Content uploaded by Samy Chikhi
Author content
All content in this area was uploaded by Samy Chikhi on Sep 26, 2024
Content may be subject to copyright.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 1
Effects of one session of theta or high alpha neurofeedback on EEG activity and working
memory
Samy Chikhi1,4*, Nadine Matton2, 3, Marie Sanna1, Sophie Blanchet1
1 Université Paris Cité, Laboratoire Mémoire, Cerveau et Cognition, F-92100 Boulogne-
Billancourt, France.
2 CLLE - Cognition, Langues, Langage, Ergonomie, Université de Toulouse, Toulouse, France.
3 Fédération ENAC ISAE-SUPAERO ONERA, Université de Toulouse, France.
4 Integrative Neuroscience and Cognition Center, Université Paris Cité, F-75006, Paris, France.
Author Note
Correspondence concerning this article should be addressed to Samy Chikhi, Integrative
Neuroscience and Cognition Center, UMR 8002, Université Paris Cité, 45 rue des Saints-Pères, 75006
Paris, France. Email: sychikhi@gmail.com
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 2
Abstract
Neurofeedback techniques provide participants immediate feedback on neuronal signals, enabling them
to modulate their brain activity. This technique holds promise in unveiling brain-behavior relationship,
and offers opportunities for neuroenhancement. Establishing causal relationships between modulated
brain activity and behavioral improvements requires rigorous experimental designs, including appropriate
control groups and large samples. Our primary objective was to examine whether a single neurofeedback
session, designed to enhance working memory through the modulation of theta or high-alpha
frequencies, elicits specific changes in electrophysiological and cognitive outcomes. Additionally, we
explored predictors of successful neuromodulation. One hundred and one healthy adults were assigned
to groups trained to increase frontal theta, parietal high alpha, or random frequencies (active control
group). We measured resting-state EEG, working memory performance, and self-reported psychological
states before and after one neurofeedback session. Although our analyses revealed improvements in
electrophysiological and behavioral outcomes, these gains were not specific to the experimental groups.
An increase in the frequency targeted by the training has been observed for the theta and high alpha
groups, but training aimed at increasing randomly selected frequencies appears to induce more
generalized neuromodulation compared to targeting a specific frequency. Among all the predictors of
neuromodulation examined, resting theta and high alpha amplitudes predicted specifically the increase
of those frequencies during the training. These results highlight the challenge of integrating a control
group based on enhancing randomly selected frequency bands and suggest potential avenues for
optimizing interventions (e.g., by including a control group trained in both up- and down-regulation).
Keywords: neurofeedback, working memory, theta, high alpha, active control group
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 3
Introduction
The brain consists of numerous neurons forming intricate networks. Within these networks,
information transmission occurs through rhythmic electrical activity known as neural oscillations (Buzsaki
& Draguhn, 2004; Engel et al., 2001; Weisz & Keil, 2022). Neural oscillations play a crucial role in various
cognitive processes (Başar et al., 2001), including sensory processing (Haegens & Golumbic, 2018),
attention (Clayton et al., 2015) and working memory (Chikhi et al., 2022; Pavlov & Kotchoubey, 2022).
Given the foundational role of neural oscillations in cognitive processes, modulating specific brain activity
parameters holds promise for improving cognition (Chiasson et al., 2023; Gruzelier, 2014; Viviani & Vallesi,
2021) and elucidating the causal relationship between these parameters and cognitive processes
(Kvamme et al., 2022a; Ramot & Martin, 2022).
Neurofeedback is a brain-computer interface that trains participants to voluntarily modulate a specific
parameter of their brain activity through real-time feedback on their activity (Enriquez-Geppert et al.,
2017; Sitaram et al., 2017). Participants engage in a closed loop where their mental actions (i.e., changing
mental states) modify the feedback signal, subsequently influencing their forthcoming mental actions
(Birbaumer et al., 2013; Lubianiker et al., 2022). Through repetition and trial-and-error learning, positive
feedbacks that reward the mental state corresponding to the targeted neuronal state (e.g., increase of
theta frequency above a certain threshold) will promote the reoccurrence of this mental state (Birbaumer
et al., 2013; Gaume et al., 2016; Sherlin et al., 2011). Repeated production of the rewarded neuronal state
may lead to neuroplasticity, strengthening synaptic connections within the underlying neural network
(Davelaar, 2018; Shibata et al., 2019). Thus, the neurofeedback technique could enable individuals to learn
to self-regulate their brain activity autonomously, without the aid of an external device (e.g.,
neurostimulation techniques; Herrmann et al., 2016). This can be achieved not only within controlled
laboratory or clinical settings but also extended to real-world environments (Bassett & Khambhati, 2017;
Herrmann et al., 2016).
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 4
In the field of EEG-neurofeedback, numerous studies have sought to increase the amplitude of two
specific brain frequencies to enhance cognition: theta and alpha frequencies (Pfeiffer et al., 2024; Yeh et
al., 2021; Yeh et al., 2022). The frontal theta frequency (4-8 Hz), intricately linked with cognitive control
(Cavanagh, & Frank, 2014; Senoussi et al., 2022), has been targeted to improve attention (Brandmeyer &
Delorme, 2020; Kerick et al., 2023), episodic memory (Eschmann et al., 2020; Tseng et al., 2021),
autobiographic memory (Shoji et al., 2017), memory consolidation (Reiner et al., 2014; Rozengurt et al.,
2016; Rozengurt et al., 2017; Shtoots et al., 2021), motor performance (Eschmann et al., 2022), working
memory (Li et al., 2019; Reis et al., 2016; Wang & Hsieh, 2013) and executive functions (Enriquez-Geppert
et al., 2014; Eschmann & Mecklinger, 2022; Smit et al., 2023). The posterior alpha frequency (8-12 Hz)
serves as a cortical inhibitor, suppressing irrelevant information (Klimesch et al., 2007; Jensen & Mazaheri,
2010; Van Diepen et al., 2019) and contributing to working memory processes (Chen et al., 2023; Riddle
et al., 2020). This frequency, and in particular the higher range of this frequency (referred to as high alpha,
10-12 Hz), has been a target of neurofeedback training aiming to enhance mental rotation (Hanslmayr et
al., 2005; Zoefel et al., 2011), episodic memory (Guez et al, 2015), attention (Navarro Gil et al., 2018),
executive control (Nawaz et al., 2022) and working memory processes (Bobby, & Prakash, 2017; Chen &
Sui, 2023; Dehghanpour et al., 2018; Escolano et al., 2012; Escolano et al., 2014; Esteves et al., 2021;
Hsueh et al., 2016; Naas et al., 2019; Shen et al., 2023; Takabatake et al., 2021; Wei et al., 2017). While
the training duration varies across studies (from one to twenty), some research has indicated successful
neuromodulation of the trained frequencies within a single session (Eschmann et al., 2022; Escolano et
al., 2012; Escolano et al., 2014; Reiner et al., 2014; Rozengurt et al., 2016; Rozengurt et al., 2017; Shtoots
et al., 2021). Although single neurofeedback sessions may not suffice for achieving mastery over one's
brain activity, they do induce notable modifications in cortical excitability (Ros et al., 2010) as well as
neuroplastic changes in structural and functional connectivity (Marins et al., 2019; Ros et al., 2013;
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 5
Sampaio-Baptista et al., 2021). These observations suggest that brief neurofeedback training can
transiently affect brain activity and behavioral outcomes.
These effects are particularly interesting given the existing dilemma within the field of cognitive
training. Indeed, a choice often has to be made between a large sample size or conducting a larger number
of training sessions, due to limited resources. The field of neurofeedback is not exempt from these
constraints, and numerous protocols suffer from insufficient statistical power. This often results in an
inflation of effect sizes and low reproducibility of results (Button et al., 2013; Szucs & Ioannidis, 2017;
Thibault & Pedder, 2022). In fact, a recent systematic review highlighted that only 4% of neurofeedback
studies had sample sizes exceeding 50 participants, with an average number of participants per group of
n = 16.64 (Chiasson et al., 2023). Hence, it is crucial for the neurofeedback field to rely on larger samples
to rigorously evaluate the effects of this technique. Additional concerns arise regarding the potential
influence of non-specific factors on neurofeedback outcomes. Non-specific factors encompass various
extraneous variables that are not directly related to the neurofeedback protocol but can significantly
contribute to the observed effects (Ros et al., 2020). Such factors include expectancy effects, placebo
effects, and unmeasured confounding variables (e.g., La Marca et al., 2018; Thibault et al., 2017).
Inadequate control of these non-specific factors may compromise the validity and interpretability of
neurofeedback research findings (Thibault et al., 2016). Lastly, the psychological, cognitive, and
electrophysiological variables that determine the ability to modulate brain activity remain insufficiently
identified (Haugg et al. 2021; Jeunet et al., 2018; Kadosh & Staunton, 2019; Weber et al., 2020). Identifying
these variables would optimize participant recruitment, mitigate inefficiency issues (Alkoby et al., 2018),
and enable the conduct of more powerful studies without increasing sample size (Thibault & Pedder,
2022).
This study aimed to investigate the specific effects of a single neurofeedback session on brain
frequencies amplitude and working memory. We compared the performance of a group that was trained
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 6
to increase the amplitude of frontal theta frequency with a group trained to increase the amplitude of
posterior high alpha frequency during a single neurofeedback session. Theta oscillations facilitate
cognitive control and item organization in working memory (e.g., Cooper et al., 2019; Deiber et al., 2007;
Hanslmayr et al., 2019; Hsieh et al., 2011; Itthipuripat et al., 2013; Pesonen et al., 2007; Riddle et al., 2020;
Roux & Uhlhaas, 2014; Sauseng et al., 2010, 2009, 2005), while alpha oscillations may inhibit task-
irrelevant information (e.g., de Vries et al., 2020; Hsu & Hämäläinen, 2022; Pesonen et al., 2007; Sghirripa
et al., 2021). We hypothesize that neurofeedback-based amplitude modulation of these frequencies will
improve performance on working memory tasks. To better distinguish between the specific and non-
specific effects of the neurofeedback, we included an active control group that was trained to modulate
randomly selected frequencies. By selecting random frequency bands, participants engage in a genuine
neurofeedback task that targets their own neurophysiological signals. However, because they cannot
learn to modulate a specific frequency, this approach minimizes potential interference with their cognitive
processes. Additionally, unlike sham control groups (Kvamme et al., 2022b), providing real EEG feedback
reduces the risk of inducing a sense of learned helplessness in participants (Aliño et al., 2016). We assessed
brain frequency amplitude, performance on working memory tasks (including digit span, n-Back, and Corsi
blocks), as well as self-reported psychological states related to emotional and attentional aspects before
and after the training period. We also aimed to identify factors (psychological, cognitive or
electrophysiological) that could predict neuromodulation success during the neurofeedback task.
