fNIRS-based neurofeedback targeting the right dorsolateral prefrontal cortex (dlPFC) in children at familial risk for substance use disorder and/or severe mental illness: a proof-of-concept study from eastern India

Source abstract

Background Children who have first-degree relatives diagnosed with substance use disorders and/or severe mental illnesses show altered prefrontal cortical development and an elevated risk of executive function impairment. fNIRS-based neurofeedback (NF) is a portable, child-tolerant technique for the modulation of cortical hemodynamics, with preliminary evidence of dlPFC-dependent cognitive benefits in adult populations. However, evidence evaluating its efficacy in pediatric at-risk cohorts from the Indian subcontinent is currently lacking. Methods This single-arm, uncontrolled proof-of-concept study evaluated the feasibility and tolerability of an eight-session, fNIRS-based NF protocol targeting two right dlPFC channels (Ch10, Ch15) in 11 children (aged 6–12 years), recruited from an inpatient psychiatric unit and an addiction treatment facility. Real-time oxygenated hemoglobin (HbO) feedback was provided via Turbo-Satori BCI. Session-wise HbO amplitude and Ch10↔Ch15 functional connectivity (FC) were extracted for all 11 children. Participants completed pre- and postintervention assessments of executive function. Exploratory Spearman correlations examined the associations between dlPFC HbO modulation and pre–post behavioral changes. Results All 11 children completed the protocol without adverse events. The ROI-averaged right dlPFC HbO amplitude showed a positive session-wise increase that did not reach conventional significance ( β = 0.00003, t (950) = 1.90, p = 0.058). Individual channel slopes were positive but non-significant after FDR correction. Inter-channel FC increased from session 1 ( r = 0.21) to session 4 ( r = 0.41) before declining by session 8 ( r = 0.18), with no significant linear trend. Behavioral measures showed numerically higher accuracy across tasks, and both the Raven’s CPM and the Stroop task showed shorter reaction times postintervention, but none of the comparisons survived Bonferroni correction. A moderate-to-strong correlation was observed between right dlPFC HbO change and Stroop incongruent accuracy improvement, but it was non-significant ( ρ = 0.679, p = 0.094), with no significant associations for the N -back or CPM outcomes. Conclusion This eight-session right dlPFC fNIRS–NF protocol in children with elevated familial risk was feasible and well tolerated. It showed a positive, but not statistically significant, increase in HbO amplitude. The connectivity between targeted channels and behavioral scores showed a positive correlation, but it was not statistically significant. A strong-to-medium correlation was observed between right dlPFC HbO change and Stroop incongruent accuracy scores. These findings provide preliminary effect-size estimates to inform the design of future sham-controlled trials in this underrepresented pediatric population.