Abstract
Electroencephalographic neurofeedback (EEG-NF) enables individuals to self-regulate specific EEG features with real-time sensory feedback. Despite clinical and cognitive-enhancement applications, the mechanisms underlying the EEG modulation remain poorly understood. Particularly, alpha activity (8-12 Hz) upregulation may occur independently of the participants' volitional control. We previously reported spontaneous increases in alpha power during a passive neurofeedback-like task using pre-recorded EEG feedback. In the present study, we replicated this protocol while implementing an EEG-NF procedure using real-time alpha power. Thirty-two healthy adults observed a gray circle whose size was either fixed (control) or continuously updated at 1, 5 or 10 Hz (experimental conditions). Importantly, participants were not informed of the nature of the feedback and received no instructions for self-regulation. We evaluated the effects of (i) trial repetition, (ii) the presence (control vs. experimental conditions), (iii) the frequency (1, 5 or 10 Hz) and (iv) the source (online vs. offline alpha) of feedback update on EEG features classically targeted by EEG-NF. Importantly, we observed robust increases in alpha power over time independently of the presence, frequency and source of feedback update. The presence, frequency and source of feedback update did not influence the EEG features considered. These findings suggest that alpha EEG-NF upregulation may arise from spontaneous dynamics, such as time-on-task effects, rather than the hypothesized self-regulation mechanism. The assumption that observed alpha increases reflect successful neurofeedback learning is thus called into question. More broadly, the present study highlights the importance of including control conditions and accounting for non-specific effects when evaluating EEG-NF outcomes.