Method
Sample
G-Power (Faul et al., 2007) was utilized to determine the sample size needed for an analysis of
variance with repeated measures focusing on within-between interactions. For the investigation of the
impact of training on working memory, where a comparison was made between pre- and post-training
performances, we computed the required sample size based on the following parameters: effect size f =
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 7
.20, power level = .90, number of groups = 3, number of measures = 2, correlation between repeated
measures = .50. An estimated total sample size of 84 participants was obtained. 70 participants were
randomly assigned to either the Theta group, aiming to increase the amplitude of the theta frequency in
the frontal cortex, or to the Control group, aiming to increase randomly selected frequencies within each
block. Data from some participants had to be excluded (N = 8, 11%) due to low-quality EEG data (i.e., more
than 20% of the signal was removed after artifact rejection, N = 7) and one participant was too fatigued
to complete the entire protocol and chose to discontinue the experiment. The data from the group trained
to increase parietal high alpha amplitude were obtained through a prior identical experimental protocol
where the research question aimed to investigate the effects of a strategy list on neuromodulation (Chikhi
et al., 2023). A total of 101 participants were included in our analyses. The Theta group consisted of 28
participants (Mean age = 19.4 ± 1.81, 75% woman), the Control group comprised 34 participants (Mean
age = 21.41 ± 8.28, 91% woman) and the High alpha group consisted of 39 participants (Mean age = 23.28
± 5.96, 82% woman). The inclusion criteria were normal or corrected-to-normal vision, no psychiatric or
neurological disorders, and no usage of medication. All participants were naïve to the neurofeedback
technique.
Neurofeedback task and procedure
Each participant underwent a single experimental session lasting approximately 120 minutes
(Figure 1) and received course credits for their participation. Prior to and after neurofeedback session,
participants completed computerized tasks assessing working memory (digit span, spatial span and n-back
task) from the Psychology Experiment Building Language (Mueller & Piper, 2014). Additionally,
participants completed a shortened version of the Neurofeedback Evaluation & Training questionnaire
(NExT questionnaire; Bismuth et al., 2020; Jaumard-Hakoun et al., 2017; Pillette et al., 2021) to evaluate
their subjective state during the neurofeedback protocol. Neurofeedback training began by measuring
the dimensions of the participants' heads to determine electrode locations. Next, conductive gel and paste
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 8
were applied to prepare the skin and improve signal quality at each electrode location. Once the correct
impedance was achieved at each electrode, a baseline was measured by recording the resting-state EEG
signal during 3 minutes. Participants were instructed to maintain their gaze on a black screen positioned
in front of them. They were required to remain relaxed with their eyes open, refraining from blinking,
making any bodily movements, or clenching their jaw.
After recording the baseline, the average amplitude of theta and high alpha frequencies were
extracted to determine the threshold for activating positive feedback. In the Control group, participants
received feedback based on a threshold that automatically adapted to the average amplitude of the
targeted brain signal. The frequency bands used for the participants in the Control group were randomly
generated 2 Hz bands for each block and each participant. The randomly selected frequencies included a
range of 2Hz, situated between 1 and 30 Hz, except for the frequency bands targeted by the experimental
groups (4-8Hz and 10-12Hz). Participants were always required to increase the amplitude of the selected
frequency band. Irrespective of the group, all participants were informed that the gauge shown on the
left side of the screen depicted the real-time amplitude of a specific brain frequency. When the gauge
exceeded the threshold (as denoted by an orange bar), a screen animation (depicting a roller coaster) was
triggered, accompanied by enjoyable music. The participants' goal was to 'keep the level of the gauge
above the threshold for as long and as frequently as possible'. All participants were informed that the
audiovisual feedback would convey information about their brain activity and aid in achieving a mental
state that triggers the reward feedback. They were not provided with particular mental strategies for
regulating their brain activity. Following the completion of the 10 training blocks, the resting-state EEG
signal was once again recorded. Subsequently, the electrodes were removed, the participants' scalps were
cleaned, and they proceeded to complete the NExT questionnaire and computerized working memory
tasks once more.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 9
Figure 1
Experimental protocol overview
Note. Preceding and subsequent to the neurofeedback session, they engaged in computerized tasks
designed to evaluate their working memory (Mueller & Piper, 2014). Additionally, they responded to a
concise questionnaire aimed at assessing their psychological state (Bismuth et al., 2020; Jaumard-Hakoun
et al., 2017). The neurofeedback protocol encompassed measuring a 3-minute resting-state EEG before
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 10
and after the 10 neurofeedback training blocks. For the Theta and High alpha groups, the positive
feedback activation threshold was determined based on the mean amplitude of theta and high alpha
frequencies, respectively. Both groups were provided real-time feedback, presented as a gauge positioned
on the left side of the screen. Exceeding the established threshold led to the activation of animation and
music. The primary objective was to sustain the gauge level above this threshold. Following the 10 training
blocks and the second resting-state EEG, participants undertook post-session working memory tasks and
questionnaire.
EEG acquisition and processing
An encoder (ProComp T740M) was connected to a laptop via a fiber optic cable and USB interface
for recording the EEG signal. The BioGraph Infiniti software (version 6.1, Thought Technology Ltd.,
Montreal, QC, Canada) was used to analyze the signal and convert the targeted brain activity into audio-
visual feedback. The High alpha group was trained using an electrode positioned at Pz. For the Theta
group, the electrode was positioned at Fz. In the Control group, the electrode was placed at the vertex
(Cz), which is equidistant from Fz and Pz (Pimenta et al., 2018). Additionally, a reference electrode and a
ground electrode were positioned on the right and left earlobes, respectively. The signal was sampled at
256 Hz, filtered using a bandpass filter (0.1 to 60 Hz), and a notch filter (50 Hz) to eliminate electrical
interference. The software extracted the amplitude of the signal trained for each group (theta, high alpha,
or random frequency) and converted it into audio-visual feedback. The main brain frequencies were
recorded for offline analysis: delta (1-4 Hz), theta (4-8 Hz), low alpha (8-10 Hz), high alpha (10-12 Hz), beta
(12-35 Hz), and gamma (35-64 Hz) using fast Fourier transform (FFT). Nuprep Skin Prep Gel (Weaver and
Company, Aurora, USA) was used to clean the skin prior to electrode placement. Electrodes were fixed
with Ten20 Conductive Paste (Weaver and Company, Aurora, USA). Impedance was assessed using an
EEG-Z sensor and kept below 10 kΩ. An automatic artifact rejection threshold was set at 100 μV before
visually inspecting the raw data and manually rejecting artifacts. Additional pre-processing and artifact
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 11
correction steps could not be performed due to the software used. The feedback was presented as an
animation illustrating a roller coaster, accompanied by calming music (Wang & Hsieh, 2013).
For the groups trained to increase the theta or high alpha amplitude, positive feedback (animation
and music) was activated when the average amplitude of the trained frequency exceeded 80% of the
baseline's average value. For the Control group, a fixed threshold could not be applied because the trained
frequency varied between each block. Therefore, we employed an adaptive threshold based on the
amplitude of the selected frequency. This threshold corresponded to +1% of the average signal amplitude,
with a delay of 0.2 seconds. This adaptive positive threshold ensures that participants receive feedback
that is contingent upon their own neural activity, reducing the risk of frustration and learned helplessness
(Sorger et al., 2019).
Working memory tasks
To assess different facets of participants' working memory, three distinct working memory tasks
were administered via the Psychology Experiment Building Language (Mueller & Piper, 2014). The tasks
included the backward digit span (Wechsler, 1955), the n-back task with 1, 2, and 3-back conditions (Owen
et al., 2005), and the backward spatial span task, also known as Corsi's blocks task (Corsi, 1972). The
backward digit span task required participants to recall a series of randomly presented digits in reverse
order (Richardson, 2007). To complete this task, participants utilized the number pad of a standard
keyboard. Throughout the task, the length of the digit sequence gradually increased. The dependent
variable was defined as the longest successfully recalled sequence. The n-back task involves presenting
participants with a series of stimuli (e.g., a letter), presented one after another in a sequential manner.
The task required participants to identify if the current stimulus matches the one presented 'n' position
earlier in the sequence. For example, during the 2-back condition, if the letter sequence displayed on the
screen was F-N-B-N-B-F, participants were instructed to press the ‘yes’ key when the second 'N' and the
second 'B' appeared. They had to press the 'no' key for all other letters. The n-back task included three
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 12
conditions: 1-back, 2-back, and 3-back. The dependent variable for each condition was the percentage of
correct responses. The computerized Corsi block task comprised nine fixed squares displayed on a
computer screen in a pseudo-random arrangement. These squares change color individually over a short
period. In the backward condition, participants were instructed to click on the squares in the reverse order
of their original sequence. As the experiment advanced, the length of the block sequence progressively
increased. The dependent variable was the longest accurately recalled sequence's length (Mueller & Piper,
2014). The digit span backward task was used to engage the executive control component of working
memory (Lezak et al., 2004). The Corsi blocks backward task was employed to involve visuospatial working
memory processing (Kessels et al., 2008). The n-back task involves simultaneously retaining previously
presented stimuli, processing the currently presented stimulus, and updating the contents of working
memory (Lezak et al., 2004; Soveri et al, 2017). By using tasks that target various working memory
processes, our aim was to distinguish the cognitive effects of theta and high alpha amplitude modulation.
Self-reported questionnaire
To effectively assess the psychological state of participants before, during, and after
neurofeedback, while minimizing the burden of the experimental procedure and cognitive load, we
utilized a condensed version of the NExT questionnaire (Bismuth et al., 2020; Jaumard-Hakoun et al., 2017;
Pillette et al., 2021). The questionnaire items were adapted from established and validated questionnaires
(see Supplementary material in Chikhi et al., 2022 for item sources). Each participant was asked to rate
their experiences using a 5-point Likert scale. The questionnaire covered five distinct dimensions:
emotional state (‘calm,’ ‘energetic,’ ‘happy,’ ‘relaxed,’ and ‘satisfied’), attentional state (‘receptive’ and
‘focused’), cognitive load, motivation, and agency (‘sense of control over the feedback signal’ and
‘predictability of the feedback signal’). The questionnaire was administered on two occasions. Firstly, it
only consisted of the first two dimensions, assessing emotional and attentional state just before the
neurofeedback session. This measurement aimed to evaluate the effect of self-reported psychological
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 13
states before training on neuromodulation ability. Secondly, the questionnaire was administered
immediately after the neurofeedback session and included all five dimensions. This second administration
aimed to assess the impact of neurofeedback on participants' psychological state changes (via the first
two dimensions). The remaining three dimensions were used to compare participants' psychological state
during the neurofeedback task, including cognitive load, motivation, and sense of agency. This enabled an
investigation into the potential influence of these variables on the participants' success in the
neurofeedback task. For example, participants in the control group might exhibit a reduced sense of
agency towards the neurofeedback device, indicating a weaker perception of control over the feedback
signal, in contrast to the experimental groups. Such a disparity could potentially affect their motivation
and consequently lead to a decline in performance both during the neurofeedback task and in the
subsequent behavioral assessments.
Statistical analysis
Statistical analyses were performed with the RStudio environment (v.4.0.2; R Core Team, 2020)
with the following packages: lme4 (v.1.1-27.1; Bates et al., 2015) and lmer (v. 3.1-3; Kuznetsova et al.,
2017) for classical and mixed linear regression, multcomp (v. 1.4-17; Hothorn et al., 2008) and modelbased
(v. 0.8.0; Makowski et al., 2020a) for post hoc contrasts, and report (v. 0.5.1; Makowski et al., 2020b),
ggplot2 (v. 3.3.6; Wickham, 2016) and smplot2 (v. 0.1.0; Min & Zhou, 2021) for formatting and
visualization of results. For repeated measures, mixed-effects linear models were estimated using the
Restricted Maximum Likelihood (REML) method and fitted with a random intercept (1| Participant).
