones to the link between puberty and mental health problems and represent an important target for educational strategies.

Effects of Individual Research Practices on fNIRS Signal Quality and Latent Characteristics

Functional near-infrared spectroscopy (fNIRS) is an increasingly popular tool for cross-cultural neuroimaging studies. However, the reproducibility and comparability of fNIRS studies is still an open issue in the scientific community. The paucity of experimental practices and the lack of clear guidelines regarding fNIRS use contribute to undermining the reproducibility of results. For this reason, much effort is now directed at assessing the impact of heterogeneous experimental practices in creating divergent fNIRS results. The current work aims to assess differences in fNIRS signal quality in data collected by two different labs in two different cohorts: Singapore (N=74) and Italy (N=84). Random segments of 20s were extracted from each channel in each participant's NIRScap and 1280 deep features were obtained using a deep learning model trained to classify the quality of fNIRS data. Two datasets were generated: ALL dataset (segments with bad and good data quality) and GOOD dataset (segments with good quality). Each dataset was divided into train and test partitions, which were used to train and evaluate the performance of a Support Vector Machine (SVM) model in classifying the cohorts from signal quality features. Results showed that the SG cohort had significantly higher occurrences of bad signal quality in the majority of the fNIRS channels. Moreover, the SVM correctly classified the cohorts when using the ALL dataset. However, the performance dropped almost completely (except for five channels) when the SVM had to classify the cohorts using data from the GOOD dataset. These results suggest that fNIRS raw data obtained by different labs might possess different levels of quality as well as different latent characteristics beyond quality per se. The current study highlights the importance of defining clear guidelines in the conduction of fNIRS experiments in the reporting of data quality in fNIRS manuscripts.

A review of information sources and analysis methods for data driven decision aids in child and adolescent mental health services

OBJECTIVE

Clinical data analysis relies on effective methods and appropriate data. Recognizing distinctive clinical services and service functions may lead to improved decision-making. Our first objective is to categorize analytical methods, data sources, and algorithms used in current research on information analysis and decision support in child and adolescent mental health services (CAMHS). Our secondary objective is to identify the potential for data analysis in different clinical services and functions in which data-driven decision aids can be useful.

MATERIALS AND METHODS

We searched related studies in Science Direct and PubMed from 2018 to 2023(Jun), and also in ACM (Association for Computing Machinery) Digital Library, DBLP (Database systems and Logic Programming), and Google Scholar from 2018 to 2021. We have reviewed 39 studies and extracted types of analytical methods, information content, and information sources for decision-making.

RESULTS

In order to compare studies, we developed a framework for characterizing health services, functions, and data features. Most data sets in reviewed studies were small, with a median of 1,176 patients and 46,503 record entries. Structured data was used for all studies except two that used textual clinical notes. Most studies used supervised classification and regression. Service and situation-specific data analysis dominated among the studies, only two studies used temporal, or process features from the patient data. This paper presents and summarizes the utility, but not quality, of the studies according to the care situations and care providers to identify service functions where data-driven decision aids may be relevant.

CONCLUSIONS

Frameworks identifying services, functions, and care processes are necessary for characterizing and comparing electronic health record (EHR) data analysis studies. The majority of studies use features related to diagnosis and assessment and correspondingly have utility for intervention planning and follow-up. Profiling the disease severity of referred patients is also an important application area.

