Your heart rate naturally changes throughout the night.
As you move from wakefulness into non-rapid eye movement (NREM) sleep, heart rate usually slows and becomes more stable. During deeper NREM sleep, cardiovascular activity often reaches some of its lowest and calmest levels of the night. During rapid eye movement (REM) sleep, heart rate can become faster and more variable again.
These are general physiological patterns. They do not mean every heart-rate dip proves you entered deep sleep or every spike proves you were in REM.
Sleep stage, circadian timing, brief awakenings, movement, breathing, exercise, alcohol, stress, illness, and other factors can all influence your nighttime heart-rate curve.
This guide explains how heart rate typically changes across NREM and REM sleep, what nighttime variation can be normal, and how to interpret consumer wearable sleep-stage data without treating it as clinical sleep staging.
A typical healthy overnight pattern looks broadly like this:
| Sleep State | Typical Heart-Rate Pattern | Autonomic Context |
|---|---|---|
| Awake before sleep | Usually higher than during established sleep | More waking activity and environmental response |
| N1 | Begins to slow | Transition from wakefulness into sleep |
| N2 | Usually lower and more stable | Greater parasympathetic influence |
| N3 / deep sleep | Often among the lowest and most stable levels of the night | Strong NREM cardiovascular downshift |
| REM sleep | Often faster and more variable than NREM | Greater autonomic variability and sympathetic bursts |
| Brief awakening | Can rise rapidly | Temporary return toward waking physiology |
The important word is typically.
Individual heart rates differ substantially, and the same person can show different patterns from one night to another.
The transition from wakefulness to sleep changes cardiovascular regulation.
As you relax and enter NREM sleep:
This creates the familiar overnight decline in heart rate seen in many healthy sleepers.
The change is gradual rather than an instant switch at the moment you close your eyes.
Clinical sleep staging divides normal sleep into three NREM stages followed by REM sleep:
N1 and N2 are often grouped under the consumer-friendly term light sleep.
N3 is commonly called deep sleep or slow-wave sleep.
REM is the stage most strongly associated with vivid dreaming, although dreaming can occur outside REM as well.
N1 is the transition between wakefulness and established sleep.
During this stage:
Because N1 is transitional, your heart rate may still resemble waking values more closely than it will later in the night.
Small fluctuations are also common because brief movements or incomplete sleep onset can interrupt the transition.
N2 represents a more established NREM state.
Compared with wakefulness, heart rate is generally:
Breathing and body temperature also continue moving toward the physiological patterns associated with sleep.
Because adults often spend a substantial portion of the night in N2, this stage contributes significantly to the overall sleeping-heart-rate average.
N3 is the deepest stage of NREM sleep.
During deep sleep, cardiovascular regulation generally shifts toward a particularly calm state.
Heart rate often:
Blood pressure also tends to fall.
This makes deep NREM sleep one of the clearest examples of the cardiovascular downshift that occurs during normal sleep.
No.
Deep sleep is associated with low average cardiovascular activity, but your single lowest heart-rate reading does not have to occur during a segment labeled deep sleep.
The timing of your minimum can also be influenced by:
Use sleep stages to provide context around the whole-night curve instead of searching for a perfect one-to-one match.
REM sleep has a distinctly different autonomic pattern.
Compared with stable NREM sleep, REM can include:
These changes are associated with greater autonomic variability during REM.
As a result, a short heart-rate increase during a REM-rich portion of the night can fit normal sleep physiology.
REM combines an unusual set of physiological conditions.
The brain becomes highly active while most skeletal muscles remain strongly inhibited.
Autonomic activity becomes less stable than during NREM.
This can create temporary cardiovascular surges that include:
Heart rate during some REM episodes can move closer to waking levels before falling again when sleep returns to NREM.

Dream-related brain activity may contribute to cardiovascular variability during REM, but wearable data cannot tell you that a specific heart-rate spike was caused by a particular dream.
Other possible contributors include:
Use the heart-rate curve as physiological information rather than attempting to reconstruct dream content from it.
Autonomic regulation differs between the two states.
Parasympathetic cardiac influence generally increases while sympathetic activity decreases.
This contributes to:
Autonomic activity becomes more variable and sympathetic bursts occur more frequently.
This creates a less stable cardiovascular pattern.
| NREM | REM | |
|---|---|---|
| Heart rate | Generally lower | Often higher and more variable |
| Blood pressure | Generally lower | More variable |
| Breathing | Usually more regular | Often more irregular |
| Autonomic pattern | Greater parasympathetic influence | Greater autonomic variability |
You do not move through N1, N2, N3, and REM once and remain there.