Classical linear regression models were estimated using Ordinary Least Squares (OLS) method. P-values
for Type III ANOVA F-tests were obtained using Kenward-Roger approximation for degrees of freedom
(Kenward & Roger, 1997; Luke, 2017). In cases where an effect was statistically significant, we applied
Tukey's post hoc contrasts with Holm correction for multiple comparisons (Schad et al., 2020). Only
significant differences from post hoc comparisons were detailed. The significance threshold for all
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 14
analyses was set at α = 0.05. Effect sizes were interpreted as follows: ηp2 = 0.01 as a small effect, ηp2 = 0.06
as a moderate effect, and ηp2 = 0.14 as a large effect.
Results
Effects of neurofeedback on targeted brain signal (Figure 2)
Effect of neurofeedback on theta amplitude during training
A mixed linear model was employed to predict the amplitude of theta frequency with Group
(Theta, High alpha, Control) and Training Block (1 to 10) as fixed effects, and Participant as a random
effect. The ANOVA revealed a significant and strong main effect of Group (F(2) = 14.14, p < .001; ηp2 = 0.22,
95% CI [0.11 - 1.00]). Post hoc analysis indicated that the average theta frequency amplitude was
significantly lower in the High alpha group compared to the Theta group (β = -1.37, SE = 0.44, p = .004).
The difference between the Theta group and the Control group was not statistically significant. The main
effect of Training Block was also statistically significant and weak (F(9) = 2.10, p = 0.027; ηp2 = 0.02, 95% CI
[0.001 - 1.00]), however, post hoc comparisons revealed no significant differences in amplitude between
the blocks. The interaction between Group and Training Block was not statistically significant (p = .053).
The absence of global interaction among the three groups and the training blocks might mask
divergent effects depending on the pair of groups under consideration. Thus, we conducted two
complementary analyses investigating the Group x Training Block interaction, specifically comparing (1)
Theta versus High alpha group and (2) Theta versus Control group. In the first mixed linear model, we
predicted the amplitude of theta frequency with Group (Theta x High alpha) and Training Block as fixed
effects, and Participant as a random effect. The ANOVA revealed a significant interaction between Group
and Training Block (F(9) = 2.43, p = .010; ηp2 = 0.04, 95% CI [0.005, 1.00]), and subsequent post hoc analysis
revealed that the average theta frequency amplitude was significantly lower in the High Alpha group
compared to the Theta group across blocks 3 to 10 (SE = 0.44, all p < .01). The second mixed linear model,
predicting the amplitude of theta frequency with Group (Theta x Control) and Training Block as fixed
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 15
effects, and Participant as a random effect, indicated that the interaction between Group and Training
Block was not statistically significant (F(9) = 1.17, p = .313; ηp2 = 0.02, 95% CI [0.00, 1.00]).
Effect of neurofeedback on theta amplitude during resting-state
A mixed linear model was performed to predict the amplitude of theta frequency recorded during
resting-state EEG, considering Group and Resting-EEG (pre- and post-training EEG recording) as fixed
effects and Participant as a random effect. The ANOVA revealed a statistically significant and moderate
main effect of Group (F(2) = 5.64, p = .005; ηp2 = 0.10, 95% CI [0.02 – 1.00]). Post hoc analysis indicated
that the average theta amplitude was significantly lower in the High alpha group compared to the Control
group (β = -1.67, ES = .60, p = .02), while no significant difference was observed between the Theta group
and the Control group (p = .52) nor between the Theta group and the High alpha group (p = .09).
Additionally, the main effect of Resting-EEG was statistically significant and weak (F(1) = 4.82, p = 0.03; ηp2
= 0.05, 95% CI [0.002 – 1.00]), with a slightly higher mean amplitude observed after training (M = 10.13 ±
2.86) compared to before (M = 9.94 ± 2.42). However, neither the overall interaction between the Group
and Resting EEG reached significance (p = .11), nor did the pairwise comparisons (Theta x High Alpha, p =
.42; Theta x Control, p = .24). To assess the impact of resting EEG activity on the neurofeedback task, we
applied a classical linear model to compare the average theta amplitude activity during the first resting-
state among the groups. The ANOVA indicated that the main effect of Group was statistically significant
and medium (F(2, 98) = 5.11, p = .008; ηp2 = 0.09, 95% CI [0.02, 1.00]). Post-hoc comparisons revealed that
the mean theta amplitude was significantly lower in the High alpha group compared to the Control group
(β = 1.67, p < .01). No significant differences were observed between the Theta group and the Control
group (p = .48) or between the Theta group and the High alpha group (p = .06).
Effect of neurofeedback on high alpha amplitude during training
A mixed linear model was performed to predict the amplitude of high alpha frequency with Group
and Training Block as fixed effects, and Participant as a random effect. The ANOVA revealed that the main
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 16
effect of Training Block was statistically significant yet weak (F(9) = 5.60, p < .001; ηp2 = 0.05, 95% CI [0.02,
1.00]). Post hoc analysis revealed no significant differences in amplitude between the blocks. Moreover,
neither the main effect of Group nor the interaction between Group and Training Block reached statistical
significance (p = .21 and p = .08, respectively).
Again, the lack of global interaction between the three groups and the training blocks may mask
differential effects depending on the pair of groups considered. We conducted two complementary
analyses testing the Group x Block interaction when specifically comparing (1) Theta vs. High alpha group
and (2) Theta vs. Control group. Using a mixed linear model, we predicted the amplitude of high alpha
frequency with Group (High alpha and Theta) and Training Block as fixed effects, and Participant as a
random effect. The ANOVA indicated a significant interaction between Group and Training Block (F(9) =
2.25, p = .018; ηp2 = 0.03, 95% CI [3.02e-03, 1.00]). However, post hoc analysis did not indicate any
amplitude difference across blocks. The second mixed linear model, predicting the amplitude of high alpha
frequency with Group (High alpha and Control) and Training Block as fixed effects, and Participant as a
random effect, indicated that the interaction between Group and Block was not statistically significant
(F(9) = 0.60, p = .795; ηp2 = 0.008, 95% CI [0.00, 1.00]).
Effect of neurofeedback on high alpha amplitude during resting-state
A mixed linear model was performed to predict the amplitude of high alpha frequency recorded
during resting-state EEG, with Group and Resting-EEG as fixed effects and Participant as a random effect.
The ANOVA revealed a statistically significant and robust main effect of Resting-EEG (F(1) = 22.37, p <
.001; ηp2 = 0.19, 95% CI [0.08 - 1.00]), indicating a higher average amplitude after training (M = 6.18 ± 2.29)
compared to before (M = 5.83 ± 2.17). However, the main effect of Group (p = .39) and the interaction
between Group and Resting EEG (global: p = .99; Theta x High Alpha, p = .97; Theta x Control, p = .89) was
not statistically significant. We also assessed the difference in EEG activity during the first resting-state
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 17
using a classical linear model. The ANOVA indicated that the main effect of Group was not statistically
significant (p = .36).
Figure 2
Evolution of theta and high alpha amplitude during neurofeedback and resting-states
Note. A. Changes in theta amplitude during neurofeedback and resting-states. The first line graph shows
that, after an initial decrease, there has been a gradual increase of theta amplitude in the Theta and
Control group. The second graph represents the mean amplitude and individual data points of theta
frequency amplitude during the resting state, measured just before and just after the neurofeedback
session. The third and final plot is an interaction plot showing the differential effects of neurofeedback
training on the theta amplitude of each group. B. Changes in high alpha amplitude during neurofeedback
and resting-states. The first line graph shows that there has been an increase of high alpha amplitude in
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 18
the three groups. The second graph shows the mean amplitude of the high alpha frequency and the
individual data points for each resting-state and the last graph is an interaction plot.
Effects of neurofeedback on all brain frequencies
Figure 3 illustrate the effect of neurofeedback on all frequency bands. Statistical analysis of
neurofeedback effects on other frequencies amplitude during training and resting-state are presented in
Supplementary material (Section 1.1).
Figure 3
Effect of neurofeedback on EEG frequency bands.
Note. Forest plot illustrating the mean slope of frequency band amplitudes across the training, error
bars represent standard errors.
Effect of neurofeedback on working memory performance (Figure 4)
Effect of neurofeedback on backward digit span task performance
We conducted a mixed linear model to predict the score on a backward digit span task, with Group
and Cognitive Screening (pre-and post-training working memory measurement) as fixed effects and
Participant as a random effect. The ANOVA revealed that the main effect of Cognitive Screening was
statistically significant and strong (F(1) = 29.39, p < .001; ηp2 = 0.23, 95% CI [0.12 – 1.00]), with a higher
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 19
mean score after training (M = 6.71 ± 1.62) compared to before (M = 5.80 ± 1.69). However, the main
effect of Group and the interaction between Group and Cognitive Screening were not statistically
significant (p = .12 and p = .18, respectively).
Effect of neurofeedback on 1-back and 2-back tasks performance
We conducted a mixed linear model to predict the score on the 1-back task, with Group and
Cognitive Screening as fixed effects and Participant as a random effect. The ANOVA revealed that the main
effect of Cognitive Screening was statistically significant and moderate (F(1) = 6.69, p = .011; ηp2 = 0.06,
95% CI [0.008 – 1.00]), with a higher mean score after training (M = 0.96 ± 0.08) compared to before (M =
0.93 ± 0.10). The main effect of Group and the interaction between Group and Cognitive Screening were
not statistically significant (p = .19 and p = .16, respectively). We then conducted a mixed linear model to
predict the score on the 2-back task, with Group and Cognitive Screening as fixed effects and Participant
as a random effect. The ANOVA revealed a statistically significant and robust main effect of Cognitive
Screening (F(1) = 25.64, p < .001; ηp2 = 0.21, 95% CI [0.10 – 1.00]). Post-training scores (M = 0.93 ± 0.08)
were significantly higher compared to pre-training scores (M = 0.88 ± 0.09). However, the main effect of
Group and the interaction between Group and Cognitive Screening were not significant (p = .66 and p =
.56).
Effect of neurofeedback on 3-back and Corsi block tasks performance
Further analyses were conducted to evaluate the effects of Group, Cognitive Screening, and their
interaction on 3-back scores and Corsi block performance. Our mixed models and ANOVA results indicated
that the main effects of Group (p = .70 and p = .08) and Cognitive Screening (p = .08 and p = .22), as well
as the interaction between these variables (p = .53 and p = .12), were not statistically significant.
Effect of brain activity modulation on working memory performance
To assess the link between brain activity and cognitive performance, we evaluated the statistical
relationship between the slope of theta and high alpha amplitude during neurofeedback and the changes
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 20
in working memory task performance (measured through the difference between post-training and pre-
training scores). Five linear models were performed to predict the changes in performance across working
memory tasks based on the modulation of the amplitudes of two frequencies targeted during training
(slope of the amplitude evolution for theta and high alpha frequencies) and the Group variable as
predictor variables. The ANOVA tests showed no significant main effects nor interactions (results are
presented in the Supplementary material, Section 1.2).
