Neural dynamics necessary and sufficient for transition into pre-sleep induced by EEG NeuroFeedback
Introduction
State of mind transitions, such as when one shifts their focus from the external world inward, are a common daily occurrence that manifests most strikingly as one falls asleep. Such transition may also occur spontaneously during mind wandering or when willfully regulating relaxation. Disturbance in sleep onset is prevalent among individuals suffering from depression or anxiety disorders (Hamilton, 1989, Neylan et al., 1998). However, healthy individuals are also prone to such difficulties, when experiencing daily concerns and tension (Augner, 2011) or as a result of aging (Foley et al., 1995).
The transition into pre-sleep is well defined by an EEG-based marker of a decline in the alpha amplitude followed by an increase in theta power while alpha remains low (De Gennaro et al., 2001, Hori et al., 1994). The time at which theta becomes greater than alpha is referred to as the theta/alpha (T/A) “crossover” period and is assumed to indicate reduced vigilance and consciousness during the transition into a deep relaxation/pre-sleep state (Johnson et al., 2013, Peniston et al., 1993). This EEG marker of shifts in wakefulness has become a complementary measure for researchers using metabolic based imaging techniques such as fMRI and PET to indicate the transition into sleep. The modulation in EEG characteristics allows one to distinguish between arousal states revealing changes in the activity among large brain areas. Using this approach, fMRI and PET studies have found increased activity in the anterior cingulate cortex, the parietal cortices, and the temporal cortices (Olbrich et al., 2009), as well as in the bilateral hippocampus (Picchioni et al., 2008), while decreased activity was found in the frontoparietal cortices, the thalamus lobes (Kjaer et al., 2002, Olbrich et al., 2009), and the cerebellum (Kjaer et al., 2002). Although brain imaging studies generally indicate that many different brain regions are involved in the mental transition into pre-sleep, it is not yet clear which core neural network is necessary for such a transition while also taking into account on-going modulations of EEG markers.
The aim of the current study was to unveil the brain dynamics underlying this transition, using a well-established T/A EEG neurofeedback (EEG-NF) protocol (Peniston and Kulkosky, 1991). It has been repeatedly demonstrated that people can be trained to modulate their T/A ratio, yielding both physiological and psychological benefits (Hammond, 2011, Sokhadze et al., 2008). We therefore asserted that the T/A-EEG-NF training procedure can be used to investigate the trajectory of the mental transition into pre-sleep in a controlled fashion and within a short time period of a few minutes. For the validation of the reduced vigilance state we used heart rate variability (HRV) analysis. Vagal activity which acts to lower the heart rate was found to be a major contributor to the high-frequency (0.15 to 0.4 Hz) component of the power spectrum of heart rate variability (HR–HF)(Malik, 1996). Elevation of the HR–HF index (also referred to as parasympathetic HRV) has been linked to entering a state of relaxation (Malik, 1996, Malik, 2007) and early sleep stages (Calcagnini et al., 1994). Accordingly we assumed that there would be an increase in HR–HF power as individuals enter the pre sleep stage.
Simultaneous recording of fMRI provided high spatial resolution for the identification of the distinct brain network associated with the mental transition into pre-sleep. On the basis of previous imaging studies of arousal and attention, we hypothesized that brain areas related to external and internal monitoring and awareness would be essential during the initial stage of transition into sleep, possibly marked by the crossover time point. The thalamus in particular has been consistently found to be a key structure in relaying sensory signals and regulation of levels of attention and arousal states (Fiset et al., 1999, Ward, 2011). In addition, alpha rhythm has been repeatedly demonstrated as correlating with the thalamus activity as demonstrated in simultaneous combined imaging studies (Ben-Simon et al., 2008, Schreckenberger et al., 2004). Taken together we therefore expect that reduced EEG alpha power will be manifested in reduced thalamus fMRI activity. On the other hand, limbic/paralimbic medial and lateral temporal regions are suspected to be involved in the occipital theta modulation post crossover point. This is based on EEG studies showing that occipital theta is modulated specifically during the transition into pre-sleep (Peniston and Kulkosky, 1991) as well as sensitive to the processing of emotional stimuli (Aftanas et al., 2001, Uusberg et al., 2014).
