Alpha-EEG-Feedback and Declarative Memory

The individually dominating alpha frequency (IAF), manifested by a peak in the EEG spectrum (peak alpha frequency, PAF), has been shown to correlate with memory performance. Reduced IAF/PAF can be observed in various neurological diseases causing memory impairments (e.g. Alzheimer`s disease, traumatic brain injury, stroke). Other studies, either based on a fixed (traditional) EEG spectrum or on individually adjusted frequency ranges, showed that high power in the upper alpha band (= 10 - 12 Hz range in a fixed EEG spectrum) correlates with good cognitive performance. However, controlled and reliable studies to determine the potentials of upper alpha neurofeedback training for cognitive enhancement and an evaluation of neurofeedback as a method per se are still missing. This project intends to provide physiological and psychological evidence for the ability to voluntarily enhance upper alpha activity and thereby improve memory performance. 70 healthy subjects, aged 18 to 30 years, are randomly assigned to either a neurofeedback training group ("NFT- group", n=35) or a pseudo-neurofeedback group ("pseudo-NFT-group", n=35). All subjects are blind to the feedback condition. The "NFT-group" will be trained to selectively increase their individual upper alpha activity by receiving continuous visual feedback about the progress of their upper alpha power in 10 training sessions within 2 weeks. To control for factors such as contact, motivation and placebo effects, a control group ("pseudo-NFT- group") will be engaged in 10 pseudo-neurofeedback sessions, in which the feedback is applied independently of the upper alpha frequency in random order. After completion of NFT or pseudo-NFT sessions all subjects will be tested on a paired-associate word list. 160 semantically unrelated word pairs are presented sequentially on a monitor. In the pre-stimulus interval (before each word-pair) subjects will be instructed to apply their explicitly learned ability of selective upper alpha enhancement. Directly after learning as well as after a consolidation period of 90 minutes (nap period) subjects are retested. Performance of learning and control group are compared and influences of neurofeedack on memory consolidation during sleep will be analysed. Our study should reveal whether neurofeedback-induced upper alpha enhancement could provide a powerful tool to facilitate and optimize the encoding and retrieval of an explicit declarative memory task in young healthy individuals.

An EEG frequency component often investigated in terms of cognition is the alpha band, an oscillatory rhythm most prominent above occipital brain regions that has been characterized as being reflective of cognitive functioning. Current models concerning the role of alpha activity suggest that it plays a role in attention processing. The so-called "alpha inhibition hypothesis" proposes that alpha oscillations serve as a general mechanism for proper resource-allocation by preventing in turn task-irrelevant processes from interfering with memory consolidation. Neurofeedback (NFB) is a technique that typically uses EEG to provide information about brain activity; it trains clients to alter their brainwave patterns. Using complex algorithms to amplify, extract, and transform the raw EEG signal, selected frequency components are displayed through a visual and/or auditory user interface. Subjects then learn via operant conditioning how to control the immediate feedback they receive about their brain activity. It is assumed that this training process will lead to changes in the EEG, which in turn produce behavioral changes. Apart from its apparent therapeutic benefit, for instance in the treatment of attention-deficit disorders or epilepsy, NFB has also been claimed to enhance cognitive performance in healthy subjects. The aim of the present project was to investigate how NFB impacts on alpha activity and performance during a subsequent declarative memory task (paired-associates word test). A second aspect dealt with potential effects of NFB on sleep quality during a 90 min nap. For control, sham-feedback was employed. Generally, genuine feedback led to a significant rise in power (a parameter that reflects the number of neurons discharging synchronously) in the upper and lower alpha bands compared to sham-feedback. Likewise, significantly higher coherence - an EEG measure that can provide information on large-scale cooperation or coupling between cortical areas - was observed for the NFB group in the upper alpha band. However, both groups showed significant improved recall scores after training with regard to the paired-associates word test. Concerning sleep quality, total wake time decreased significantly after training in the neurofeedback group but not in the sham- feedback cohort. We conclude that that feedback may affect alpha domains but has no immediate impact on learning and memory. However, its influence remains inconclusive. The fact that both groups increased their word-pair test recall score after training may point to an equifinality scenario. More likely, the improvement can be explained by training effects due to the repetition of a similar task. Alternatively, NFB on other electrode sites may have led to more distinct effects. Finally, the paired-associates word test may not have been the proper choice for our specific setup to investigate declarative memory performance.