How does the autonomic nervous system's circadian regulation (sympathetic vs. parasympathetic tone) modulate the perception of LBBAP pacemaker-related discomfort throughout the night, and what role do sleep stages (e.g., REM, NREM) play in this modulation?
The honest starting point, again: no study measures LBBAP-related discomfort as a function of autonomic phase or sleep stage. The specific mapping this question asks for has not been investigated. But it can be assembled from adjacent, well-established evidence — and doing so exposes a distinction the question almost forces into the open: there are really two things happening at night, and only one of them is what most people mean by "discomfort." First, there is what the heart and autonomic system actually do across the night, which is genuinely sleep-stage- and circadian-dependent. Second, there is perception — and perception is itself gated by sleep stage, because for most of the night you are not conscious to perceive anything at all. Most of what matters here is not specific to LBBAP.
The autonomic state of your heart at 3 AM is set by two overlapping timers. The circadian clock imposes a roughly 24-hour rhythm that is independent of sleep itself: in a 27-day forced-desynchrony protocol, vagal (parasympathetic) tone was maximal during the circadian phase corresponding to the usual sleep episode, with an acrophase around 4–5 AM, while sympathovagal balance was at its minimum in the late morning.[1] Layered on top is the sleep-stage clock — the ultradian cycle of NREM and REM that repeats roughly every 90 minutes and shifts across the night, with deep NREM front-loaded into the first half and REM periods lengthening toward morning.
During non-REM sleep the sympathovagal balance shifts toward parasympathetic dominance; heart rate and its variability settle, and this vagal drive deepens with sleep depth from light N1/N2 into slow-wave N3.[2] This is the calmest the autonomic system gets — a low, steady, vagally-braked heart.
REM sleep is the opposite. Direct microneurographic recording in healthy subjects showed that REM is associated with profound sympathetic activation, with bursts of sympathetic nerve traffic that can reach roughly twice waking levels, and correspondingly erratic heart rate and blood pressure.[3] Frequency-domain heart-rate-variability work confirms the same picture: a shift back toward sympathetic modulation and instability during REM compared with the parasympathetic quiet of NREM.[2] REM has therefore been described as a physiologically vulnerable window in which nocturnal arrhythmias can surface.
Translate those two clocks onto a device that is watching a lower rate limit and you get a predictable, if unproven, pattern. Under the deep vagal brake of NREM and the circadian vagal peak toward 4–5 AM, the intrinsic rate falls; more beats drop below the programmed lower rate, and so the proportion of the night that is paced tends to rise unless a lower sleep/rest rate is programmed. If any felt sensation tracks with pacing itself, this is the phase where there is the most of it. REM cuts both ways: the sympathetic surges raise the intrinsic rate (potentially reducing pacing dependence) but increase ectopy and instability, and — importantly — REM-related bradyarrhythmias and even sinus arrests unrelated to sleep apnoea have been described in otherwise healthy hearts, a reminder that the autonomic swings of REM are not uniformly "faster."[4]
One device-adjacent fact is worth keeping in view because it is easy to misattribute: the ventricular pacing threshold does vary diurnally, but not through autonomic tone — the study that looked directly at this found diurnal variation in threshold with no relationship to the circadian autonomic pattern, and the programmed safety margin normally absorbs it.[5] So "my threshold is a bit different at night" is real, but it is not the autonomic story, and it is not usually something you would feel.
Here is the reframing. All of the above describes what the heart does. Whether any of it is perceived as discomfort depends on a second gate that is itself controlled by sleep stage: during consolidated NREM and REM you are largely unconscious and are not monitoring your heartbeat at all. The felt experience of a pacemaker at night is therefore not a continuous stream — it is concentrated at the seams: sleep onset, brief arousals, the awakenings that punctuate normal sleep, and the transition to morning wakefulness. Palpitation is fundamentally an awareness of the heartbeat; its sensory mechanism is poorly understood, and rest and nighttime typically reflect a heightened awareness of an otherwise normal rhythm rather than a new arrhythmia.[6]
This is what gives REM its outsized role in perceived discomfort — not because REM is LBBAP-specific, but because REM is the stage from which you are most likely to wake with vivid awareness, and an arousal out of REM deposits you into a high-sympathetic, faster-heart state that is very easy to attribute to the device. The same beat that passed unnoticed in slow-wave sleep becomes salient the moment a REM arousal makes you conscious of it. Add the non-specific amplifiers of any quiet night — supine posture increasing venous return so each beat is felt more forcefully, a left-lateral position bringing the heart nearer the chest wall, and a dark silent room stripping away the daytime signals that normally mask cardiac sensation — and most of the "night-long discomfort" resolves into a handful of perceptual windows, each colored by whichever autonomic state you happened to surface from.
The two clocks do not peak independently. The last third of the night stacks several factors in the same place: the circadian vagal acrophase near 4–5 AM (more pacing), the lengthening REM periods of early morning (more sympathetic surges and more arousals), and the accumulating pressure to wake. That convergence — more pacing and more arousals landing in the same pre-dawn window — is the most plausible candidate for a "worst" period, and it fits the common patient report of the small hours being the least comfortable. But every clause of that is mechanistic reasoning from adjacent data, not a measured LBBAP finding. The honest status is "plausible, unproven."
The tractable, patient-runnable version is to triangulate three data streams over a couple of weeks: a timestamped symptom diary (when discomfort is noticed, and on waking from what); device interrogation data (nocturnal versus daytime pacing percentage, rate histograms, threshold/impedance trends, any recorded episodes); and, if available, a wearable that estimates sleep stages and heart-rate variability. Two questions then become answerable that pure reasoning cannot settle: does the discomfort coincide with pacing at all, and does it cluster in the last third of the night or around REM arousals? If the timestamps do not line up with pacing episodes, the device is not the driver, and the target shifts to sleep architecture, arousals, posture, and interoception — none of which are fixed by reprogramming the pacemaker, but several of which are modifiable in other ways worth discussing with the device clinic.