Figure 4
Changes in working memory tasks performance
Note. Average performance and individual data points on working memory tasks, namely digit span,
spatial span and n-back (1, 2 & 3-back). Descriptively, there was an average increase in performance on
the digit span, 1-back and 2-back tasks. As this increase was observed in all three groups, we did not
observe an interaction effect. The span block task presented less unequivocal results, with an average
increase for the Theta group, a decrease for the Control group and stagnation of performance for the High
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 21
alpha group. Conversely, on the 3-back task, the Control group showed the greatest improvement in
performance, followed by the High alpha group and then the Theta group, whose performance stagnated.
However, as presented in the text, we observed no interaction effect between groups and those two
behavioral measures.
Effect of neurofeedback on subjective psychological state (Figure 5)
Calm score
We conducted a mixed-effects linear model to predict the calm score using Group and Self-
Reported Measure as fixed effects and Participant as a random effect. The ANOVA indicated that the main
effect of the Self-Reported Measure was statistically significant and moderate (F(1) = 9.22, p = .003; ηp2 =
0.09, 95% CI [0.02 – 1.00]), with a higher average score after training (M = 3.19 ± 0.97) compared to before
(M = 2.78 ± 1.05). The main effect of the Group and the interaction between Group and Self-Reported
Measure were not significant (p = .92 and p = .52, respectively).
Energetic score
We performed a mixed-effects linear model to predict the energy score using Group and Self-
Reported Measure as fixed effects and Participant as a random effect. The ANOVA revealed that the main
effect of the Self-Reported Measure was statistically significant and strong (F(1) = 41.51, p < .001; ηp2 =
0.30, 95% CI [0.18 – 1.00]), with a lower average score after training (M = 1.16 ± 0.97) compared to before
(M = 1.88 ± 0.98). The main effect of the Group and the interaction between Group and Self-Reported
Measure were not significant (p = .19 and p = .38).
Relaxed score
We fitted a mixed-effects linear model to predict the relaxation score using Group and Self-
Reported Measure as fixed effects and Participant as a random effect. The ANOVA demonstrated that the
main effect of the Self-Reported Measure was statistically significant and strong (F(1) = 38.52, p < .001;
ηp2 = 0.28, 95% CI [0.16 – 1.00]), with a higher average score after training (M = 3.08 ± 0.94) compared to
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 22
before (M = 2.25 ± 1.28). The main effect of the Group and the interaction between Group and Self-
Reported Measure were not significant (p = .69 and p = .78).
Satisfaction score
We conducted a mixed linear model to predict the satisfaction score, with Group and Self-
Reported Measure as fixed effects and Participant as a random effect. The ANOVA revealed a statistically
significant and moderate main effect of the Self-Reported Measure (F(1) = 7.58, p = .007; ηp2 = 0.07, 95%
CI [0.01 – 1.00]), indicating a higher average score after the training (M = 2.69 ± 0.98) compared to before
(M = 2.42 ± 1.19). However, the interaction effect between Group and Self-Reported Measure was
statistically significant and moderate (F(2) = 4.70, p = .011; ηp2 = 0.09, 95% CI [0.01, 1.00]). Post hoc analysis
revealed that for the Control group, the average score was significantly higher after the training compared
to before (β = 0.59, SE = 0.19, p = .03). The main effect of Group was not significant (p = .053).
Happiness, Receptivity, and Concentration scores
We conducted three mixed linear models to predict the happiness score, receptivity score, and
concentration score, with Group and Self-Reported Measure as fixed effects and Participant as a random
effect. The ANOVA indicated that the main effects of Group, Cognitive Screening, and the interaction
between these two variables were not significant for these three items (results are presented in the
Supplementary material, Section 1.3).
Cognitive Load score
We performed a linear model to predict the subjective cognitive load score during neurofeedback
training, with Group as the predictor variable. The ANOVA revealed a statistically significant and strong
main effect of Group (F(2, 98) = 38.19, p < .001; ηp2 = 0.44, 95% CI [0.32 – 1.00]). Post hoc analysis
demonstrated that the mean cognitive load score was significantly lower in the High Alpha group
compared to the Theta group (β = -1.92, t(98) = 7.78, SE = 0.25, p < .001) and the Control group (β = -1.63,
t(98) = 7.01, SE = 0.23, p < .001).
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 23
Flow, Mastery, and Coherence scores
We employed three linear models to predict the subjective feeling of flow, mastery and coherence
of the feedback signal during the neurofeedback training, with Group as the predictive variable. The
ANOVAs revealed that there were no significant main effects of Group on any of the scores (p = .37, p =
.86 and p = .39, respectively).
Figure 5
Statistically significant changes in self-reported psychological state
Note. Average score and individual data point of questionnaire scores measured before and after
neurofeedback. In a similar way for the three groups, there was an increase in the calm and relaxation
scores and a decrease in the energy score. In terms of satisfaction, the Control group showed the strongest
increase, as evidenced by the significant interaction effect (see Results section), followed by the Theta
group. The High alpha group showed a slight decrease. *** p < .001.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 24
Predictors of neuromodulation
To better understand the factors involved in neuromodulation, we utilized resting-state EEG
activity, pre-training scores on working memory tasks and questionnaires to predict the neuromodulation
of targeted signals through neurofeedback. A mixed linear model was employed to predict the slope of
theta amplitude during neurofeedback using Group and resting-state brain frequencies (delta, theta, low
alpha, high alpha, beta, and gamma) as predictor variables. ANOVA results revealed a statistically
significant and strong main effect of resting-state theta amplitude (F(1, 92) = 14.06, p < .001; ηp2 = 0.13,
95% CI [0.04 - 1.00]). The other main effects were not significant (see Supplementary material, Section
1.4.a). A mixed linear model was utilized to predict the slope of high alpha amplitude during
neurofeedback using Group and resting-state brain frequencies (delta, theta, low alpha, high alpha, beta,
and gamma) as predictor variables. ANOVA indicated a statistically significant and moderate main effect
of resting-state low alpha amplitude (F(1, 92) = 7.80, p = .006; ηp2 = 0.08, 95% CI [0.01 - 1.00]) and a strong
main effect of resting-state high alpha amplitude (F(1, 92) = 21.97, p < .001; ηp2 = 0.19, 95% CI [0.09 -
1.00]). The other main effects were not significant (see Supplementary material, Section 1.4.b). The
performance scores on working memory tasks and the questionnaire measured before neurofeedback did
not predict the changes in the amplitude of the signals targeted by the training (results are presented in
the Supplementary material, Section 1.4).
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 25
Figure 6
Scatter plots of the two predictors of neuromodulation
Note. We predicted the evolution of the amplitude (i.e., slope) of the signals targeted by neurofeedback
(i.e., theta and high alpha) via psychological (self-reported questionnaire), cognitive (working memory
performance) and electrophysiological (resting-state EEG) variables measured before neurofeedback.
Only resting-state EEG of the trained signal predicted the slope of those signals (A: theta slope, B: high
alpha slope). The two scatter plots show the slope of the trained signal as a function of the mean
amplitude recorded in the resting-state. The points represent individual data, the line represents the fitted
linear model and the shaded area the 95% confidence interval of the fitted values.
Discussion
In this study, we investigated the effect of a single neurofeedback session on brain oscillations
amplitude, working memory task performance, and self-reported psychological states. We hypothesized
that the neurofeedback's effects on brain activity would vary depending on whether the group was trained
to enhance the amplitude of the theta frequency, the high alpha frequency, or random frequencies (active
control group). This modulation of the targeted brain signal was expected to result in improved
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 26
performance in working memory tasks. To gain a deeper understanding of the self-regulation mechanism,
we utilized resting-state EEG activity, working memory task performance, and self-reported psychological
states to predict the neuromodulation of the trained signal. While our analyses did reveal
electrophysiological and behavioral improvements, these gains were not exclusive to the trained signals.
An increase in the frequency targeted by the training has been observed for the theta and high alpha
groups, but no significant difference has been observed between these groups and the active control
group. Among all the predictors of neuromodulation that have been examined, theta and high alpha
amplitudes at rest have been shown to predict the ability to modulate the frequencies targeted by
neurofeedback. Participants who showed the greatest increase in amplitude for the trained frequency
during neurofeedback were those with the highest amplitudes during resting EEG measurements. Overall,
these findings contribute to a better understanding of the factors influencing the outcomes of a
neurofeedback training and highlight the challenges associated with using random frequencies training as
an active control group.
Effect of neurofeedback on the amplitude of trained brain oscillations
The main goal of this study was to investigate the effects of a neurofeedback session designed to
enhance the amplitude of theta and high alpha frequencies. The first hypothesis was that engaging in
neurofeedback to increase the amplitudes of theta and high alpha frequencies would indeed affect the
amplitudes of the targeted frequencies.
Theta frequency
When comparing the three groups together, we observed a main effect of neurofeedback training
on the average amplitude of theta frequency but no interaction effect. Nonetheless, the absence of a
global interaction effect may mask specific differences among the groups. Indeed, we observed a
significant interaction between the Theta and High Alpha groups, indicating an increase in theta amplitude
in the former compared to the latter. This interaction effect is consistent with previous research that
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 27
indicates an increase in theta amplitude during a single training session (Eschmann et al., 2021; Rozengurt
et al., 2016; Rozengurt et al., 2017; Shtoots et al., 2021), and suggests that the theta training had specific
effects compared to a group trained to modulate another frequency. However, we did not observe
significant differences between the Theta group and our active Control group. In the previously cited
studies, the active control group underwent training to enhance the amplitude of the beta frequency,
resulting in slight decreases (Rozengurt et al., 2017), no changed (Rozengurt et al., 2016), or slight
increases in theta amplitude (8%, Shtoots et al., 2021). Another study utilized neurofeedback to increase
beta1 frequency (15-18 Hz) amplitude and showed a slight decrease in theta amplitude during the training
(Pimenta et al., 2018). The lack of difference could be attributed to our control group, which used random
frequencies instead of the beta frequency. According to our findings, training to increase multiple
frequencies appears to induce a more generalized neuromodulation than training targeting a specific
frequency. Thus, a more appropriate control group would be one trained either to decrease randomly
selected frequencies or to increase a different and uncorrelated frequency (e.g., beta). Concerning the
effect of training on resting brain activity, our findings revealed an increase in the amplitude of resting-
state theta amplitude after the training. However, no significant differences were observed between the
groups concerning changes in theta amplitude, whether through assessing the overall interaction among
the three groups or through additional pairwise comparisons. This lack of interaction effect cannot be
directly compared to the findings of earlier studies (Eschmann et al., 2021; Rozengurt et al., 2016;
Rozengurt et al., 2017; Shtoots et al., 2021), as these previous studies solely assessed theta frequency
amplitude at rest before the neurofeedback, without considering post-training measurements.
High alpha frequency
Consistent with previous studies (Escolano et al., 2012; Escolano et al., 2014), we observed that a
single high alpha neurofeedback training session leads to an increase in the amplitude of the high alpha
frequency both during training (compared to the theta group) and in subsequent resting states. However,
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 28
we did not observe an interaction effect when comparing the high alpha group with the active control
group, as the latter also exhibited an increase in high alpha frequency. Again, this suggests that training
to increase different frequencies across neurofeedback training may result in nonspecific, broad effects
on frequency amplitude.