Sleep cycles repeat several times.
Your overnight heart-rate curve therefore often contains repeated:
decline → stability → variation → decline
rather than one smooth downward line.
Sleep-stage distribution changes as the night progresses.
The earlier part of a normal night generally contains more deep NREM sleep.
REM episodes tend to become longer and more prominent later in the sleep period.
This means the second half of the night may naturally contain:
This is one reason a nighttime heart-rate curve should be interpreted across the full sleep period.
Sleep stage is only one factor controlling nighttime cardiovascular activity.
Your internal circadian system continues operating while you sleep.
Research has demonstrated circadian rhythms in heart rate and autonomic measures even when sleep stage is considered.
This means:
sleep stage + time of night
both contribute to the heart-rate pattern.
A heart-rate increase near wake time can reflect several overlapping factors:
A gradual rise toward your normal waking level can therefore be part of an ordinary overnight pattern.
This four-step framework helps you interpret sleeping heart rate without overreading individual points.
Was the wearable estimating light sleep, deep sleep, REM, or wake?
Was this early in the night, near the middle, or close to waking?
Consider:
Does the pattern repeat across several nights?
The final step usually carries more practical value than any isolated stage-specific number.

There is no single ideal curve, but one common pattern is:
| Part of Night | Possible Heart-Rate Pattern |
|---|---|
| Falling asleep | Heart rate begins declining |
| Early NREM | Continues settling |
| Deep NREM | Low and relatively stable |
| REM episode | Temporarily more variable or higher |
| Return to NREM | Often settles again |
| Later-night REM | More variability may appear |
| Approaching wake | Gradual rise may occur |
Your personal curve may differ while still fitting healthy sleep physiology.
Many healthy adults have sleeping heart rates below their daytime resting values.
A broad commonly cited range for adults is approximately 40–60 bpm during sleep, but this should be treated as population context rather than a personal target.
Your normal sleeping heart rate depends on:
Highly trained people may normally reach lower values.
For long-term wearable tracking, your own nighttime baseline is usually more useful than forcing every night into a population range.
See how to interpret resting heart rate during sleep for a broader baseline framework.
There is no universal number of beats per minute that separates NREM from REM.
A rule such as:
“REM should always be 10 bpm higher than deep sleep”
would be misleading.
The difference depends on:
Focus on relative patterns instead of fixed stage-specific thresholds.
A sleeping heart-rate chart does not need to be perfectly flat.
Brief increases can occur during:
A short spike followed by a return toward your usual sleeping range is different from a heart rate that remains elevated throughout much of the night.
Imagine two nights:
The second pattern provides much stronger evidence that something changed in the night's recovery context.
Wakefulness causes a rapid shift toward daytime autonomic activity.
If you briefly wake, move, and fall asleep again, heart rate can rise quickly.
A consumer wearable may estimate the surrounding minutes as REM, light sleep, or wake depending on the available physiological signals.
This is one reason a short nighttime spike should not be assigned to a sleep stage with certainty.
This distinction is essential when interpreting wearable heart-rate data.
Clinical sleep staging is performed using polysomnography.
A formal sleep study can include measurements such as:
Sleep specialists use these signals to classify sleep according to standardized clinical criteria.
Consumer wearables cannot reproduce the full signal set of polysomnography.
Depending on the device, algorithms may use combinations of:
The result is an estimated sleep-stage timeline.
That estimate can be useful for observing repeated patterns across nights, but it should not be treated as equivalent to clinical polysomnographic staging.
Movement provides a relatively strong clue about whether someone is awake or sleeping.
Separating N1, N2, N3, and REM requires much more detailed physiological classification.
Validation research shows that consumer wearables can provide useful sleep information while still making stage-classification errors compared with polysomnography.
This is especially important when interpreting one night's exact REM or deep-sleep minutes.
This is an important limitation that is easy to overlook.
Many wearable sleep algorithms use heart-rate or HRV information as part of the process for estimating sleep stages.
That means you may see:
REM label + higher heart rate
partly because the same cardiovascular information contributed to the classification itself.
You should therefore avoid reasoning:
“The wearable says REM, so REM independently proves why heart rate increased.”
Instead, treat the two as related pieces of an algorithmic sleep estimate.

A more useful interpretation is:
“This heart-rate increase occurred during a period the wearable estimated as REM, which is physiologically compatible with greater cardiovascular variability.”
