Frequency-Specific Operant Learning in Neurofeedback Reveals Distinct Cortical Mechanisms: Evidence from Double-Blind ERSP and ERP Dissociations
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
Neurofeedback reliably alters EEG activity, but the cortical mechanism by which reward-contingent feedback shapes oscillatory dynamics remains unresolved. In particular, no study has examined reward-locked event-related spectral perturbations (ERSP) under double-blind, active-placebo-controlled conditions.
Methods
Forty participants underwent five training sessions and a 3–5 week retention session of single-channel EEG biofeedback (C3 SMR 12–15 Hz, n = 8; C3 Beta 15–18 Hz, n = 8; C4 SMR 12–15 Hz, n = 8; active-placebo sham, n = 16), with concurrent 64-channel EEG recording. ERSP was computed from reward-locked epochs (approximately 600–700 trials per session) using Morlet wavelets (3–40 Hz) across four sessions.
Results
Active groups produced frequency-specific event-related desynchronization (ERD) in the rewarded band (pooled Active vs Sham d = −1.23, p adj = 0.001; C3 Beta and C4 SMR FDR-significant, | d | ≥ 1.12; C3 SMR trended, d = −0.80, p adj = 0.081), absent in sham. A double dissociation emerged at C3: beta training produced the strongest ERD ( d = −2.38), whereas C3 SMR training produced the largest P2 suppression ( d = −1.33, BF 01 = 0.10; smaller P2 at the trained electrode). Only SMR groups showed lasting plasticity, with increased eyes-closed alpha at follow-up (C3 SMR d = 0.97; C4 SMR d = 0.78) and significant across-session accumulation ( β = 1.44, p = 0.004). ERD magnitude predicted long-term resting-state change ( r = 0.54, p = 0.009) but not within-session shifts ( r = −0.09, p = 0.67), dissociating transient from consolidating effects. An ICA-based sensitivity analysis confirmed convergence of all primary findings.
Conclusions
Neurofeedback engages frequency-specific, contingency-dependent cortical mechanisms. The ERD–P2 dissociation suggests that beta and SMR training recruit distinct circuits (C3 Beta: local cortical ERD with preserved P2; C3 SMR: putative thalamocortical relay ERD with suppressed P2 at the trained site) with different capacities for consolidation. These findings establish a multi-timescale model in which immediate reward-locked desynchronization drives durable plasticity only when supported by deeper circuit dynamics.