Our findings show that T/A EEG-NF induces a state of pre-sleep that corresponds with an increased high-frequency heart rate variability (i.e. parasympathetic). In addition, successful training sessions portrayed distinct changes in fMRI activation related to pre- and post-crossover point, induced by either EEG alpha or theta modulations.
Section snippets
Subjects
45 healthy subjects aged 24–37 years (22 males) signed an informed consent form approved by the ethical committees of the Tel Aviv Sourasky Medical Center and participated in a two-stage NF experiment; T/A EEG-NF training outside the MRI scanner and two sessions of T/A EEG-NF inside the MRI scanner.
EEG-NF practice outside the scanner
This experimental stage was designed to enable the subjects to become familiar with the neurofeedback procedure and setup. Participants were given a set of headphones (Trust International, Dordrecht,
Preprocessing
Matlab (Mathworks Inc, Natick, MA) and EEGLAB ((Delorme and Makeig, 2004)) were used for all calculations. Removal of MR gradients and cardio ballistic artifacts included a FASTR algorithm (Niazy et al., 2005) implemented in the FMRIB plug-in for EEGLAB. To reduce computational complexity the EEG signal was subsequently down-sampled to 250 Hz. Time frequency representation of the EEG was calculated using the Stockwell transform (Stockwell et al., 1996) with a time resolution of 1/250 s and a
Results
Simultaneous fMRI and EEG-neurofeedback were used to initiate and explore the mental state transition from wakefulness to pre-sleep. The purpose of this study was to unfold the core networks involved in the transition and, specifically, to pinpoint the essential neural dynamics for the change. Thus, we chose to contrast successful vs. unsuccessful transitions into pre-sleep sessions identified by the theta/alpha (T/A) ratio crossover time point.
Discussion
Using fMRI simultaneously with EEG-NF provided a unique opportunity to explore and unfold the various neural dynamics that underlie successful transition from wakefulness to pre-sleep. T/A crossover was used to categorize sessions as successful or unsuccessful while an increased HR parasympathetic index verified the relaxed state of the successful ones. The fMRI activity suggests that the mental transition into pre-sleep requires two-staged interleaved processes. The initiation of transition
Conclusions
Using EEG Theta/Alpha NeuroFeedback (T/A-NF) along with simultaneous fMRI, we identified four different periods that designated the neural dynamics of the transition into pre-sleep. We found that pre-sleep initiation depends on reduced activation in sub cortical regions involved in sensory gating (e.g. medial thalamus). In contrast, for sustainment of the pre-sleep state, opposite activation of anterior versus posterior salience network was necessary. This opposition possibly stands for
Acknowledgments
Support for this research was provided by the U.S. Department of Defense award number W81XWH-11-2-0008 and from the Israeli Defense Forces Medical Corps and Israeli Ministry of Defense Funding Program for Research in Military Medicine. We thank A. Solski for assistance in manuscript preparation.
References (52)
- et al.
Affective picture processing: event-related synchronization within individually defined human theta band is modulated by valence dimension
Neurosci. Lett.
(2001) - et al.
Human parahippocampal activity: non-REM and REM elements in wake–sleep transition
Brain Res. Bull.
(2005) - et al.
The boundary between wakefulness and sleep: quantitative electroencephalographic changes during the sleep onset period
Neuroscience
(2001) - et al.
EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis
J. Neurosci. Methods
(2004) - et al.
Contribution of the cerebellar vermis to cardiovascular control
J. Auton. Nerv. Syst.
(1996) - et al.
Modulation of the brain's functional network architecture in the transition from wake to sleep
Prog. Brain Res.
(2011) - et al.
Monitoring, metacognition, and executive function: elucidating the role of self-reflection in the development of self-regulation
Adv. Child Dev. Behav.
(2011) - et al.
Procedural learning and sleep hippocampal low frequencies in humans
Neuroimage
(2008) - et al.
Removal of FMRI environment artifacts from EEG data using optimal basis sets
Neuroimage
(2005) - et al.
EEG-vigilance and BOLD effect during simultaneous EEG/fMRI measurement
Neuroimage
(2009)