HRV and Heart Rate During Sleep: Why the Night Is the Best Window

What happens to heart rate and HRV at night?

When you fall asleep, the heart slows down and the small beat-to-beat changes in its rhythm — heart rate variability (HRV) — change character. In non-REM sleep the balance of the autonomic nervous system shifts toward parasympathetic influence, mainly through more parasympathetic activity rather than less sympathetic activity [S1]. In a laboratory study of fourteen young healthy people over three nights, this shift happened abruptly at sleep onset [S1].

Two processes overlap. One is sleep itself and the sequence of sleep stages. The other is the body clock: heart rate and high-frequency HRV follow a circadian rhythm in wakefulness and in every non-REM stage [S3]. In the same body of work, heart rate also showed an effect of time across the night beyond the stage mix [S1]. What this page adds to the general picture of why HRV changes from day to day is the night itself: what changes within it, and why it is the usual window for tracking.

How does it work across sleep stages?

Non-REM sleep (light and deep stages; the deepest is called slow-wave sleep). Sleep onset and the move into deeper stages are associated with a shift toward greater parasympathetic modulation of the heart [S3]. Heart rate is lowest and breathing is most regular here, which is why deep sleep is the calmest stretch of the night on most recordings. Within each stage, the high- and low-frequency HRV components stayed roughly constant across the night, so much of the change over the night follows the changing mix of stages [S1].

REM sleep. Autonomic balance in REM sleep was, in general, similar to wakefulness [S1], with a shift back toward sympathetic modulation compared with deep sleep [S3]. Breathing and heart rate become more irregular.

Awakenings and stage transitions. Heart rate rises abruptly at each transition from deeper to lighter sleep and at awakenings [S2]. Short awakenings are normal and are often not remembered.

What depends on the metric. "Higher" and "lower" HRV in sleep depend on which number is computed. In a small laboratory study, SDNN — which captures slower swings as well as beat-to-beat ones — was lowest in slow-wave sleep and high in REM sleep and during awakenings [S2]. Measures that track faster, breathing-linked changes behave differently. So a stage can look "high" on one metric and "low" on another; this is a property of the metrics, not a contradiction.

What is debated. One study found that sleep stages, not the body clock, mainly shape the overnight profile [S2]; another found a clear circadian rhythm within sleep stages [S3]. Both were small laboratory studies with different designs. The practical point holds either way: a shorter, later or broken night changes the reading partly through the sleep itself.

How is nightly HRV measured?

In research, overnight HRV usually comes from the ECG channel of a polysomnogram (PSG) — the full laboratory sleep recording [S4]. Wearables instead estimate it from the pulse signal of a ring, watch or strap, and each maker chooses how to turn a night into one number:

• Oura measures HRV only during sleep; the value shown is the mean of five-minute samples taken while you sleep [S13].
• Garmin HRV Status shows an average HRV calculated from data across the entire sleep period [S14].
• Apple Watch. Apple Health records HRV as SDNN [S15]. Apple's documentation for that data type does not describe a nightly averaging window, so this page does not state one.

Other devices weight parts of the night differently, for example deep sleep or the last part of the night. Different windows, different metrics (SDNN or RMSSD) and different pulse-processing steps mean that two devices worn on the same night can disagree, and neither number transfers to a laboratory value. How a single value should be read is on interpreting HRV.

Why is the night the best window?

During sleep most of the everyday disturbances to the heart's rhythm are absent: you are not moving, eating, drinking coffee, talking or changing posture. Many experimental, demographic and environmental factors influence how HRV is measured and how reliable it is [S5], and the night holds more of them steady from one recording to the next. That is why a nightly value is a useful input for a personal trend — the reference described on HRV baseline.

This also explains why a morning reading on waking or a daytime spot check gives a different number. It is recorded in a different state, often over a few minutes, with breathing, posture and recent activity shaping it. Neither is wrong; they are different measurements, and they should be compared only with readings taken the same way. The same applies to resting heart rate and breathing rate recorded at night.

The night is steadier, not clean. Sleep stages, awakenings and the timing of sleep still move the value (see above), which is why a single night says little on its own.

What affects nightly values?

Short or broken sleep, alcohol, a hard training day, infection, stress, a late meal and the menstrual cycle can all shift a night's heart rate and HRV. Their mechanisms and evidence are set out on why HRV changes from day to day and are not repeated here.