Other frequencies
The analysis of the neurofeedback effect on the amplitude of other brain frequencies has shown
an increase in delta frequency amplitude within the Theta and Control group, a reduction in beta
frequency amplitude among the High alpha group, and a decrease in gamma frequency amplitude across
all groups. Thus, our findings corroborate the absence of selectivity resulting from a single neurofeedback
session on EEG activity (Dessy et al., 2020; Jurewicz et al., 2018; Pimenta et al., 2018). This outcome was
expected within the Control group, which underwent training to regulate various brain frequencies, but
not within the Theta and High alpha groups. A unique neurofeedback session does not appear to enable
the modulation of brain activity exclusively within the frequency targeted by the neurofeedback device.
Effect of training on working memory performance
The second hypothesis of this study was that enhancing the amplitude of theta or high alpha
frequencies through neurofeedback would improve performance on working memory tasks. While
significant improvements were observed in performance on the digit span task and in two conditions of
the n-back task (1-back and 2-back), these gains were observed across all groups. These improvements in
performance across the three groups may be attributed to the neuromodulation observed in each of the
groups. To test this hypothesis, we predicted the difference in working memory performance (post-
training minus pre-training) using the slope of the evolution of theta and high alpha amplitude during
training. Similar to the study conducted by Eschmann and colleagues (2021), which evaluated the link
between the modulation of theta amplitude and changes in performance on a 2-back task, no statistically
significant relationship was observed. The absence of statistically significant relationships in both studies
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 29
underscores the possibility that improvements in behavioral tasks may not directly result from the
targeted neurofeedback training, but rather from non-specific factors such as increased task familiarity or
increased motivation. It must, however, be emphasized that our ability to make psychophysiological
inferences is limited here given that we did not record brain activity during the behavioral tasks. Further
research examining cerebral activity during working memory tasks may yield more robust and
comprehensive explanations.
Effect of training on self-reported questionnaire
The analysis of the questionnaires assessing participants' emotional state and consciousness
revealed significant changes. A substantial decrease in the sense of energy was observed, alongside an
increase in feelings of calmness and relaxation, regardless of the experimental group. The absence of
differences between the groups suggests that these effects were not limited to the trained frequency and
may indicate an increase in participants' levels of fatigue. Despite having short breaks between each block,
the 30-minute neurofeedback training demanded continuous vigilance in focusing on the feedback signal.
This increase in fatigue could potentially explain the observed modulation of various brain frequencies—
namely, delta, theta, alpha, and beta. A recent meta-analysis emphasized the connection between fatigue
and changes in amplitude within these frequencies (Tran et al., 2020). Interestingly, self-reported
questionnaire responses indicated that both the Theta and Control groups experienced higher cognitive
loads during training than the High alpha group. While this may support the established link between
theta amplitude and cognitive load (Chikhi et al., 2022), we found no statistically significant correlation
between theta activity and perceived cognitive load. Nonetheless, this result, along with similar
motivation and agency scores across groups, suggests that participants of the Control group were equally
engaged in the task as participants in the Theta and High alpha groups.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 30
Towards specificity of EEG neurofeedback
Here, a single neurofeedback session aimed at increasing one frequency amplitude (theta or
high alpha), does not elicit specific modulation of brain activity when compared to a group trained to
increase random frequencies. First, this absence of significant differences may be attributed to the brief
duration of our training, which was limited to a single session. Although there is no consensus on the
time required for learning to occur, theoretical models suggest that the repeated production of the
neural state rewarded by feedback induces neuroplasticity, facilitating the re-establishment of the
neural state targeted by the training (Birbaumer et al., 2013; Davelaar, 2018; Lubianiker et al., 2022; Niv,
2013; Ros, 2014; Shibata et al., 2019). Thus, a prolonged training duration could strengthen the neural
connections associated with the specific targeted frequencies, leading to more durable and robust
neural changes compared to a group trained to modulate random signals. Consequently, by prioritizing
sample size over the number of training sessions, our findings do not preclude the possibility that
mastering the self-regulation of theta or high alpha frequencies over an extended period could produce
distinct effects compared to a group trained to increase multiple frequency bands.
However, brain frequencies may be inter-correlated (Canolty & Knight, 2010; Klimesch, 2018), to
the extent that EEG-neurofeedback training leads to non-specific modulation in the amplitude of brain
frequencies. In this respect, the use of conventional EEG frequency bands could be sub-optimal when
targeting a specific parameter of brain activity to positively impact a specific cognitive component.
Therefore, it may be crucial to find functionally relevant biomarkers that can be trained independently.
Hence, recent studies have used innovative methods for analyzing EEG signals in neurofeedback
protocols. For example, studies have used multivariate pattern analysis (Haynes & Rees, 2006;
Taschereau-Dumouchel et al., 2021) to "decode" cognitive states using machine learning techniques,
thereby targeting specific brain networks involved in the cognitive state to be modified (Bu et al., 2019;
Faller et al., 2019; Keynan et al., 2019; Tuckute et al., 2021; Rana et al., 2020; Wang et al., 2021). Other
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 31
studies have applied micro-state analysis, which identifies short, stable, distinct patterns of electrical
activity measured by EEG across the scalp (Férat et al., 2022; Khanna et al., 2015; Michel & Koenig,
2018), representing various cognitive processes (Tarailis et al., 2024). Currently, few studies have
attempted to modulate these neural states via neurofeedback (Asai et al., 2022; Diaz Hernandez et al.,
2016), and it is necessary to assess larger samples in well-controlled protocols to highlight the potential
of these methods compared to the use of conventional brain frequencies.
Predictive variables of neuromodulation of trained signals
In this research, our second objective was to identify factors that could potentially predict the
modulation of brain signals through the use of a neurofeedback. We found that neither working memory
performance nor pre-training questionnaire scores could predict the neuromodulation of the trained
signals. Instead, the sole predictor for both trained frequencies (theta and high alpha) was the average
amplitude of the targeted frequency recorded during rest before the training. This effect, already
observed for alpha frequency (Su et al., 2021; Nan et al., 2018; Wan et al., 2014), was replicated here for
theta frequency. This effect is particularly intriguing due to the specificity of the result: neuromodulation
during training is not predicted by other brain signals or variables recorded in our study. The sole
exception to this specificity pertains to the high alpha frequency, where modulation is also predicted by
the amplitude of the resting low alpha. This result may suggest that the two sub-bands of the alpha
frequency are not totally independent, at least within the temporal scope of our study. This effect could
also contribute to an explanation for the observed increase in theta amplitude in the Control group.
Indeed, the Control group exhibited higher resting theta activity compared to the High Alpha group and
demonstrated similar activity to the Theta group, reflecting the neuromodulation pattern observed during
neurofeedback. Our study appears to validate the existing relationship between neuromodulation
capacity and resting-state signal amplitude. Thus, it may be relevant to consider using this measure as an
inclusion criterion to select a sample with a higher proportion of learners. This approach would enhance
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 32
the statistical power of studies without increasing the sample size (Thibault & Pedder, 2022). What
remains to be elucidated, however, is the mechanism behind this statistical relationship. In cognitive
training, a lower baseline level is usually associated with greater improvement in the trained variable
(Roheger et al., 2020; Traut et al., 2021). In the case of neuromodulation, our findings suggest an inverse
relationship: higher amplitude of the trained frequency seems to indicate a higher likelihood of neuronal
synchronization in the targeted brain region. One hypothesis could be that the oscillatory activity level
(i.e., amplitude) recorded at rest reflects the level of neuronal excitability in the targeted frequency (Ogata
et al., 2019; Schutter & Hortensius, 2011). To further explore this hypothesis, experimental testing could
stimulate brain activity immediately prior to neurofeedback session (Orendáčová & Kvašňák, 2021).
Conclusion
Our neurofeedback training led to electrophysiological and behavioral improvements in all groups,
including the active control group. Although these results may suggest a lack of specific neurofeedback
effect, identifying the non-specific factors contributing to neuromodulation remains complex due to the
sensory, cognitive, and affective factors involved in the self-regulation task (Lubianiker et al., 2019). A
factor that appears robustly associated with brain activity neuromodulation is the resting-state activity
recorded before training. Still, longer-term protocols with sufficient statistical power are still necessary to
draw solid conclusions about the mechanisms involved in EEG activity modulation by neurofeedback.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 33
Declarations
Funding
S.C.’s thesis was funded by the Agence de l’innovation de défense (AID) and the Ecole Nationale
de l’Aviation Civile (ENAC). The views expressed are those of the authors and not those of the AID or the
ENAC. The authors have no competing interests to declare that are relevant to the content of this
article.
Conflicts of interest/Competing interest
The authors have no relevant financial or non-financial interests to disclose.
Ethics approval
This study was performed in line with the principles of the Declaration of Helsinki. Approval was
granted by the Ethics Committee of the University Paris Cité (IRB No. = 00012019-41).
Consent to participate/Consent for publication
Informed consent was obtained from all individual participants included in the study.
Availability of data and code (Open Practices Statement)
The data and R script are available on this OSF repository:
https://osf.io/zs97e/?view_only=7d01849a74454cdfa42a09a92938cf64
References
Asai, T., Hamamoto, T., Kashihara, S., & Imamizu, H. (2022). Real-time detection and feedback of
canonical electroencephalogram microstates: Validating a neurofeedback system as a function
of delay. Frontiers in Systems Neuroscience, 16, 786200.
https://doi.org/10.3389/fnsys.2022.786200
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 34
Alkoby, O., Abu-Rmileh, A., Shriki, O., & Todder, D. (2018). Can we predict who will respond to
neurofeedback? A review of the inefficacy problem and existing predictors for successful EEG
neurofeedback learning. Neuroscience, 378, 155-164.
Başar, E., Başar-Eroglu, C., Karakaş, S., & Schürmann, M. (2001). Gamma, alpha, delta, and theta
oscillations govern cognitive processes. International journal of psychophysiology, 39(2-3), 241-
248.
Bassett, D. S., & Khambhati, A. N. (2017). A network engineering perspective on probing and perturbing
cognition with neurofeedback. Annals of the New York Academy of Sciences, 1396(1), 126-143.
Bates, D., Kliegl, R., Vasishth, S., & Baayen, H. (2015). Parsimonious mixed models. arXiv preprint
arXiv:1506.04967.
Birbaumer, N., Ruiz, S., & Sitaram, R. (2013). Learned regulation of brain metabolism. Trends in cognitive
sciences, 17(6), 295-302.
Bismuth, J., Vialatte, F., & Lefaucheur, J. P. (2020). Relieving peripheral neuropathic pain by increasing
the power-ratio of low-β over high-β activities in the central cortical region with EEG-based
neurofeedback: Study protocol for a controlled pilot trial (SMRPain study). Neurophysiologie
Clinique, 50(1), 5-20.
Bobby, J. S., & Prakash, S. (2017). Upper alpha neurofeedback training enhances working memory
performance using LabVIEW. International Journal of Biomedical Engineering and
Technology, 25(2-4), 120-132.
Brandmeyer, T., & Delorme, A. (2020). Closed-loop frontal midlineθ neurofeedback: A novel approach
for training focused-attention meditation. Frontiers in Human Neuroscience, 14, 246.
Button, K. S., Ioannidis, J. P., Mokrysz, C., Nosek, B. A., Flint, J., Robinson, E. S., & Munafò, M. R. (2013).
