What is your sleep data saying?
What is your sleep data saying?
Here's. a simple guide to analyze the metrics your smartwatch throws at you each morning-
and how you can make it actionable. (8-minute read)
Here's. a simple guide to analyze the metrics your smartwatch throws at you each morning-
and how you can make it actionable. (8-minute read)
|
Reviewed by Prof. Ritika Jain, IIT Madras Zanzibar
|
Co-Founder, Remster
Co-Founder, Remster
Reviewed by Prof. Ritika Jain, IIT Madras Zanzibar
[1]. Sleep trackers statistics - The Quantified Scientist- Summary spreadsheet
[2]. Dijk D. J. (2009). Regulation and functional correlates of slow wave sleep. Journal of clinical sleep medicine : JCSM: American Academy of Sleep Medicine, 5(2), S6-S15.
[3]. Walker, M. (2018). Why We Sleep. Penguin Books.
[4]. Ohayon, M. M., et al. (2004). Meta-analysis of quantitative sleep parameters from childhood to old age in healthy individuals: developing normative sleep values across the human lifespan. Sleep, 27(7), 1255-1273.
[5]. Kondo M. et al. (2007). Physiological significance of cyclic changes in room temperature around dusk and dawn for circadian rhythms of core and skin temperature. Journal of Physiological Anthropology, 26(4), 429-436.
[6]. Cajochen, C. et al. (2024). Ultradian sleep cycles: Frequency, duration, and associations with individual and environmental factors—A retrospective study. Sleep Health, 10(1), S52-S62.
[7]. Bonnet, M. H., & Arand, D. L. (1997). Heart rate variability: sleep stage, time of night, and arousal influences. Electroencephalography and clinical neurophysiology, 102(5),
390-396.
[1]. Sleep trackers statistics - The Quantified Scientist- Summary spreadsheet
[2]. Dijk D. J. (2009). Regulation and functional correlates of slow wave sleep. Journal of clinical sleep medicine : JCSM: American Academy of Sleep Medicine, 5(2), S6-S15.
[3]. Walker, M. (2018). Why We Sleep. