EEG Theta-Beta Ratio Neurofeedback-training

Overview

This study will investigate the effects of an EEG theta/beta ratio Neurofeedback training using a case series multiple baseline design. Twelve participants will be assigned to one of three different groups that have a different time of intervention onset, or include placebo sessions only. The investigators will examine whether the theta/beta ratio Neurofeedback training will reduce resting state theta/beta ratio. As a pilot-exploration for future studies, also performance on a cognitive interference task will be monitored to see if any improvement after NFT on cognitive performance might be a viable subject for future studies. All participants (healthy, but selected for above-median theta/beta ratio in previous studies in this lab) will start with three resting state EEG baseline sessions, the first participant group then continues with 14 theta/beta ratio Neurofeedback sessions, the second group continues with six extra baseline EEG measurements, followed by eight Neurofeedback training sessions while the third group will continue with 14 placebo training sessions only. EEG resting state theta/beta ratio and interference performance will be measured during every session.

Background Resting state EEG measures can produce useful spectral indices of power in different frequency bands. One robust finding in attention-deficit/ hyperactivity disorder (ADHD) has been increased theta band power (typically 4-7 Hz) relative to beta band power (13-30 Hz) otherwise expressed as the 'theta beta ratio' (TBR; Barry, Clark \& Johnstone, 2003). TBR possibly reflects interactions between subcortical and prefrontal cortical areas in the brain (Schutter \& Knyazev, 2012). A recent study from the Leiden University lab showed that increased TBR was related to less self-reported attentional control and correlated to a stronger decline in attentional control after stress-induction (Putman et al., 2014) which are core elements of cognitive performance anxiety (CPA); a phenomenon defined as when fears compromise an individual's capacity to execute a task. Given that theta/beta ratio might serve as a marker for PFC-mediated attentional control and vulnerability to the effects of CPA, the eventual goal of our research program is to test whether theta/beta ratio reduction might be effective in reducing symptoms of CPA. Neurofeedback training is a method to induce changes in EEG frequency power bands. Online feedback on people's EEG spectrum while asking to increase or decrease power in certain frequency bands can eventually lead to the ability to do this (Vernon, 2005). Application of a TBR Neurofeedback training with the aim to reduce TBR might be effective in reducing attentional dysfunction as in ADHD patients (Lofthouse et al., 2012) and might possibly increase resilience to deleterious effects of stress on cognition (Putman et al., 2014). The investigators would ultimately like to study effects of this training in individuals with elevated vulnerability for CPA; having a higher than average TBR. The primary goal of the current study is only to see if it is possible to change the EEG TBR signature in this population by means of our NFT protocol. Additionally, we would like to explore the effects on prefrontal mechanisms regulating attention to interfering distractors by administering a letter string interference task (Bishop, 2009) on fixed moments during the training. Data pertaining to this cognitive measure will only be used to inform design of future studies and any finding will be mentioned in publications explicitly as being exploratory (if at all) and not providing evidence to accept or reject any hypotheses about effects of our NFT protocol on cognitive performance. Our design will be different from the widely used randomized controlled trials design (RCT). RCT designs are generally considered to be superior to other methods in estimating an interventions true effect. However other designs such as case studies or well-matched comparison group - studies may provide information that is just as valuable in clinical intervention studies in the case that the intervention provides compelling evidence of its effect. The current study will serve as a first try-out whether we can reduce TBR using a Neurofeedback training for which an RCT design might unnecessarily cost a considerable lot of time and effort. We propose an alternative to the RCT design, namely using smaller groups of participants that start with EEG resting state measurements as a baseline and switch to real Neurofeedback training on different points in time; so called 'case series multiple baseline design'. Practical advantages of the multiple baseline design are that the design requires fewer participants and that participants act as their own controls. Visual inspection and single-case statistics will be used to determine the effectiveness of this Neurofeedback training by observing differences of TBR per group over time (see section 'Statistical analysis plan'). Aims, research question and hypothesis The main aims of this study will be to evaluate the effects of a Neurofeedback training on reducing EEG theta/beta ratio in individuals having elevated baseline TBR. Furthermore, with a very second degree of importance, the investigators will test if a reduction in TBR increases cognitive performance on a response conflict task. The research questions will be; can a TBR Neurofeedback training in healthy (non-ADHD) people with elevated TBR reduce TBR and (very much secondary) will the training perhaps increase cognitive performance on a response conflict task? If so we might continue with further studies to assess NFT effects on cognitive performance in CPA in healthy people. It is expected that TBR will reduce after training onset. Also, it is expected that performance on the interference task will improve after participants have started with the TBR Neurofeedback training. Design The investigators will use a single blind case series multiple baseline design with a baseline varying prior to training onset. By randomizing 12 participants over three conditions of different training onset points in time, the investigators will use a variation in baseline which offers the possibility to differentiate between time effects of the Neurofeedback training. All participants will start with a three-session baseline phase which will include a session with questionnaires, interference task training (in order to quickly reach a learning effect ceiling to better be able to observe additional performance change after intervention) and a resting state EEG measurement followed by two other sessions of resting state EEG measurements and the interference task. Participants in group A will then continue with 14 sessions of TBR Neurofeedback training. Group B will receive six extra baseline EEG measurements before they continue with eight TBR Neurofeedback sessions. Group C on the other hand will only receive 14 placebo training sessions after the short baseline phase. Before every TBR Neurofeedbacktraining or placebo training, a four-minute EEG resting-state baseline measurement will be performed. The minimal number of real training sessions (eight) are necessary for making it possible to observe the onset of an effect of the Neurofeedback training, based on previous studies using a TBR Neurofeedback training (e.g. Studer et al., 2014; Doppelmayr \& Weber, 2011; Leins et al., 2007; Gevensleben et al., 2009). Also, a higher number of training sessions is important to ensure that a possible delayed effect of the Neurofeedback training can still be observed. Procedure Before the first lab session, the participants will be stratified randomly assigned to one of the three groups. All participants will visit the lab 17 times. During the first visit, participants will fill in an informed consent, demographic information and questionnaires. They will continue with a training for the letter string interference task to avoid learning effects. At the end of the first visit, EEG resting state will be measured. During the second and third visit, participants will only receive a resting state EEG measurement and perform the letter string interference task. In the sessions after these baseline measurements, participants will receive a 4 minute EEG eyes-open resting state measurement, following either TBR Neurofeedback training or a placebo training depending on the participant group. After the Neurofeedback training, resting state measurement or placebo training, participants will always perform the letter string interference task. Sessions will be (almost) daily and approximately take between 50-60 minutes; except for the first questionnaires/interference training session which will take 90 minutes. The complete experiment will therefore take 17-19 hours per participant in 4-6 weeks. Materials and training Questionnaires The participants will complete demographic information and fill in the trait and state version of the State-Trait Anxiety Inventory (STAI-t, STAI-s; Spielberger, 1983; Van der Ploeg et al., 1981) adapted to only report about the past week before the first session, and the trait and state version of the Attentional Control Scale (ACS; Derryberry \& Reed, 2002). The STAI-t assesses trait anxiety (range 20-80), STAI-s assesses the participant's current state of anxiety, and the ACS assesses attentional inhibition, attentional shift, and the capacity to flexibly generate new thoughts (range 20-80). The internal consistency of these questionnaires is generally good. STAI- t, ACS-t and a demographic questionnaire will only be measured during the first session and during the last session (after the training). The STAI-s and ACS-s will be assessed on every session before the EEG measurement or training. Letter string interference task The Letter String interference task will be as in Bishop (2009) with the only difference that in this study the neutral trials will be left out as they have no informative value on cognitive interference. Neurofeedback theta/beta ratio training. The TBR Neurofeedback signal will be given by means of a NeXus-4 amplifier and recording system (Mindmedia, The Netherlands). One Ag/AgCl disposable electrode will be applied on the participant's scalp between locations Cz and FCz. For a high-quality and accurate EEG measurement, 16 extra in-cap electrodes (BioSemi, The Netherlands) will be added on locations F3, Fz, F4, C3, Cz, C4, P3, Pz, P4, T7, TP7, T8, TP8, O1, Oz and O2) during every session. Each session will consist of a four-minute resting state measurement, followed by a 25 minute measurement/Neurofeedback training or placebo training. The thresholds will be set 20% above (beta) or below (theta) the participant's amplitude, calculated and adjusted automatically every 15 seconds. During the training, theta and beta activity will be visualized in separate bar graphs on the screen and participants will be instructed to "try to move down the theta activity and move up the beta activity below or above the threshold". No instructions about how to influence their activity will be given. When the EEG theta power goes below the threshold, and the beta power above its threshold simultaneously, the participant will be rewarded by the continuation of a video. If they fail to reach this point, the video will be paused. The placebo training will be a previously recorded Neurofeedback session of a participant in group A (receiving 14 real Neurofeedback sessions). Before the participants in group C (placebo trainings) start, they will be matched to a participant in group A, and receive the exact same (prerecorded) feedback per session of their matched participant. In this way, participants in the placebo training are then made to think that the video and EEG feedback is their own, however they are actually not able to influence the continuation of the video. Data analysis For interpreting the results, visual inspection will be primarily used to determine the effectiveness of the TBR Neurofeedback training. If TBR in group A reduces after the introduction of the real Neurofeedback training compared to no changes in group C, the experiment provides compelling evidence for the effectiveness of the Neurofeedback training. The effect will be even more strongly supported if we see a similar reduction after Neurofeedback training onset in group A and B (as in after an equal number of actual NFT sessions as in group A, regardless the longer duration of the baseline). These changes after onset in group A and group B are assumed to be absent in group C, where we expect no changes at all. The expected effect in group B can also be considered as a direct replication of the effect in group A. The idealized expected changes in TBR are depicted in Figure 3 (each individual line represents changes in TBR per group). These visual inspections serve as testing of the primary outcome measures. In addition to the primary test by visual inspection, secondary, we will perform statistical tests between the average TBR of group A and B together on an equal number of sessions after training onset compared to group C on the same number of sessions as group A (i.e. session 10 for group A averaged with session 14 for group B compared to session 10 for group C). Also, we will calculate eight separate contrast scores of average TBR within group A of session 1 minus session 17; session 2 minus session 16; session 3 minus session 15 etc. The same will be done for group B and C followed by a statistical comparison of the contrast scores of group A and B (taken together as one group) with the contrast scores of group C. Tertiary, only in the case that in group A there seems to be evidence of change in TBR at approximately the same session in time as per visual inspection,, we will test a direct replication of this effect in group B using a similar permutation testing method as Ter Kuile et al., (2009, see also Bulté and Onghena, 2008 ) The approximate time-point of TBR change in group A might serve to designate the per-session TBR data as either pre- or post-intervention.