Plasma proteomics discovery of mental health risk biomarkers in adolescents

Adolescence is a period of significant biological, psychosocial, cognitive, and emotional changes, making it a critical window for both cognitive improvement and the onset of mental disorders. Many mental health conditions, including ADHD, phobias, OCD, eating disorders, substance abuse, mood disorders, and social anxiety, begin before adulthood, often peaking around age 14. Early identification and intervention are crucial to improve outcomes and reduce the socioeconomic impact of these disorders. Globally, 10–20% of adolescents experience mental health conditions, but most remain undiagnosed and untreated due to social stigma, perceptions of mental health care needs, and limited resources. Current diagnostic methods rely on ICD and DSM classifications, but objective biomarkers could significantly improve early identification and reduce misdiagnosis or overdiagnosis. This study explores the use of plasma proteomics to identify potential biomarkers for mental health risk in adolescents.

Existing research indicates alterations in plasma proteins are associated with various mental disorders such as depression (MDD), schizophrenia (SCZ), psychotic disorders, and bipolar disorders. Commonly affected pathways include the complement cascade and interleukin signaling. Studies have shown the predictive value of self-reported SDQ scores in clinical diagnostics, particularly when combined with parent or teacher reports. The SDQ is a reliable and valid tool for assessing behavioral problems, and although not a diagnostic instrument itself, it effectively screens for adolescents at high risk of mental disorders. However, research on plasma proteomic biomarkers in adolescent mental health is still a relatively new field.

This study utilized a subsample of 91 adolescents (11–16 years) from the Spanish WALNUTS cohort study. Participants completed the SDQ, and plasma samples were collected and stored at −80°C. In 2021, samples underwent protein depletion and proteomic analysis using liquid chromatography–tandem mass spectrometry (LC-MS/MS) at the Turku Proteomics Facility. Linear modeling with DeqMS was used to investigate associations between SDQ scores and protein abundances. Bioinformatic analyses (STRINGdb, ReactomePA, IPA) characterized biological processes and pathways associated with differentially abundant proteins. A QLattice algorithm generated predictive models to identify proteins that best differentiated between low and raised SDQ score groups. Fivefold cross-validation with logistic regression ensured model generalizability. Potential confounding factors, such as sex, age, parental education, media consumption, social media engagement, drug and alcohol use, and physical activity, were considered in the analyses.

Mass spectrometry identified 1485 proteins; after removing contaminants and proteins detected in less than 80% of samples, 983 proteins were analyzed. 58 proteins showed significant associations with SDQ scores. Enriched pathways included immune responses, blood coagulation, complement cascade, neuronal degeneration, and neurogenesis. Blood coagulation and immune responses were the most prominently altered pathways. Proteins like clusterin, vitronectin, complement C2, and coagulation factor XI showed alterations, consistent with previous studies in mental disorders. QLattice analysis generated five predictive models containing eleven proteins, four of which have previously reported connections to the central nervous system (CNS), neurogenesis, or mental health. These included amyloid beta precursor-like protein 1 (APLP1), calcium/calmodulin-dependent protein kinase II beta (CAMK2B), reticulon 4 (RTN4), and cadherin 11 (CDH11). The study also noted minor differences in self-reported puberty changes between low and raised SDQ groups, but sex was included as a confounding factor in linear models. No other significant differences were found in socio-economic, sociodemographic, or other factors analyzed.

This study's findings align with the understanding of altered immune responses and blood coagulation in the pathophysiology of mental disorders. The identified proteins, particularly those with CNS connections (APLP1, CAMK2B, RTN4, CDH11), warrant further investigation as potential biomarkers for adolescent mental health risk. The negative association of APLP1 and the positive association of CAMK2B with SDQ scores are noteworthy, given their roles in brain development and neuronal plasticity. The involvement of RTN4, implicated in neurodegeneration, and CDH11, crucial for dendrite formation and synaptogenesis, also highlights the potential impact on neurodevelopment. The clustering of these proteins suggests a potential interconnected network involved in adolescent mental health. The study's results support the concept of early changes in coagulation and complement cascades in the predisposition to mental health issues.

This exploratory study identified several plasma protein alterations associated with SDQ scores in adolescents, offering potential protein-based susceptibility biomarker candidates for mental health dysfunction. Further research in larger cohorts, with longitudinal follow-up, is needed to validate these candidates and assess their association with the transition to clinical mental disorders. Future work could focus on investigating the identified protein network and its role in brain development and mental health during adolescence.

The primary limitation is the relatively small sample size compared to the number of identified proteins. While linear modeling and group comparisons enhanced statistical power, larger cohorts are needed for robust validation. The use of non-fasting plasma samples might have influenced protein concentrations. Further studies are required to confirm these findings and investigate the clinical utility of these potential biomarkers.