## Project

# EEG alpha-theta (cross-frequency) dynamics during arithmetic performance, mind wandering and meditative states

Neural oscillations have been shown to be functionally relevant for human behaviour. In this way, brain rhythms oscillating at different frequencies have been associated to different cognitive functions. Consequently, the interplay between brain rhythms in different frequency ranges (i.e. cross-frequency coupling) is thought to be essential for cognition to emerge. This thesis focuses on cross-frequency dynamics between EEG alpha (8-14 Hz) and theta (4-8 Hz) rhythms during different cognitive states. Specifically, it investigates the functional relevance of alpha:theta cross-frequency numerical ratios during arithmetic performance, mind wandering and meditative states. This approach to alpha-theta cross-frequency dynamics is based on a recent theory positing that the formation of different cross-frequency numerical ratios between the peak frequencies of two brain rhythms is reflective of their level of interaction. In this way, it is proposed that harmonic cross-frequency arrangements (e.g. 2:1 numerical ratio) would enable cross-frequency coupling. The rationale behind this premise is that only harmonic cross-frequency ratios allow stable and regular excitatory phase meetings between two neural populations (i.e. coincidence of time periods in which spiking is more likely to occur).

The working hypothesis in this thesis was that alpha and theta rhythms would arrange more often in harmonic positions during both working memory tasks and mind wandering. This hypothesis was based on previous literature suggesting that alpha (8-14Hz) and theta (4-8 Hz) rhythms reflect different components of working memory that need to be integrated when information has to be stored and manipulated in the brain. This hypothesis was assessed throughout the four studies encompassing this thesis. In the first study (Chapter 2), we assess the incidence of different alpha:theta cross-frequency ratios during an arithmetic task with a strong working memory component, rest and meditation practice. It was shown that alpha and theta rhythms separated in the frequency domain during arithmetic performance (relative to rest and meditation) thereby increasing the incidence of cross-frequency numerical ratios between 2 and 3 (and therefore 2:1 and 3:1 phase synchrony). These changes were accompanied by a decreased occurrence of ratios between 1 and 1.6. Interestingly, the separation between alpha and theta rhythms in the frequency domain (i.e. increased occurrence of ratios around 2 and 3 and decreased occurrence of ratios between 1 and 1.6) were positively associated to arithmetic performance, thereby underlining their functional relevance. A similar pattern of results was observed in the second study (Chapter 3), in which the same paradigm was adopted but with participants that were highly experienced in meditation practice. In addition to the previously reported changes during arithmetic task (relative to rest and meditation practice), we report that meditation was associated to a decreased incidence of alpha:theta ratios between 2 and 3 and an increased incidence of alpha:theta ratios between 1 and 1.6 when compared to rest and arithmetic. Based on these latter results, it was speculated that these changes in alpha-theta cross-frequency dynamics could be attributed to reduced mind wandering during meditation. To further investigate the influence of meditation training in alpha:theta cross-frequency dynamics, we assessed in a third study (Chapter 4) whether the compliance to a meditation training course was significantly correlated to changes in the incidence of different alpha:theta cross-frequency ratios during meditation. In this way, we show that meditation training (i.e. minutes of attendance plus minutes of practice at home) was associated to an approximation of alpha and theta rhythms in the frequency domain (i.e. a decreased occurrence of alpha:theta cross-frequency ratios around 3 and an increased occurrence of cross-frequency ratios around 1.6). In line with the previous study, we speculated that these inter-individual differences in the incidence of different alpha:theta numerical ratios during mediation were associated to mind wandering. Finally, in the fourth study (Chapter 5) we directly assess whether the occurrence of different alpha:theta cross-frequency ratios were associated to mind wandering in the context of meditation practice. For this purpose, a sample of novice meditators were repeatedly interrupted during a breath focus meditation to report whether they were mind wandering or focusing on their breath. In line with previous findings, our results showed that mind wandering is associated to a separation of alpha and theta rhythms in the frequency domain (i.e. an increased incidence of alpha:theta ratios between 2 and 3 at the expense of the occurrence of alpha:theta ratios between 1 and 1.6).

Together, our results consistently show a separation of alpha and theta rhythms in the frequency domain (higher mean alpha:theta numerical ratio) during both arithmetic performance (relative to rest) and mind wandering (in the context of meditation practice). Although these changes in alpha:theta cross-frequency ratios led to greater 2:1 and 3:1 harmonicity and phase synchrony between alpha and theta rhythms, the observed changes in the frequency architecture (as indexed by the incidence of different cross-frequency ratios) do not unequivocally reflect changes in the level of interaction between alpha (8-14Hz) and theta (4-8 Hz) rhythms. Therefore, we cannot conclude that here studied cognitive states involve different levels of communication between the neural populations that are entrained by neural oscillations in the alpha and theta range. In this way, changes in the incidence of different ratios cross-frequency ratios would remain descriptive and open to interpretation until future studies empirically disentangle whether: i) alpha and theta rhythms encompass two separate neural oscillations with exclusively sinusoidal properties and ii) different cross-frequency ratios reflect different levels of information exchange between neural rhythms. Regardless the interpretation we give to the incidence of different alpha:theta ratios, the here presented studies suggest the existence of a neurocognitive mechanism that supports both working memory task performance and mind wandering.