That statement respects both:
For a deeper explanation, see how to read wearable REM sleep trends without overinterpreting them.
Heart rate gives you the speed of the heartbeat.
HRV describes variation in the timing between heartbeats.
Across stable NREM sleep, you may commonly see:
REM produces greater autonomic variability, so both heart rate and HRV can change.
The relationship is complex enough that a single HRV number cannot identify a sleep stage by itself.
See how nighttime HRV fits into sleep and recovery tracking.
Heart rate and HRV are related but distinct signals.
The lowest average heart rate does not automatically produce the highest possible HRV value.
Both are influenced by:
Review each metric according to its own personal baseline.
Heart rate naturally interacts with respiration.
Small beat-to-beat changes occur as you inhale and exhale.
Breathing itself also changes by sleep stage.
Breathing is generally slower and more regular.
Breathing can become more irregular.
This is another reason REM heart rate and HRV may look less stable.
A wearable pattern involving repeated heart-rate increases together with changes in respiratory rate or SpO2 deserves more context than a heart-rate spike alone.
Possible explanations can range from ordinary brief arousals to sleep-disordered breathing or other physiological changes.
A consumer wearable cannot determine the diagnosis from this pattern.
If you are concerned about breathing during sleep, see how to interpret respiratory-rate trends during sleep.
Hard or late exercise can change the entire night's cardiovascular baseline.
Instead of seeing only a brief REM-related rise, you may notice:
This is a workload and recovery effect layered on top of normal sleep-stage variation.
A large meal close to bedtime keeps digestion and metabolic activity active during the early part of sleep.
Some people notice:
If the pattern repeatedly follows late meals, timing becomes useful behavioral context.
Alcohol can produce a recognizable overnight pattern that includes:
In this situation, interpreting each heart-rate rise as simply “REM” would miss the broader behavioral effect.
Psychological stress or temporary illness may increase cardiovascular strain independently of sleep stage.
You may notice:
The whole-night baseline has shifted.
That pattern should be interpreted differently from one temporary REM spike.
| Pattern | More Useful Interpretation |
|---|---|
| Short HR rise during estimated REM | Can fit normal REM cardiovascular variability |
| Short HR rise during a brief awakening | Can reflect transition toward waking physiology |
| Higher HR throughout most NREM and REM | Review broader recovery and lifestyle context |
| Repeated spikes with disrupted sleep | Review breathing, movement, stress, illness, and sleep quality |
| Multi-night increase above personal baseline | Persistent trend deserves more attention |
Consider two healthy people:
| Person A | Person B | |
|---|---|---|
| Typical sleeping HR | 45–52 bpm | 60–68 bpm |
| Deep-sleep pattern | Low 40s | High 50s |
| REM pattern | 50s–60s | 60s–70s |
Both can have internally consistent overnight patterns.
The better comparison is:
your night vs. your recent nights
rather than:
your heart rate vs. someone else's heart rate.
A useful nighttime review asks:
This approach provides more information than one nightly average.

Suppose two nights both have an average sleeping heart rate of 55 bpm.
The average is identical.
The cardiovascular recovery pattern is very different.
Your nightly low point can shift later after:
A single late minimum has limited meaning.
A repeated relationship with a behavior is more useful.
First check whether the difference is small and isolated.
Possible explanations include:
A consumer “deep sleep” label does not guarantee that every minute within the segment would receive the same N3 classification on polysomnography.
REM heart rate does not have to be high during every episode.
REM is characterized by greater cardiovascular variability at the population level.
Individual REM periods can still contain relatively calm heart-rate intervals.
A stage-specific pattern should therefore be evaluated over multiple cycles and nights.
No single heart-rate threshold can reliably identify a sleep stage.
For example:
Heart rate below 50 bpm = deep sleep
is not a valid universal rule.
Someone with a waking resting heart rate of 45 bpm could remain below 50 across several different sleep stages.
Another person may never fall below 50 during the night.
NREM and REM are defined by patterns in brain, eye, muscle, and physiological activity.
That is why clinical sleep studies use multiple simultaneous channels.
Consumer algorithms attempt to approximate those states from a smaller sensor set.
The resulting stage estimate is useful for longitudinal pattern recognition, with appropriate limits.
Wearable stage data can help you notice whether:
This is stronger than using one night's exact stage percentages as a clinical result.
Wearable sleep-stage estimates cannot independently confirm:
Formal evaluation may require clinical history and appropriate sleep testing.