Does Grit Compensate for Family Background Disadvantage in Predicting Mental Health Difficulties? A Longitudinal Study of Chinese Migrant and Urban Children

The significant population of Chinese rural-to-urban migrant children has sparked considerable domestic and international concern regarding their disadvantaged family circumstances and their escalating prevalence of internalizing and externalizing problems. Derived from the resource substitution hypothesis, non-cognitive factors such as personality traits may act as "substitution" resources for educational outcomes of children from less privileged families. Yet, the compensatory role of personality traits as substitution resources in children's mental health has received limited attention, including that of migrant children. This study examined the interplay of trait-like grit and family SES on emotional and conduct problems among Chinese migrant and urban children. The current sample consisted of 770 migrant children (Mage = 10.45 and SDage = 0.68 years; 38.4% girls) and their 222 urban counterparts (Mage = 10.34 and SDage = 0.46 years; 45.5% girls). Moderated polynomial regressions with response surface analysis on a two-wave data with an interval of over six months showed that grit served as a "substitution" resource for the less socioeconomically advantaged children. The compensatory effect of perseverance of effort on urban children's emotional problems and that of consistency of interest on migrant children's conduct problems were visualized. Moreover, the two compensatory effects were found to be robust and unique, even after children's effortful control, a grit-related construct, was taken into account. These findings not only support the resource substitution hypothesis, but also underscore the protective role that grit plays in children under less privileged environments.

Brain network connectivity during peer evaluation in adolescent females: Associations with age, pubertal hormones, timing, and status

Despite copious data linking brain function with changes to social behavior and mental health, little is known about how puberty relates to brain functioning. We investigated the specificity of brain network connectivity associations with pubertal indices and age to inform neurodevelopmental models of adolescence. We examined how brain network connectivity during a peer evaluation fMRI task related to pubertal hormones (dehydroepiandrosterone and testosterone), pubertal timing and status, and age. Participants were 99 adolescents assigned female at birth aged 9-15 (M = 12.38, SD = 1.81) enriched for the presence of internalizing symptoms. Multivariate analysis revealed that within Salience, between Frontoparietal - Reward and Cinguloopercular - Reward network connectivity were associated with all measures of pubertal development and age. Specifically, Salience connectivity linked with age, pubertal hormones, and status, but not timing. In contrast, Frontoparietal - Reward connectivity was only associated with hormones. Finally, Cinguloopercular - Reward connectivity related to age and pubertal status, but not hormones or timing. These results provide evidence that the salience processing underlying peer evaluation is jointly influenced by various indices of puberty and age, while coordination between cognitive control and reward circuitry is related to pubertal hormones, pubertal status, and age in unique ways.

Unveiling Adolescent Suicidality: Holistic Analysis of Protective and Risk Factors Using Multiple Machine Learning Algorithms

Adolescent suicide attempts are on the rise, presenting a significant public health concern. Recent research aimed at improving risk assessment for adolescent suicide attempts has turned to machine learning. But no studies to date have examined the performance of stacked ensemble algorithms, which are more suitable for low-prevalence conditions. The existing machine learning-based research also lacks population-representative samples, overlooks protective factors and their interplay with risk factors, and neglects established theories on suicidal behavior in favor of purely algorithmic risk estimation. The present study overcomes these shortcomings by comparing the performance of a stacked ensemble algorithm with a diverse set of algorithms, performing a holistic item analysis to identify both risk and protective factors on a comprehensive data, and addressing the compatibility of these factors with two competing theories of suicide, namely, The Interpersonal Theory of Suicide and The Strain Theory of Suicide. A population-representative dataset of 173,664 Norwegian adolescents aged 13 to 18 years (mean = 15.14, SD = 1.58, 50.5% female) with a 4.65% rate of reported suicide attempt during the past 12 months was analyzed. Five machine learning algorithms were trained for suicide attempt risk assessment. The stacked ensemble model significantly outperformed other algorithms, achieving equal sensitivity and a specificity of 90.1%, AUC of 96.4%, and AUCPR of 67.5%. All algorithms found recent self-harm to be the most important indicator of adolescent suicide attempt. Exploratory factor analysis suggested five additional risk domains, which we labeled internalizing problems, sleep disturbance, disordered eating, lack of optimism regarding future education and career, and victimization. The identified factors provided stronger support for The Interpersonal Theory of Suicide than for The Strain Theory of Suicide. An enhancement to The Interpersonal Theory based on the risk and protective factors identified by holistic item analysis is presented.