Neural and Behavioral Adaptations to Frontal Theta Neurofeedback Training: A Proof of Concept Study

Previous neurofeedback research has shown training-related frontal theta increases and performance improvements on some executive tasks in real feedback versus sham control groups. However, typical sham control groups receive false or non-contingent feedback, making it difficult to know whether observed differences between groups are associated with accurate contingent feedback or other cognitive mechanisms (motivation, control strategies, attentional engagement, fatigue, etc.). To address this question, we investigated differences between two frontal theta training groups, each receiving accurate contingent feedback, but with different top-down goals: (1) increase and (2) alternate increase/decrease. We hypothesized that the increase group would exhibit greater increases in frontal theta compared to the alternate group, which would exhibit lower frontal theta during down- versus up-modulation blocks over sessions. We also hypothesized that the alternate group would exhibit greater performance improvements on a Go-NoGo shooting task requiring alterations in behavioral activation and inhibition, as the alternate group would be trained with greater task specificity, suggesting that receiving accurate contingent feedback may be the more salient learning mechanism underlying frontal theta neurofeedback training gains. Thirty young healthy volunteers were randomly assigned to increase or alternate groups. Training consisted of an orientation session, five neurofeedback training sessions (six blocks of six 30-s trials of FCz theta modulation (4-7 Hz) separated by 10-s rest intervals), and six Go-NoGo testing sessions (four blocks of 90 trials in both Low and High time-stress conditions). Multilevel modeling revealed greater frontal theta increases in the alternate group over training sessions. Further, Go-NoGo task performance increased at a greater rate in the increase group (accuracy and reaction time, but not commission errors). Overall, these results reject our hypotheses and suggest that changes in frontal theta and performance outcomes were not explained by reinforcement learning afforded by accurate contingent feedback. We discuss our findings in terms of alternative conceptual and methodological considerations, as well as limitations of this research.