What do sleep disorders do to night-time heart rhythm?

Sleep apnea. In obstructive sleep apnea, the airway repeatedly narrows or closes during sleep. In a large clinical series, heart rate slowed progressively at the onset of each breathing pause and then sped up abruptly when breathing resumed; this repeating pattern is mediated by the autonomic nervous system [S9]. In a small study of patients with obstructive sleep apnea, sympathetic nerve activity and blood pressure did not fall during any stage of sleep, unlike in healthy sleep [S10]. These swings add variability to the overnight heart rhythm rather than signalling rest. HRV patterns have been used to screen people for referral to a sleep laboratory [S4], but a wearable value cannot diagnose apnea. Diagnosis uses polysomnography or a technically adequate home sleep apnea test [S12]. Nasal versus mouth breathing in sleep is covered on nasal breathing.

Insomnia. It is widely assumed that people with insomnia have lower HRV, but a critical review could not confirm that HRV is reliably impaired in insomnia, because the studies differed widely and most were at high risk of bias [S11].

When to see a doctor. Loud snoring with pauses in breathing that someone else notices, waking up gasping or choking, heavy daytime sleepiness, or long-standing trouble falling or staying asleep are reasons to see a doctor or a sleep specialist. Chest pain, fainting, severe breathlessness or palpitations with dizziness need urgent care. How to bring wearable data to that appointment is covered in the ONDA report for your sleep specialist and the wider doctors and your data series.

What does the evidence show?

Established. Non-REM sleep shifts the heart toward parasympathetic influence and REM sleep looks more like wakefulness [S1] [S3] [S4]. Sleep apnea produces a cyclical slow-then-fast heart rate pattern [S9]. Recording conditions shape HRV, which is why comparable conditions are recommended (expert consensus) [S5]. Consumer devices have known limits in classifying wake during the sleep period [S7].

Context-dependent. Heart rate and high-frequency HRV follow a circadian rhythm within sleep stages [S3]. In healthy young adults in a laboratory, sleep trackers detected sleep well but their stage assessments were inconsistent with PSG [S6]; the sample was thirty-four healthy young adults over three consecutive nights [S6].

Emerging. A manufacturer-run analysis reported better sleep-stage agreement when autonomic and circadian signals were added to movement [S8]. Sustained sympathetic activity during sleep in obstructive sleep apnea comes from a small study [S10].

Debated. How much of the night-time profile belongs to sleep stages and how much to the body clock [S2] [S3]. Whether HRV is lower in insomnia [S11].

Unknown. How well any specific current wearable model reproduces laboratory overnight HRV in a given person, and how nightly HRV relates to sleep quality as a person experiences it.

What it does not tell you

"A watch accurately determines my sleep stages from HRV." Not as a general rule. In a laboratory comparison with polysomnography, the devices' sleep-stage assessments were inconsistent, and performance was worse on disrupted nights [S6]. An expert review lists misclassification of wake during the sleep period and unclear performance in some groups of people [S7]. The more favourable stage results come from studies run by device makers [S8]. Stage graphs are estimates, not a sleep study.

"High nightly HRV means good sleep." Not by itself. SDNN was high in REM sleep and during awakenings and lowest in deep sleep in a small laboratory study [S2], and the repeating heart rate swings of sleep apnea add variability rather than signal rest [S9]. A higher number can come from a more broken night on some metrics.

• No single night is a verdict. Stages, awakenings and timing move the value from one night to the next.
• Devices are not interchangeable. Windows, metrics and processing differ [S13] [S14] [S15], and the methods used in research vary as well [S4] [S5].
• A wearable is not a sleep study. It cannot diagnose sleep apnea or insomnia [S12].

In ONDA

ONDA builds its personal baseline from nightly values stored in Apple Health — from Apple Watch or another device that syncs heart data there [S16]. The window is 14 days, and ONDA compares each night with your own corridor — the average of your recent nights plus or minus one standard deviation — and flags a night only when it is at least 1.5 standard deviations outside and has changed by a minimum amount. Apple Health records HRV as SDNN, so ONDA's HRV baseline is an SDNN baseline. ONDA does not stage sleep from HRV and does not diagnose sleep apnea, insomnia or any other condition.