Multi-level treatment outcome evaluation in adolescents with autism spectrum disorder

Abstract

Background: Aberrant resting state electroencephalography (rsEEG) is a well-established indicator of psychopathological brain activity in clinical disorders. In Autism Spectrum Disorder (ASD), a substantial body of research reports reduced Alpha activity in the electrocortical resting state of affected individuals. However, effective interventions based on neurophysiological patterns and objective biological markers of treatment outcome remain scarce.

Methods: In this randomized controlled trial, the primary objective was to examine rsEEG changes in adolescents with ASD following 24 sessions of slow cortical potential neurofeedback training (n = 21) compared to a treatment-as-usual control group (n = 20). A repeated-measures analysis of variance was used to assess group differences over time. Additionally, Pearson correlation analyses were conducted to exploratorily investigate associations between rsEEG measures and clinical psychopathology and affective well-being, as assessed via parental and self-report questionnaires at baseline and post-intervention.

Results: Analyses revealed significant differences in the development of rsEEG between the intervention groups: while Alpha activity increased in the experimental neurofeedback group, it decreased in the control group, demonstrating an opposite trend. Exploratory analyses showed that Delta activity decreased in both groups, with a more pronounced decrease in the experimental group. Correlational analyses revealed significant associations between subjective-psychological and electrocortical levels: lower alpha power at baseline was related to greater severity of ASD symptoms, while both lower alpha and higher delta power were associated with greater negative affect at baseline. Increases in alpha power after NF-training were linked with enhanced positive affect, whereas reductions in delta power corresponded to decreases in negative affect.

Conclusions: This study provides insights into changes in resting-state neural activity before and after clinical interventions alongside clinical-psychological assessment, overcoming single-level assessments and emphasizing the need for multi-level outcome measures for a more comprehensive treatment evaluation.

Clinical trial registration: DRKS00012339.

Keywords: Alpha brain activity; Autism spectrum disorder; Brain regulation training; Neurofeedback; Resting state EEG.

Fig. 1

At the start of each trial, a triangle indicated the required SCP polarity: upward for negative shifts (increased cortical activation), downward for positive shifts (decreased cortical activation / inhibition). Following baseline recording, SCP activity was displayed in real time as a moving object (e.g., a fish or moon) on the participant’s screen, with upward movement reflecting increased and downward movement decreased cortical activation. Participants were instructed to volitionally control the object’s movement by producing SCP shifts in the indicated direction. Successful regulation—defined as a shift in the required polarity lasting at least 2 consecutive seconds within the final 4 s of a trial—was rewarded with a sun symbol and verbal encouragement from the trainers. Each session began with an eye movement calibration task to enable real-time artifact correction and minimize eye movement influences

Fig. 2

Increases in Alpha band from before (T1, red bars) to after intervention (T2, green bars) in the NF - EG; Decreases in Alpha band from before (T1, red bars) to after intervention (T2, green bars) in the TAU - CG

Fig. 3

Changes in Delta band from before (T1, red bars) to after intervention (T2, green bars) in the NF - EG and the TAU– CG for the EO Condition on the left side (a) and for the EC Condition on the right side (b)

Fig. 4

Correlation between SRS - Social Communication subscale at baseline and EO - rs Alpha Power (µV²) at baseline. The higher the SRS score (ASD psychopathology) at baseline, the lower the Alpha power at baseline

Fig. 5

Correlation between Anxiety Baseline (PANAS) and EO - Baseline Alpha Power (µV²). The higher the PANAS - Anxiety score at baseline, the lower the Alpha power at baseline

Fig. 6

Correlation between Changes in Alpha and Delta power (post minus pre) and changes in PANAS– affects (post minus pre). a The more Alpha power is increasing from before to after intervention, the more the positive affect Inspired is also increasing. b The more Delta power is decreasing from before to after intervention, the more the negative affect Irritable is decreasing