Peak Alpha Frequency: an Electroencephalographic Measure of Cognitive Preparedness

Background.

Electroencephalographic (EEG) peak alpha frequency (PAF) has been shown to correlate with a variety of phenomena, including age, memory performance in healthy and demented individuals, different emotional states, schizophrenia, anxiety, recovery from stroke, cerebral blood flow (CBF) velocity, brain oxygenation, as well as acute administration of stimulant and nootropic substances. These studies have shown that PAF varies between healthy and clinical individuals, with the latter consistently having lower PAF. Moreover, PAF varies between healthy individuals, reflecting cognitive performance, with better performance being associated with increased PAF. Finally, PAF varies within individuals both between developmental stages and between different cognitive tasks, or physiological states induced by administration of various substances.

The present study suggests that among other phenomena PAF reflects a trait or state of cognitive preparedness, using three independent datasets from healthy and brain injured individuals. Based on the preceding literature, the following hypotheses were generated. First, that PAF is an index of optimal brain function, being suppressed under traumatic brain injury (TBI). Second, that PAF is negatively correlated with time since TBI, and that it is increased after cognitive rehabilitation of individuals with TBI. Third, that PAF shows cognitive preparedness within individuals, reflecting task performance differences at different days. Fourth, that PAF is increased after tasks more in those individuals who had it lower at initial baseline, being affected by the task that forces them to correct their initial unpreparedness.

Method.

Dataset A involved EEG recordings from 15 healthy young adults before, during, and after a set of reading tasks (task duration 5-20 minutes) . Dataset B involved EEG recordings from 10 individuals with TBI and 12 healthy age and sex matched controls, before, during, and after tasks of visual and auditory attention (task duration 20 minutes) (Captain's Log, Braintrain). Dataset C involved EEG recordings from 19 healthy young adults before and after a 3-minute working memory task (WAIS-R Digit Span). In this dataset, the procedure has been repeated in two different days, so within individual differences in PAF and performance could be measured. EEG was recorded at 19 scalp electrodes using the 10/20 international electrode placement system. Average PAF for each recording was reported using the EEG Workstation 2.0 software (NovatechEEG, inc.).

Results.

PAF showed significantly lower values in individuals with TBI as compared to matched healthy controls during a post-task eyes-open baseline. PAF recorded at day 1 was significantly correlated with Digit Span performance of the same day but not with Digit Span performance of day 2. Likewise, PAF recorded at day 2 was significantly correlated with Digit Span performance of the same day but not with Digit Span performance of day 1. Moreover, PAF was significantly increased after Digit Span for those participants who had it below the sample median before the task, whereas it did not increase significantly for those who had it above the sample median. However, this was not replicated with PAF before and after reading tasks. Finally, PAF was not found to be significantly correlated to time since TBI, and it did not increase significantly after cognitive rehabilitation of individuals with TBI.

Conclusions.

As expected, individuals with brain injury had lower PAF from healthy controls, confirming the consistent direction of PAF differences between normal brain function and pathology, as is the case with other neurological or psychiatric syndromes, including stroke, dementia, and schizophrenia. Interestingly, these differences between individuals with TBI and non-clinical controls were mostly prominent during a baseline that followed a set of cognitive tasks, resembling other physiological indices that require stressing the organ to be assessed (e.g. electrocardiogram).

In addition to its sensitivity to gross brain pathology, PAF was found to be particularly sensitive to brain states within individuals during different days. PAF significantly predicted cognitive performance on a working memory task that was performed immediately after EEG recording, whereas it did not predict performance within a few days. In addition to predict performance, PAF was found to be affected by a working memory task, extending previous research and supporting a strong dual relationship between PAF and cognitive performance. The nature and duration of this state of relationship, however, needs further investigation. Although a short working memory task increased PAF in individuals who had it lower immediately before the task, a longer set of reading tasks failed to replicate this phenomenon. It is suggested that EEG normative databases include PAF into their statistical reports and that neurofeedback protocols to increase PAF are attempted to improve cognitive performance in both clinical and non-clinical populations.