Heartbeat-related activity in the anterior thalamus differs between phasic and tonic REM sleep

Abstract

Rapid eye movement (REM) sleep is a fundamental sleep state associated with diverse functions from elemental physiological processes to higher order neurocognitive functions. A growing body of research indicates that REM sleep with eye movements (phasic REM) differs from REM periods without ocular activity (tonic) in terms of spontaneous and evoked neural responses. Studies using auditory stimulation consistently observed enhanced evoked responses in tonic versus phasic REM, indicating that external processing is largely diminished when the eyes move during REM sleep. Whereas exteroceptive processing during sleep is widely studied, investigations on interoception (the processing of bodily signals) during sleep are scarce, and limited to scalp electroencephalographic recordings. Here we studied interoceptive processing in a group of epileptic patients (N = 11) by measuring their heartbeat-related neural activity in the anterior nuclei of the thalamus (ANT) during phasic and tonic REM sleep and resting wakefulness. Evoked potentials and beta-low gamma spectral power locked to the heartbeat were significantly different in phasic REM compared with tonic REM and wakefulness. Heartbeat-related neural signals exhibited pronounced inter-trial phase synchronization at lower (7-20 Hz) oscillatory activity in all vigilance states, but reduced gamma synchronization at later time points in phasic REM only. Tonic REM and wakefulness did not show significant differences in heartbeat-related activity in the ANT. Our findings indicate that heartbeat-related neural activity is detectable at the level of the ANT, showing distinct signatures of interoceptive processing in phasic REM compared with tonic REM and wakefulness. KEY POINTS: We studied interoceptive processing in the anterior the thalamus (ANT). The ANT tracks cardiac signals during wakefulness and rapid eye movement (REM) sleep. Phasic REM shows distinct patterns of heartbeat-related oscillatory activity. Interoceptive processing might be attenuated during REM periods with eye movements.

Keywords: REM sleep; anterior thalamic nucleus; heartbeat; interoception; thalamography.

Figure 1. Heartbeat‐evoked potential (HEP) in the anterior thalamus in phasic and tonic REM sleep and resting wakefulness.

A, HEP amplitudes in phasic and tonic REM and wakefulness. HEPs differed at late potentials (over ∼400 and 600 ms) between phasic REM and tonic REM, as well as between phasic REM and wakefulness. Black and green vertical lines indicate significant differences along the time axis between phasic and tonic REM, and phasic REM and wakefulness, respectively. The vertical dashed line marks the time point of the R‐peak. B, HEP amplitudes averaged along the time ranges where significant differences emerged across conditions. C, averaged ECG amplitudes in the three conditions. No significant differences were observed between the three vigilance states in ECG amplitudes. D, ECG amplitudes averaged along the time ranges where significant HEP differences emerged between conditions. No significant differences in ECG amplitudes over the time range of interest emerged, indicating that HEP differences were not confounded by ECG activity. [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 2. Power modulations in the anterior thalamus locked to the R‐peak of the ECG signal.

A, changes in time‐frequency power relative to the baseline period (between ‐200 and ‐50 ms) are visualized for phasic REM, tonic REM, and wakefulness. Values above 1 indicate increased power, whereas values below 1 mean reductions in power relative to the baseline. B, pairwise statistical comparisons across conditions highlighting the value of the statistical test (T value) along the time and frequency axis. C, uncorrected significant differences (P < 0.05) at each time × frequency point and significant clusters remaining after the correction for multiple comparisons are visualized (with red contour). Power changes locked to the R‐peak in the beta and low gamma range (∼20–40 Hz) were relatively reduced in phasic REM as compared with wakefulness and tonic REM sleep. The vertical dashed line marks the time point of the R‐peak. [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 3. Inter‐trial phase coherence (ITPC) in the anterior thalamus relative to the R‐peak of the ECG signal.

A, changes in ITPC relative to the baseline period (between ‐200 and ‐50 ms) are visualized for phasic REM, tonic REM and wakefulness. Values above 1 indicate increased power, whereas values below 1 mean reductions in power relative to the baseline. B, pairwise statistical comparisons of ITPC changes across conditions highlighting the value of the statistical test (T value) along the time and frequency axis. C, uncorrected significant differences (P < 0.05) at each time × frequency point are visualized. Changes in ITPC locked to the R‐peak were not significantly different across conditions. [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 4. Changes in inter‐trial coherence (ITPC) in the anterior thalamus locked to the R‐peak.

A, statistical comparisons of ITPC between the baseline period (‐200–‐50 ms) and the post‐baseline periods (50–650 ms) in phasic REM, tonic REM, and wakefulness. Statistical parameters (T value) of within‐state comparisons along the time × frequency axis are highlighted. B, significant clusters at each time × frequency point after cluster‐based permutation statistics are highlighted. [Colour figure can be viewed at wileyonlinelibrary.com]