Power failure: why small sample size undermines the reliability of neuroscience. Nature reviews
neuroscience, 14(5), 365-376.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 35
Bu, J., Young, K. D., Hong, W., Ma, R., Song, H., Wang, Y., ... & Zhang, X. (2019). Effect of deactivation of
activity patterns related to smoking cue reactivity on nicotine addiction. Brain, 142(6), 1827-
1841. https://doi.org/10.1093/brain/awz114
Buzsaki, G., & Draguhn, A. (2004). Neuronal oscillations in cortical networks. Science, 304(5679), 1926-
1929.
Canolty, R. T., & Knight, R. T. (2010). The functional role of cross-frequency coupling. Trends in cognitive
sciences, 14(11), 506-515.
Cavanagh, J. F., & Frank, M. J. (2014). Frontal theta as a mechanism for cognitive control. Trends in
cognitive sciences, 18(8), 414-421.
Chen, X., Ma, R., Zhang, W., Zeng, G. Q., Wu, Q., Yimiti, A., ... & Zhang, X. (2023). Alpha oscillatory
activity is causally linked to working memory retention. PLoS Biology, 21(2), e3001999.
Chen, X., & Sui, L. (2023). Alpha band neurofeedback training based on a portable device improves
working memory performance of young people. Biomedical Signal Processing and Control, 80,
104308.
Chiasson, P., Boylan, M. R., Elhamiasl, M., Pruitt, J. M., Ranjan, S., Riels, K., ... & Keil, A. (2023). Effects of
Neurofeedback training on performance in laboratory tasks: A systematic review. International
Journal of Psychophysiology, 189, 42-56.
Chikhi, S., Matton, N., & Blanchet, S. (2022). EEG power spectral measures of cognitive workload: A
meta‐analysis. Psychophysiology, 59(6), e14009.
Chikhi, S., Matton, N., Sanna, M., & Blanchet, S. (2023). Mental strategies and resting state EEG: Effect
on high alpha amplitude modulation by neurofeedback in healthy young adults. Biological
Psychology, 178, 108521.
Clayton, M. S., Yeung, N., & Kadosh, R. C. (2015). The roles of cortical oscillations in sustained
attention. Trends in cognitive sciences, 19(4), 188-195.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 36
Cooper, P. S., Karayanidis, F., McKewen, M., McLellan-Hall, S., Wong, A. S., Skippen, P., & Cavanagh, J. F.
(2019). Frontal theta predicts specific cognitive control-induced behavioural changes beyond
general reaction time slowing. Neuroimage, 189, 130-140.
https://doi.org/10.1016/j.neuroimage.2019.01.022
Corsi, P. M. (1973). Human memory and the medial temporal region of the brain. Dissertation Abstracts
International, 34(2-B), 891.
Davelaar, E. J. (2018). Mechanisms of neurofeedback: a computation-theoretic
approach. Neuroscience, 378, 175-188.
Dehghanpour, P., Farokhi, F., & Rostami, R. (2018). Improvement of working memory performance by
parietal upper alpha neurofeedback training. International Journal of Smart Electrical
Engineering, 7(02), 77-81.
Deiber, M. P., Missonnier, P., Bertrand, O., Gold, G., Fazio-Costa, L., Ibanez, V., & Giannakopoulos, P.
(2007). Distinction between perceptual and attentional processing in working memory tasks: a
study of phase-locked and induced oscillatory brain dynamics. Journal of cognitive
neuroscience, 19(1), 158-172. https://doi.org/10.1162/jocn.2007.19.1.158
Dessy, E., Mairesse, O., Van Puyvelde, M., Cortoos, A., Neyt, X., & Pattyn, N. (2020). Train your brain?
Can we really selectively train specific EEG frequencies with neurofeedback training. Frontiers in
Human Neuroscience, 14, 22.
de Vries, I. E., Slagter, H. A., & Olivers, C. N. (2020). Oscillatory control over representational states in
working memory. Trends in cognitive sciences, 24(2), 150-162.
https://doi.org/10.1016/j.tics.2019.11.006
Diaz Hernandez, L., Rieger, K., Baenninger, A., Brandeis, D., & Koenig, T. (2016). Towards using
microstate-neurofeedback for the treatment of psychotic symptoms in schizophrenia. A
feasibility study in healthy participants. Brain topography, 29, 308-321.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 37
Dobrushina, O. R., Vlasova, R. M., Rumshiskaya, A. D., Litvinova, L. D., Mershina, E. A., Sinitsyn, V. E., &
Pechenkova, E. V. (2020). Modulation of intrinsic brain connectivity by implicit
electroencephalographic neurofeedback. Frontiers in human neuroscience, 192.
Domingos, C., Peralta, M., Prazeres, P., Nan, W., Rosa, A., & Pereira, J. G. (2021). Session frequency
matters in neurofeedback training of athletes. Applied Psychophysiology and Biofeedback, 46,
195-204.
Engel, A. K., Fries, P., & Singer, W. (2001). Dynamic predictions: oscillations and synchrony in top–down
processing. Nature Reviews Neuroscience, 2(10), 704-716.
Enriquez-Geppert, S., Huster, R. J., Figge, C., & Herrmann, C. S. (2014). Self-regulation of frontal-midline
theta facilitates memory updating and mental set shifting. Frontiers in behavioral
neuroscience, 8, 420.
Enriquez-Geppert, S., Huster, R. J., & Herrmann, C. S. (2017). EEG-neurofeedback as a tool to modulate
cognition and behavior: a review tutorial. Frontiers in human neuroscience, 11, 51.
Eschmann, K. C., Bader, R., & Mecklinger, A. (2020). Improving episodic memory: Frontal-midline theta
neurofeedback training increases source memory performance. NeuroImage, 222, 117219.
Eschmann, K. C., & Mecklinger, A. (2022). Improving cognitive control: Is theta neurofeedback training
associated with proactive rather than reactive control enhancement?. Psychophysiology, 59(5),
e13873.
Eschmann, K. C., Riedel, L., & Mecklinger, A. (2022). Theta neurofeedback training supports motor
performance and flow experience. Journal of Cognitive Enhancement, 1-17.
Escolano, C., Oliván, B., Lopez-del-Hoyo, Y., Garcia-Campayo, J., & Minguez, J. (2012, August). Double-
blind single-session neurofeedback training in upper-alpha for cognitive enhancement of
healthy subjects. In 2012 Annual International Conference of the IEEE Engineering in Medicine
and Biology Society (pp. 4643-4647). IEEE.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 38
Escolano, C., Navarro-Gil, M., Garcia-Campayo, J., & Minguez, J. (2014). The effects of a single session of
upper alpha neurofeedback for cognitive enhancement: A sham-controlled study. Applied
psychophysiology and biofeedback, 39, 227-236.
Esteves, I., Nan, W., Alves, C., Calapez, A., Melício, F., & Rosa, A. (2021). An exploratory study of training
intensity in EEG neurofeedback. Neural Plasticity, 2021.
Faller, J., Cummings, J., Saproo, S., & Sajda, P. (2019). Regulation of arousal via online neurofeedback
improves human performance in a demanding sensory-motor task. Proceedings of the National
Academy of Sciences, 116(13), 6482-6490. https://doi.org/10.1073/pnas.1817207116
Faul, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G* Power 3: A flexible statistical power analysis
program for the social, behavioral, and biomedical sciences. Behavior research methods, 39(2),
175-191.
Férat, V., Seeber, M., Michel, C. M., & Ros, T. (2022). Beyond broadband: Towards a spectral
decomposition of electroencephalography microstates. Human Brain Mapping, 43(10), 3047-
3061.
Gaume, A., Vialatte, A., Mora-Sánchez, A., Ramdani, C., & Vialatte, F. B. (2016). A psychoengineering
paradigm for the neurocognitive mechanisms of biofeedback and neurofeedback. Neuroscience
& Biobehavioral Reviews, 68, 891-910.
Gruzelier, J. H. (2014). EEG-neurofeedback for optimising performance. I: A review of cognitive and
affective outcome in healthy participants. Neuroscience & Biobehavioral Reviews, 44, 124-141.
Guez, J., Rogel, A., Getter, N., Keha, E., Cohen, T., Amor, T., ... & Todder, D. (2015). Influence of
electroencephalography neurofeedback training on episodic memory: A randomized, sham-
controlled, double-blind study. Memory, 23(5), 683-694.
Haegens, S., & Golumbic, E. Z. (2018). Rhythmic facilitation of sensory processing: A critical
review. Neuroscience & Biobehavioral Reviews, 86, 150-165.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 39
Hanslmayr, S., Axmacher, N., & Inman, C. S. (2019). Modulating human memory via entrainment of brain
oscillations. Trends in neurosciences, 42(7), 485-499. https://doi.org/10.1016/j.tins.2019.04.004
Hanslmayr, S., Sauseng, P., Doppelmayr, M., Schabus, M., & Klimesch, W. (2005). Increasing individual
upper alpha power by neurofeedback improves cognitive performance in human
subjects. Applied psychophysiology and biofeedback, 30, 1-10.
Haugg, A., Renz, F. M., Nicholson, A. A., Lor, C., Götzendorfer, S. J., Sladky, R., ... & Steyrl, D. (2021).
Predictors of real-time fMRI neurofeedback performance and improvement–A machine learning
mega-analysis. Neuroimage, 237, 118207.
Haynes, J. D., & Rees, G. (2006). Decoding mental states from brain activity in humans. Nature reviews
neuroscience, 7(7), 523-534.
Hothorn, T., Bretz, F., Westfall, P., Heiberger, R. M., Schuetzenmeister, A., Scheibe, S., & Hothorn, M. T.
(2016). Package ‘multcomp’. Simultaneous inference in general parametric models. Project for
Statistical Computing, Vienna, Austria. Available from http://multcomp.r-forge.r-project.org.
Hsieh, L. T., Ekstrom, A. D., & Ranganath, C. (2011). Neural oscillations associated with item and
temporal order maintenance in working memory. Journal of Neuroscience, 31(30), 10803-10810.
https://doi.org/10.1523/JNEUROSCI.0828-11.2011
Hsu, Y. F., & Hämäläinen, J. A. (2022). Load-dependent alpha suppression is related to working memory
capacity for numbers. Brain Research, 1791, 147994.
https://doi.org/10.1016/j.brainres.2022.147994
Hsueh, J. J., Chen, T. S., Chen, J. J., & Shaw, F. Z. (2016). Neurofeedback training of EEG alpha rhythm
enhances episodic and working memory. Human brain mapping, 37(7), 2662-2675.
Itthipuripat, S., Wessel, J. R., & Aron, A. R. (2013). Frontal theta is a signature of successful working
memory manipulation. Experimental brain research, 224, 255-262.
https://doi.org/10.1007/s00221-012-3305-3
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 40
Jaumard-Hakoun, A., Chikhi, S., Medani, T., Nair, A., Dreyfus, G., & Vialatte, F. B. (2017). An apparatus to
investigate western opera singing skill learning using performance and result biofeedback, and
measuring its neural correlates. Interspeech, 55, 102-111.
Jensen, O., & Mazaheri, A. (2010). Shaping functional architecture by oscillatory alpha activity: gating by
inhibition. Frontiers in human neuroscience, 4, 186.
Jeunet, C., Lotte, F., Batail, J. M., Philip, P., & Franchi, J. A. M. (2018). Using recent BCI literature to
deepen our understanding of clinical neurofeedback: a short review. Neuroscience, 378, 225-
233.
