Probing Neurophysiological Processes Related to Self-Referential Processing to Predict Improvement in Adolescents With Depression Receiving Cognitive Behavioral Therapy

Abstract

Background

Cognitive behavioral therapy (CBT) is a gold-standard approach for treating major depressive disorder in adolescents. However, nearly half of adolescents receiving CBT do not improve. To personalize treatment, it is essential to identify objective markers that predict treatment responsiveness. To address this aim, we investigated neurophysiological processes related to self-referential processing that predicted CBT response among female adolescents with depression.

Methods

At baseline, female adolescents ages 13 to 18 years (N = 80) completed a comprehensive clinical assessment, and a self-referential encoding task was administered while electroencephalographic data were recorded. Baseline electroencephalographic data were utilized to identify oscillatory differences between healthy adolescents (n = 42) and adolescents with depression (n = 38). Following the baseline assessment, adolescents with depression received up to 12 weeks of CBT. Baseline differences in electroencephalographic oscillations between healthy adolescents and those with depression were used to guide CBT prediction analysis. Cluster-based event-related spectral perturbation analysis was used to probe theta and alpha event-related synchronization (ERS)/event-related desynchronization (ERD) response to negative and positive words.

Results

Baseline analyses showed that, relative to the healthy adolescents, adolescents with depression exhibited higher levels of frontal theta ERS and greater posterior alpha ERD. Multilevel modeling identified primary neural pretreatment predictors of treatment response: greater theta ERS in the right prefrontal cortex after the onset of negative words and lower alpha ERD in both the right prefrontal cortex and posterior cingulate cortex. ERS and ERD associations with treatment response remained significant, with baseline depressive and anxiety symptoms included as covariates in all analyses.

Conclusions

Consistent with prior research, results highlighted that relative to healthy adolescents, adolescents with depression are characterized by prominent theta synchronization and alpha desynchronization over the prefrontal cortex and posterior cingulate cortex, respectively. Cluster-based event-related spectral perturbation analysis also identified key mechanisms underlying depression-related self-referential processing that predicted improved symptoms during the course of CBT. Ultimately, a better characterization of the neural underpinnings of adolescent depression and its treatment may lead to more personalized interventions.