Plasma SERPINA3 and its clinical correlates in adolescent major depressive disorder: an exploratory case-control...

Major depressive disorder (MDD) in adolescents lacks objective peripheral biomarkers that may assist diagnosis and biological stratification. Serine protease inhibitor A3 (SERPINA3) is an inflammation-related acute-phase protein, but its circulating levels and clinical relevance in adolescent MDD remain unclear.

In this exploratory case-control study, 141 adolescents with MDD and 76 healthy controls were enrolled. Plasma SERPINA3 concentrations were measured by enzyme-linked immunosorbent assay. Within the MDD cohort, participants were stratified at the median SERPINA3 concentration (380 pg/mL) for exploratory subgroup analyses. Associations between SERPINA3 and laboratory variables were examined using exploratory correlation and regression analyses. P values for the exploratory laboratory comparisons in Table 2 were adjusted using the Benjamini-Hochberg false discovery rate (FDR) procedure. Receiver operating characteristic (ROC) analysis was used to assess the discriminatory performance of SERPINA3, with internal validation by 2000 bootstrap resamples and stratified 10-fold cross-validation.

Plasma SERPINA3 concentrations were higher in adolescents with MDD than in healthy controls (P < 0.0001). Several laboratory variables showed nominal between-group differences in unadjusted analyses, but none remained statistically significant after FDR correction (all q ≥ 0.242). In exploratory multivariable analyses, Glucose, RDW-CV, and indirect bilirubin were associated with continuous SERPINA3 concentration. In the full-sample ROC analysis, the AUC was 0.8433 and the optimal cutoff was 309 pg/mL. Internal validation yielded an optimism-corrected AUC of 0.8421 in bootstrap resampling and a mean test AUC of 0.8393 in stratified 10-fold cross-validation.

In this exploratory case-control study, plasma SERPINA3 concentrations were higher in adolescents with MDD than in healthy controls. However, the observed laboratory associations were not robust after correction for multiple testing, and the discriminatory performance should be interpreted as preliminary and internally validated only within this cohort. These findings are hypothesis-generating and require confirmation in larger, longitudinal, and independent cohorts.

Not applicable.