Jurewicz, K., Paluch, K., Kublik, E., Rogala, J., Mikicin, M., & Wróbel, A. (2018). EEG-neurofeedback
training of beta band (12–22 Hz) affects alpha and beta frequencies–A controlled study of a
healthy population. Neuropsychologia, 108, 13-24.
Kadosh, K. C., & Staunton, G. (2019). A systematic review of the psychological factors that influence
neurofeedback learning outcomes. Neuroimage, 185, 545-555.
Kenward, M. G., & Roger, J. H. (1997). Small sample inference for fixed effects from restricted maximum
likelihood. Biometrics, 983-997.
Kerick, S. E., Asbee, J., Spangler, D. P., Brooks, J. B., Garcia, J. O., Parsons, T. D., ... & Robucci, R. (2023).
Neural and behavioral adaptations to frontal theta neurofeedback training: A proof of concept
study. Plos one, 18(3), e0283418.
Kessels, R. P., van Den Berg, E., Ruis, C., & Brands, A. M. (2008). The backward span of the Corsi Block-
Tapping Task and its association with the WAIS-III Digit Span. Assessment, 15(4), 426-434.
Keynan, J. N., Cohen, A., Jackont, G., Green, N., Goldway, N., Davidov, A., ... & Hendler, T. (2019).
Electrical fingerprint of the amygdala guides neurofeedback training for stress resilience. Nature
human behaviour, 3(1), 63-73. https://doi.org/10.1038/s41562-018-0484-3
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 41
Khanna, A., Pascual-Leone, A., Michel, C. M., & Farzan, F. (2015). Microstates in resting-state EEG:
current status and future directions. Neuroscience & Biobehavioral Reviews, 49, 105-113.
https://doi.org/10.1016/j.neubiorev.2014.12.010
Klimesch, W. (2018). The frequency architecture of brain and brain body oscillations: an
analysis. European Journal of Neuroscience, 48(7), 2431-2453.
Klimesch, W., Sauseng, P., & Hanslmayr, S. (2007). EEG alpha oscillations: the inhibition–timing
hypothesis. Brain research reviews, 53(1), 63-88.
Kuznetsova, A., Brockhoff, P. B., & Christensen, R. H. (2017). lmerTest package: tests in linear mixed
effects models. Journal of statistical software, 82, 1-26.
Kvamme, T. L., Ros, T., & Overgaard, M. (2022a). Can neurofeedback provide evidence of direct brain-
behavior causality?. Neuroimage, 258, 119400.
Kvamme, T. L., Sarmanlu, M., & Overgaard, M. (2022b). Doubting the double-blind: Introducing a
questionnaire for awareness of experimental purposes in neurofeedback studies. Consciousness
and Cognition, 104, 103381.
La Marca, J. P., Cruz, D., Fandino, J., Cacciaguerra, F. R., Fresco, J. J., & Guerra, A. T. (2018). Evaluation of
artifact-corrected electroencephalographic (EEG) training: a pilot study. Journal of Neural
Transmission, 125, 1087-1097.
Lezak, M. D., Howieson, D. B., Loring, D. W., & Fischer, J. S. (2004). Neuropsychological assessment.
Oxford University Press.
Li, Z., Wang, H., Wu, X., Xu, X., Wei, S., & Yao, L. (2019, February). Working Memory Training Using EEG
Neurofeedback Based on Theta Coherence of Brain Regions. In 2019 7th International Winter
Conference on Brain-Computer Interface (BCI) (pp. 1-6). IEEE.
Lubianiker, N., Paret, C., Dayan, P., & Hendler, T. (2022). Neurofeedback through the lens of
reinforcement learning. Trends in Neurosciences, 45(8), 579-593.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 42
Lubianiker, N., Goldway, N., Fruchtman-Steinbok, T., Paret, C., Keynan, J. N., Singer, N., ... & Hendler, T.
(2019). Process-based framework for precise neuromodulation. Nature human behaviour, 3(5),
436-445.
Luke, S. G. (2017). Evaluating significance in linear mixed-effects models in R. Behavior research
methods, 49(4), 1494-1502.
Makowski, D., Lüdecke, D., & Ben-Shachar, M. S. (2020a). Modelbased: Estimation of model-based
predictions, contrasts and means. R package version 0.3.0. CRAN. Available from
https://github.com/easystats/modelbased.
Makowski, D., Ben-Shachar, M.S., Patil, I. & Lüdecke, D. (2020b). Automated Results Reporting as a
Practical Tool to Improve Reproducibility and Methodological Best Practices Adoption. CRAN.
Available from https://github.com/easystats/report.
Marins, T., Rodrigues, E. C., Bortolini, T., Melo, B., Moll, J., & Tovar-Moll, F. (2019). Structural and
functional connectivity changes in response to short-term neurofeedback training with motor
imagery. Neuroimage, 194, 283-290.
Michel, C. M., & Koenig, T. (2018). EEG microstates as a tool for studying the temporal dynamics of
whole-brain neuronal networks: a review. Neuroimage, 180, 577-593.
Min, S. H., & Zhou, J. (2021). Smplot: An R package for easy and elegant data visualization. Frontiers in
Genetics, 12, 802894.
Mueller, S. T., & Piper, B. J. (2014). The psychology experiment building language (PEBL) and PEBL test
battery. Journal of neuroscience methods, 222, 250-259.
Muñoz-Moldes, S., & Cleeremans, A. (2020). Delineating implicit and explicit processes in neurofeedback
learning. Neuroscience & Biobehavioral Reviews, 118, 681-688.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 43
Naas, A., Rodrigues, J., Knirsch, J. P., & Sonderegger, A. (2019). Neurofeedback training with a low-priced
EEG device leads to faster alpha enhancement but shows no effect on cognitive performance: A
single-blind, sham-feedback study. PLoS One, 14(9), e0211668.
Nan, W., Wan, F., Tang, Q., Wong, C. M., Wang, B., & Rosa, A. (2018). Eyes-closed resting EEG predicts
the learning of alpha down-regulation in neurofeedback training. Frontiers in psychology, 9,
1607.
Navarro Gil, M., Escolano Marco, C., Montero-Marín, J., Minguez Zafra, J., Shonin, E., & García Campayo,
J. (2018). Efficacy of neurofeedback on the increase of mindfulness-related capacities in healthy
individuals: a controlled trial. Mindfulness, 9, 303-311.
Nawaz, R., Nisar, H., Yap, V. V., & Tsai, C. Y. (2022). The Effect of Alpha Neurofeedback Training on
Cognitive Performance in Healthy Adults. Mathematics, 10(7), 1095.
Ninaus, M., Kober, S. E., Witte, M., Koschutnig, K., Stangl, M., Neuper, C., & Wood, G. (2013). Neural
substrates of cognitive control under the belief of getting neurofeedback training. Frontiers in
human neuroscience, 7, 914.
Niv, S. (2013). Clinical efficacy and potential mechanisms of neurofeedback. Personality and Individual
Differences, 54(6), 676-686. https://doi.org/10.1016/j.paid.2012.11.037
Ogata, K., Nakazono, H., Uehara, T., & Tobimatsu, S. (2019). Prestimulus cortical EEG oscillations can
predict the excitability of the primary motor cortex. Brain Stimulation, 12(6), 1508-1516.
Orendáčová, M., & Kvašňák, E. (2021). Effects of transcranial alternating current stimulation and
neurofeedback on alpha (EEG) dynamics: a review. Frontiers in Human Neuroscience, 15.
Owen, A. M., McMillan, K. M., Laird, A. R., & Bullmore, E. (2005). N‐back working memory paradigm: A
meta‐analysis of normative functional neuroimaging studies. Human brain mapping, 25(1), 46-
59.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 44
Pavlov, Y. G., & Kotchoubey, B. (2022). Oscillatory brain activity and maintenance of verbal and visual
working memory: A systematic review. Psychophysiology, 59(5), e13735.
Pesonen, M., Hämäläinen, H., & Krause, C. M. (2007). Brain oscillatory 4–30 Hz responses during a visual
n-back memory task with varying memory load. Brain research, 1138, 171-177.
https://doi.org/10.1016/j.brainres.2006.12.076
Pfeiffer, M., Kübler, A., & Hilger, K. (2024). Modulation of Human Frontal Midline Theta by
Neurofeedback: A Systematic Review and Quantitative Meta-Analysis. Neuroscience &
Biobehavioral Reviews, 105696.
Pimenta, M. G., van Run, C., de Fockert, J. W., & Gruzelier, J. H. (2018). Neurofeedback of SMR and
beta1 frequencies: an investigation of learning indices and frequency-specific
effects. Neuroscience, 378, 211-224.
Pillette, L., Roc, A., N’Kaoua, B., & Lotte, F. (2021). Experimenters' influence on mental-imagery based
brain-computer interface user training. International Journal of Human-Computer Studies, 149,
102603.
R Core Team. (2020). R: A language and environment for statistical computing (4.0.3). R Foundation for
Statistical Computing.
Ramot, M., & Martin, A. (2022). Closed-loop neuromodulation for studying spontaneous activity and
causality. Trends in cognitive sciences, 26, 290-299.
Ramot, M., Grossman, S., Friedman, D., & Malach, R. (2016). Covert neurofeedback without awareness
shapes cortical network spontaneous connectivity. Proceedings of the National Academy of
Sciences, 113(17), E2413-E2420.
Rana, K. D., Khan, S., Hämäläinen, M. S., & Vaina, L. M. (2020). A computational paradigm for real-time
MEG neurofeedback for dynamic allocation of spatial attention. Biomedical engineering
online, 19, 1-17. https://doi.org/10.1186/s12938-020-00787-y
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 45
Reiner, M., Rozengurt, R., & Barnea, A. (2014). Better than sleep: Theta neurofeedback training
accelerates memory consolidation. Biological psychology, 95, 45-53.
Reis, J., Portugal, A. M., Fernandes, L., Afonso, N., Pereira, M., Sousa, N., & Dias, N. S. (2016). An alpha
and theta intensive and short neurofeedback protocol for healthy aging working-memory
training. Frontiers in aging neuroscience, 157.
Richardson, J. T. (2007). Measures of short-term memory: a historical review. Cortex, 43(5), 635-650.
Riddle, J., Scimeca, J. M., Cellier, D., Dhanani, S., & D’Esposito, M. (2020). Causal evidence for a role of
theta and alpha oscillations in the control of working memory. Current Biology, 30(9), 1748-
1754.
Roheger, M., Meyer, J., Kessler, J., & Kalbe, E. (2020). Predicting short-and long-term cognitive training
success in healthy older adults: who benefits?. Aging, Neuropsychology, and Cognition, 27(3),
351-369.
Ros, T., J. Baars, B., Lanius, R. A., & Vuilleumier, P. (2014). Tuning pathological brain oscillations with
neurofeedback: a systems neuroscience framework. Frontiers in human neuroscience, 8, 1008.
https://doi.org/10.3389/fnhum.2014.01008
Ros, T., Enriquez-Geppert, S., Zotev, V., Young, K. D., Wood, G., Whitfield-Gabrieli, S., ... & Thibault, R. T.
(2020). Consensus on the reporting and experimental design of clinical and cognitive-
behavioural neurofeedback studies (CRED-nf checklist). Brain, 143(6), 1674-1685.