Consider these two nights:
| Night A | Night B | |
|---|---|---|
| Estimated REM | Slightly lower than usual | Slightly lower than usual |
| Sleeping HR | Normal baseline | Elevated all night |
| HRV | Normal baseline | Below baseline |
| Total sleep | Normal | Short |
| Next-day feeling | Normal | Fatigued |
The REM estimate is similar.
The broader recovery picture is very different.
RingConn supports continuous heart-rate and HRV tracking and provides sleep information including estimated sleep stages.
Useful overnight information can include:
The strongest use is to compare repeated nights and identify how your own heart-rate curve changes with sleep, recovery, and daily behavior.
Instead of asking only:
“How much REM did I get?”
review:
This creates a more complete nighttime interpretation.
A seven-night review is usually more informative than one unusual chart.
Look for:
Your personal pattern becomes easier to recognize as the history grows.
RingConn Gen 3 supports continuous monitoring of heart rate and HRV along with sleep-related metrics such as sleep duration and estimated sleep stages.
This allows heart-rate changes to be reviewed as part of a broader overnight pattern rather than as isolated measurements.
Users interested in continuous day-and-night monitoring can explore RingConn Gen 3.
A single REM-related heart-rate rise can fit normal sleep physiology.
Pay closer attention when you repeatedly notice:
A wearable cannot determine the medical cause of these patterns.
Discuss persistent unexplained nighttime heart-rate changes with a healthcare professional when they are new, repeated, or associated with significant symptoms.
Seek urgent medical evaluation for symptoms such as:
Serious symptoms take priority over whether a wearable sleep stage or heart-rate value appears normal.
Heart rate normally changes throughout sleep because cardiovascular and autonomic regulation change across NREM, REM, brief awakenings, and the circadian cycle.
The broad pattern is:
N1: heart rate begins slowing
N2: lower and more stable
N3: often among the lowest and calmest periods
REM: faster and more variable cardiovascular activity
These patterns describe tendencies, not hard stage-specific heart-rate thresholds.
A brief REM-associated heart-rate increase can fit normal sleep physiology. A heart rate that remains elevated throughout much of the night represents a different pattern and should be interpreted with sleep, HRV, exercise, alcohol, stress, illness, and other context.
The most useful framework is:
Sleep Stage → Time of Night → Context → Multi-Night Trend
Consumer wearables estimate sleep stages from physiological and movement signals. They do not reproduce clinical polysomnography, so an estimated REM or deep-sleep label should be used as contextual trend information rather than a definitive clinical classification.
RingConn can support this approach by tracking heart rate, HRV, sleep, SpO2, respiratory rate, and other supported wellness signals across repeated nights.
RingConn products are intended for personal health and wellness awareness and are not medical devices. Heart rate, HRV, sleep-stage estimates, SpO2, respiratory rate, and other RingConn wellness information should not replace clinical sleep staging, professional medical advice, diagnosis, emergency assessment, or treatment.
Heart rate generally becomes lower and more stable during deep NREM sleep as parasympathetic influence increases and cardiovascular activity settles. Deep sleep often contains some of the lowest heart-rate periods of the night, although your exact nightly minimum does not have to occur during every segment labeled deep sleep.
It often becomes faster and more variable during REM compared with stable NREM sleep. Short accelerations can occur because REM involves greater autonomic variability and sympathetic activation.
Brief spikes can occur during REM, movement, brief awakenings, arousals, or changes in breathing. Duration, repetition, accompanying symptoms, and whether heart rate returns toward your normal sleeping range help provide context.
Many healthy adults fall somewhere around 40–60 bpm during sleep, but age, fitness, medication, health status, and individual physiology create substantial variation. Your personal nighttime baseline is often more useful for wearable trend tracking.
It often reaches low levels during deep NREM sleep, but the exact nightly minimum is also influenced by circadian timing, lifestyle, training, and individual physiology. A single lowest heart-rate point cannot identify sleep stage by itself.
Consumer wearables can use combinations of heart rate, HRV, movement, and other supported signals to estimate sleep stages. These estimates can be useful for longitudinal trends but are not equivalent to clinical polysomnography.
Later-night sleep contains more REM, and circadian physiology also begins shifting toward wakefulness. Brief awakenings and movement become more common as well, so a gradual rise toward waking heart rate can occur normally.
Persistent heart rate above your usual nighttime range, repeated large spikes, significant breathing changes, new palpitations, or concerning symptoms deserve further evaluation. Chest pain, fainting, significant breathing difficulty, or severe symptomatic rhythm changes require prompt medical assessment.