Abstract
Healthy humans switch seamlessly between activity states, wake up and fall asleep with regularity, and cycle through sleep stages necessary for restored homeostasis and memory consolidation each night. This study tested the hypothesis that such smooth behavioral transitions are accompanied by smooth transitions between stable neural states within the brain. A method for detecting phase discontinuities across a broad range of frequencies was created to quantify phase disruptions in the Fp-Cz EEG channel from 20 annotated sleep files. Phase discontinuities decreased with increasingly deep sleep, and increased phase discontinuity was associated with increased drowsiness, reduced deep sleep, and shorter REM sleep. A 10s phase discontinuity summary measure (the phase jump indicator) closely tracked the annotated sleep stages and enabled discrimination between short (< 10 min) and longer REM periods. Overall phase discontinuity correlated inversely with broadband EEG power, suggesting that reduced spurious signaling may facilitate increased synchronization. However, the correlation between phase discontinuity and power varied with sleep stage and age. Older individuals spent significantly more time in the Awake and Drowsy stages and less time in the deepest sleep stage and REM sleep. Interestingly, although EEG power was reduced in older individuals across all sleep stages, increased phase discontinuity only occurred in stages that showed impairment. In older patients the power vs. phase discontinuity correlation shifted to positive during drowsiness, suggesting potential deficits in cortical inhibition. These results provide evidence that phase discontinuity measures extend current capabilities for assessing sleep and may yield new insights into pathological brain states.
Significance statement Evidence continues to accumulate regarding the positive relationship between healthy sleep and brain function. Recent studies also show that more healthful sleep can be induced with timely application of non-invasive therapies. Accordingly, the ability to accurately assess sleep quality in real-time has become increasingly important. Here, a newly defined measure, referred to as phase discontinuity, enabled rapid identification of unhealthful neural patterns associated with increased drowsiness, reduced deep sleep, and early termination of REM sleep. Moreover, the measure was linked to underlying neuronal and circuit properties known to impact sleep quality. Thus, the phase discontinuity measure defined in this study provides new insight into sleep pathology and has potential implications for closed-loop therapeutic intervention.
Footnotes
The abstract has been updated to clarify the important findings.