Ros, T., Munneke, M. A., Ruge, D., Gruzelier, J. H., & Rothwell, J. C. (2010). Endogenous control of waking
brain rhythms induces neuroplasticity in humans. European Journal of Neuroscience, 31(4), 770-
778.
Ros, T., Théberge, J., Frewen, P. A., Kluetsch, R., Densmore, M., Calhoun, V. D., & Lanius, R. A. (2013).
Mind over chatter: plastic up-regulation of the fMRI salience network directly after EEG
neurofeedback. Neuroimage, 65, 324-335.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 46
Roux, F., & Uhlhaas, P. J. (2014). Working memory and neural oscillations: alpha–gamma versus theta–
gamma codes for distinct WM information?. Trends in cognitive sciences, 18(1), 16-25.
https://doi.org/10.1016/j.tics.2013.10.010
Rozengurt, R., Barnea, A., Uchida, S., & Levy, D. A. (2016). Theta EEG neurofeedback benefits early
consolidation of motor sequence learning. Psychophysiology, 53(7), 965-973.
Rozengurt, R., Shtoots, L., Sheriff, A., Sadka, O., & Levy, D. A. (2017). Enhancing early consolidation of
human episodic memory by theta EEG neurofeedback. Neurobiology of learning and
memory, 145, 165-171.
Sampaio-Baptista, C., Neyedli, H. F., Sanders, Z. B., Havard, D., Huang, Y., Andersson, J. L., ... & Johansen-
Berg, H. (2021). fMRI neurofeedback in the motor system elicits bidirectional changes in activity
and in white matter structure in the adult human brain. Cell Reports, 37(4).
Sauseng, P., Griesmayr, B., Freunberger, R., & Klimesch, W. (2010). Control mechanisms in working
memory: a possible function of EEG theta oscillations. Neuroscience & Biobehavioral
Reviews, 34(7), 1015-1022. https://doi.org/10.1016/j.neubiorev.2009.12.006
Sauseng, P., Klimesch, W., Heise, K. F., Gruber, W. R., Holz, E., Karim, A. A., ... & Hummel, F. C. (2009).
Brain oscillatory substrates of visual short-term memory capacity. Current biology, 19(21), 1846-
1852. https://doi.org/10.1016/j.cub.2009.08.062
Sauseng, P., Klimesch, W., Schabus, M., & Doppelmayr, M. (2005). Fronto-parietal EEG coherence in
theta and upper alpha reflect central executive functions of working memory. International
journal of Psychophysiology, 57(2), 97-103. https://doi.org/10.1016/j.ijpsycho.2005.03.018
Senoussi, M., Verbeke, P., Desender, K., De Loof, E., Talsma, D., & Verguts, T. (2022). Theta oscillations
shift towards optimal frequency for cognitive control. Nature Human Behaviour, 6(7), 1000-
1013.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 47
Schad, D. J., Vasishth, S., Hohenstein, S., & Kliegl, R. (2020). How to capitalize on a priori contrasts in
linear (mixed) models: A tutorial. Journal of Memory and Language, 110, 104038.
Schutter, D. J., & Hortensius, R. (2011). Brain oscillations and frequency-dependent modulation of
cortical excitability. Brain stimulation, 4(2), 97-103.
Sghirripa, S., Graetz, L., Merkin, A., Rogasch, N. C., Ridding, M. C., Semmler, J. G., & Goldsworthy, M. R.
(2021). The role of alpha power in the suppression of anticipated distractors during verbal
working memory. Brain Topography, 34, 102-109. https://doi.org/10.1007/s10548-020-00810-4
Shen, L., Jiang, Y., Wan, F., Ku, Y., & Nan, W. (2023). Successful Alpha Neurofeedback Training Enhances
Working Memory Updating and Event-related Potential Activity. Neurobiology of Learning and
Memory, 107834
Sherlin, L. H., Arns, M., Lubar, J., Heinrich, H., Kerson, C., Strehl, U., & Sterman, M. B. (2011).
Neurofeedback and basic learning theory: implications for research and practice. Journal of
Neurotherapy, 15(4), 292-304.
Shibata, K., Lisi, G., Cortese, A., Watanabe, T., Sasaki, Y., & Kawato, M. (2019). Toward a comprehensive
understanding of the neural mechanisms of decoded neurofeedback. Neuroimage, 188, 539-
556.
Shibata, K., Watanabe, T., Sasaki, Y., & Kawato, M. (2011). Perceptual learning incepted by decoded
fMRI neurofeedback without stimulus presentation. Science, 334(6061), 1413-1415.
Shoji, Y., Patti, C. R., & Cvetkovic, D. (2017, July). Electroencephalographic neurofeedback to up-regulate
frontal theta rhythms: preliminary results. In 2017 39th Annual International Conference of the
IEEE Engineering in Medicine and Biology Society (EMBC) (pp. 1425-1428). IEEE.
Shtoots, L., Dagan, T., Levine, J., Rothstein, A., Shati, L., & Levy, D. A. (2021). The Effects of Theta EEG
Neurofeedback on the Consolidation of Spatial Memory. Clinical EEG and Neuroscience, 52(5),
338-344.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 48
Sitaram, R., Ros, T., Stoeckel, L., Haller, S., Scharnowski, F., Lewis-Peacock, J., ... & Sulzer, J. (2017).
Closed-loop brain training: the science of neurofeedback. Nature Reviews Neuroscience, 18(2),
86-100.
Smit, D., Dapor, C., Koerts, J., Tucha, O. M., Huster, R. J., & Enriquez-Geppert, S. (2023). Long-term
improvements in executive functions after frontal-midline theta neurofeedback in a (sub) clinical
group. Frontiers in Human Neuroscience, 17, 1163380.
Sorger, B., Scharnowski, F., Linden, D. E., Hampson, M., & Young, K. D. (2019). Control freaks: Towards
optimal selection of control conditions for fMRI neurofeedback studies. Neuroimage, 186, 256-
265.
Soveri, A., Antfolk, J., Karlsson, L., Salo, B., & Laine, M. (2017). Working memory training revisited: A
multi-level meta-analysis of n-back training studies. Psychonomic bulletin & review, 24(4), 1077-
1096.
Su, K. H., Hsueh, J. J., Chen, T., & Shaw, F. Z. (2021). Validation of eyes-closed resting alpha amplitude
predicting neurofeedback learning of upregulation alpha activity. Scientific reports, 11(1), 19615.
Szucs, D., & Ioannidis, J. P. (2017). Empirical assessment of published effect sizes and power in the
recent cognitive neuroscience and psychology literature. PLoS biology, 15(3), e2000797.
Takabatake, K., Kunii, N., Nakatomi, H., Shimada, S., Yanai, K., Takasago, M., & Saito, N. (2021). Musical
auditory alpha wave neurofeedback: Validation and cognitive perspectives. Applied
Psychophysiology and Biofeedback, 46(4), 323-334.
Tarailis, P., Koenig, T., Michel, C. M., & Griškova-Bulanova, I. (2024). The functional aspects of resting
EEG microstates: a systematic review. Brain topography, 37(2), 181-217.
https://doi.org/10.1007/s10548-023-00958-9
Taschereau-Dumouchel, V., Cortese, A., Lau, H., & Kawato, M. (2021). Conducting decoded
neurofeedback studies. Social Cognitive and Affective Neuroscience, 16(8), 838-848.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 49
Thibault, R. T., Lifshitz, M., & Raz, A. (2016). The self-regulating brain and neurofeedback: Experimental
science and clinical promise. Cortex, 74, 247-261.
Thibault, R. T., Lifshitz, M., & Raz, A. (2017). Neurofeedback or neuroplacebo?. Brain, 140(4), 862-864.
Thibault, R. T., & Pedder, H. (2022). Excess significance and power miscalculations in neurofeedback
research. NeuroImage. Clinical, 35, 103008.
Tran, Y., Craig, A., Craig, R., Chai, R., & Nguyen, H. (2020). The influence of mental fatigue on brain
activity: Evidence from a systematic review with meta‐analyses. Psychophysiology, 57(5),
e13554.
Tseng, Y. H., Tamura, K., & Okamoto, T. (2021). Neurofeedback training improves episodic and semantic
long-term memory performance. Scientific reports, 11(1), 17274.
Tuckute, G., Hansen, S. T., Kjaer, T. W., & Hansen, L. K. (2021). Real-time decoding of attentional states
using closed-loop EEG neurofeedback. Neural Computation, 33(4), 967-1004.
Uslu, S., & Vögele, C. (2023). The more, the better? Learning rate and self-pacing in neurofeedback
enhance cognitive performance in healthy adults. Frontiers in Human Neuroscience.
Van Diepen, R. M., Foxe, J. J., & Mazaheri, A. (2019). The functional role of alpha-band activity in
attentional processing: the current zeitgeist and future outlook. Current opinion in
psychology, 29, 229-238.
Viviani, G., & Vallesi, A. (2021). EEG‐neurofeedback and executive function enhancement in healthy
adults: A systematic review. Psychophysiology, 58(9), e13874.
Wan, F., Nan, W., Vai, M. I., & Rosa, A. (2014). Resting alpha activity predicts learning ability in alpha
neurofeedback. Frontiers in human neuroscience, 8, 500.
Wang, J. R., & Hsieh, S. (2013). Neurofeedback training improves attention and working memory
performance. Clinical Neurophysiology, 124(12), 2406-2420.
EFFECTS OF ONE SESSION OF THETA OR HIGH ALPHA NEUROFEEDBACK 50
Wang, B., Xu, Z., Luo, T., & Pan, J. (2021). EEG‐Based Closed‐Loop Neurofeedback for Attention
Monitoring and Training in Young Adults. Journal of Healthcare Engineering, 2021(1), 5535810.
https://doi.org/10.1155/2021/5535810
Weber, L. A., Ethofer, T., & Ehlis, A. C. (2020). Predictors of neurofeedback training outcome: A
systematic review. NeuroImage: Clinical, 27, 102301.
Wechsler, D. (1955). Wechsler adult intelligence scale. Archives of Clinical Neuropsychology.
Wei, T. Y., Chang, D. W., Liu, Y. D., Liu, C. W., Young, C. P., Liang, S. F., & Shaw, F. Z. (2017). Portable
wireless neurofeedback system of EEG alpha rhythm enhances memory. Biomedical engineering
online, 16(1), 1-18.
Weisz, N., & Keil, A. (2022). Introduction to the special issue of human oscillatory brain activity:
Methods, models, and mechanisms. Psychophysiology, 59(5), e14038.
Wickham, H. (2016). Data analysis. In ggplot2 (pp. 189-201). Springer, Cham.
Yeh, W. H., Hsueh, J. J., & Shaw, F. Z. (2021). Neurofeedback of alpha activity on memory in healthy
participants: A systematic review and meta-analysis. Frontiers in Human Neuroscience, 14,
562360.
Yeh, W. H., Ju, Y. J., Liu, Y. T., & Wang, T. Y. (2022). Systematic review and meta-analysis on the effects of
neurofeedback training of theta activity on working memory and episodic memory in healthy
population. International Journal of Environmental Research and Public Health, 19(17), 11037.
Zoefel, B., Huster, R. J., & Herrmann, C. S. (2011). Neurofeedback training of the upper alpha frequency
band in EEG improves cognitive performance. Neuroimage, 54(2), 1427-1431.ISO 690.
