Abastract 

• Trait anxiety is an individual disposition marked by heightened anticipation of potential threats under uncertainty. It has been associated with allostatic load, the cumulative physiological cost of chronic stress, suggesting that enduring anxiety vulnerability emerges from brain–body interactions across multiple scales. Yet these domains have largely been examined separately and the temporal dynamics of brain activity remain underexplored. Using data from healthy young adults (LEMON cohort, N = 120), we implemented a graph-attention framework integrating low-frequency (slow-4 and slow-5) fMRI dynamics, structural connectivity and systemic biomarkers through cross-modal attention to predict individual trait anxiety outcome (The State-Trait Anxiety Inventory). Temporal modeling significantly enhanced prediction compared with static or amplitude-based features, highlighting the importance of time-resolved neural information. Model-derived importance mapping identified the limbic and visual systems as core predictive networks. Dynamic functional connectivity revealed that higher trait anxiety was associated with longer occupancy of states marked by strong limbic–default-mode–frontoparietal coupling and shorter occupancy of visually decoupled states. Metabolic and immune markers further contributed to prediction and transcriptomic enrichment linked these networks to neurodevelopmental and synaptic signaling pathways. Together, these findings delineate a temporally dynamic brain and body architecture underlying stable anxiety vulnerability.

Abastract 

• Youth depression is a major public health concern with long-term consequences and strong ties to suicidal behavior. Early detection is crucial for effective intervention, yet reliable predictive methods remain limited. To address this, we combined genetic predisposition with white matter microstructure data. Using Tract-Weighted Imaging (TWI) from diffusion MRI in 9–10-year-olds from the Adolescent Brain Cognitive Development (ABCD) cohort, we trained a 3D convolutional neural network (3DCNN) pretrained with Polygenic Risk Scores (PGS) for depression. This model outperformed from-scratch and traditional machine learning approaches, achieving an AUC of 0.62 in identifying Major Depressive Disorder (MDD) at baseline. Explainable AI revealed reduced fractional anisotropy in commissural tracts as key features. The model demonstrated robust zero-shot prediction of MDD and suicidal behavior two years later (MDD 2y: AUC = 0.61; MDD + Suicide Attempt: AUC = 0.66). Transfer learning validated its performance in an ethnically diverse cohort (AUC = 0.67). These findings highlight the potential of combining genetic and white matter data for early youth depression detection.  

Abstract 

• Episodic memory is a cornerstone of cognitive development and is essential for adaptive functioning across the lifespan. However, its development can be profoundly impacted by early-life adversity, making it critical to elucidate how such experiences influence brain connectivity and memory processes. In this study, we examined the effects of early adversity on hippocampal networks and episodic memory in Korean participants meeting inclusion criteria (N=152, age: 10-18, W=81), combining multi-modal MRI, genetic, environmental, and psychological data. Behaviorally, participants unexposed to early adversity demonstrated age-related improvements in episodic memory, a pattern absent in the adversity-exposed group. Neuroimaging findings revealed that adversity-exposed participants had reduced global structural connectivity efficiency, disrupted hippocampal-cingulate pathways in the brain networks. These results highlight the long-lasting impacts of early adversity on brain networks and memory development, emphasizing the significance of environmental influences on cognitive trajectories. This work underscores the value of integrating diverse disciplines to better understand and address the complexities of neurodevelopmental disruptions.