Adolescence is a critical developmental period marked by significant physiological, psychological, and behavioural changes, many of which differ between the sexes. We aimed to investigate sex-specific associations between the plasma proteome and questionnaire-based mental health measures in adolescents. Liquid chromatography - tandem mass spectrometry proteomic analysis was used to measure the plasma proteome abundances in 197 adolescents (11-16 years old) from the WALNUTs cohort. Baseline analysis of sexual dimorphism revealed 76 proteins significantly differentially abundant between sexes, which were enriched in cell adhesion, collagen fibril organisation, and ossification pathways. Bioinformatic analysis revealed 37 proteins significantly associated with the total score of the Strengths and Difficulties Questionnaire (SDQ). Modelling the sex-specificity via interaction terms revealed 40 proteins with significant associations with SDQ in females and 1 protein in males. Plasma protein abundancies in males exhibited stronger correlations with SDQ externalizing subscale scores, while in females the associations with the internalizing score were more prominent, consistent with known behavioural sex differences. Female-associated proteins were enriched for haemostasis and complement pathways, while male-associated signals suggested distinct immune and cytoskeletal processes. These findings indicate that both shared and sex-specific plasma proteomic signatures are associated with SDQ scores in adolescents and that models adjusting only for sex may obscure sex-divergent biology. These exploratory results are hypothesis-generating and support the use of sex-aware proteomic analyses to refine biomarker discovery for adolescent mental health.

To evaluate the diagnostic accuracy of the System for the Assessment of Children and Adolescents (SENA) in identifying emotional disorders and suicidal thoughts and behaviors. SENA is a widely used Spanish screening tool for assessing emotional and behavioral symptoms in youth. 526 primary and secondary pupils aged 8-16 in several Spanish regions (a subset of the EmoChild Project, n=5,652 completed SENA), completed the SENA and underwent with clinical interviews using KSADS-COMP within 3 months. We screened potential participants by identifying possible positive and negative cases based on a T-score equivalent to the 75th percentile and served as a cut-off point for accessing a diagnostic interview. SENA's Emotional subscale showed a sensitivity (SN) of 78.5% in children and 80.4% in adolescents for detecting any emotional disorder, with specificities (SP) of 65.9% and 66.1%, and Area Under the ROC curve (AUC) of .74 in children and .73 in adolescents. AUCs were adequate for all subscales (0.72-0.93) other than obsessive compulsive disorder (AUC=.67). Specific subscales performed best: Generalized Anxiety Disorder (SN=100%, SP=88.4%) and Social Anxiety Disorder-SAD (SN = 91.4%, SP = 80.3%) in children, and SAD (SN=88.1%, SP=72.5%), Post-Traumatic Stress Disorder (SN=81.2%, SP=68.2%) and suicidal thoughts in adolescents (SN=84.5%, SP=73.8%). SENA is a valuable screening tool for educational and clinical settings, facilitating early intervention through a standardized and user-friendly assessment. Nevertheless, there is need for refined thresholds to enhance specificity and clinical alignment.

Cannabis-induced psychosis (CIP) carries a high risk of relapse. Research has shown that antipsychotic medications are effective in relapse prevention after first diagnosed CIP. Given that antipsychotics carry the potential for dose-related adverse effects, understanding the optimal dose is critical. Therefore, we conducted a dose-response analysis to evaluate the real-world effectiveness of oral antipsychotics in preventing relapse after CIP. We used data from linkage of administrative and health care registers from Sweden to identify all individuals with first diagnosis of CIP (ICD-10 F12.5). We modelled oral antipsychotic exposure (aripiprazole, clozapine, risperidone, olanzapine, quetiapine, antipsychotic polytherapy, other oral antipsychotics) as time-dependent using validated PRE2DUP-method. Dose-response association of antipsychotic exposure and outcome were examined across three predefined daily dose (DDD) categories (<0.6, 0.6-<1.4, ≥1.4) using within-individual models in a stratified Cox-regression analysis. The primary outcome was hospitalization for any psychotic episode, defined as schizophrenia-spectrum disorder (F20-F29) or substance-induced psychosis (F1x.5) as the main diagnosis. We identified 1,772 individuals aged 16-64 years with first-time CIP between 2006 and 2021. Antipsychotic polytherapy was associated with reduced risk of psychosis hospitalization across all dose ranges (HRs=0.54-0.65). Clozapine (0.6-<1.4 DDDs/day), olanzapine (≥0.6 DDDs/day), aripiprazole (0.6-<1.4 DDDs/day), risperidone (<0.6 DDDs/day), and other oral antipsychotics (0.6-<1.4 DDDs/day) were effective, while quetiapine showed no significant benefit. Findings indicate dose-dependent real-world effectiveness of antipsychotics in CIP, with most agents performing best at 0.6-<1.4 DDDs/day. These results support optimizing dosing of oral antipsychotic medications for relapse prevention after CIP to balance efficacy and adverse effects.

Pharmacological treatments for attention-deficit/hyperactivity disorder (ADHD) are efficacious and safe; however, substantial interindividual variability in treatment response persists, with many patients experiencing suboptimal outcomes or early discontinuation. Although genetic factors have been proposed as contributors to this variability, clinically actionable predictors remain elusive. Here, we present the first meta-analysis evaluating whether polygenic liability for ADHD and related psychiatric and behavioral-cognitive phenotypes is associated with clinically meaningful response to methylphenidate in 1000 ADHD cases from Norway, Brazil, and Spain assessed in real-world settings. Polygenic scores (PGS) for ADHD, autism, bipolar disorder, educational attainment, major depressive disorder, neuroticism, and schizophrenia were calculated separately for each cohort. Treatment response was assessed using evaluations of global clinical improvement and harmonized by categorizing individuals as responders or non-responders. Cohort-specific associations were combined using fixed-effects meta-analysis. No PGS showed a significant association with treatment response. Effect sizes were small, consistent across cohorts, and characterized by minimal between-study heterogeneity. Sensitivity analyses incorporating clinical and treatment-related covariates yielded convergent results. As the first meta-analytic evaluation of polygenic predictors evaluating clinically meaningful ADHD stimulant response, these findings delineate the current limits of PGS in pharmacogenomic applications. Rather than supporting immediate clinical utility, our results highlight key methodological and conceptual constraints, including limited sample sizes, heterogeneous outcome definitions, and the indirect nature of susceptibility-based PGS for predicting treatment response. By mapping these boundaries, this study provides a framework to recalibrate research priorities and guide the next generation of ADHD pharmacogenomic studies toward larger, harmonized, and more informative definitions of treatment response.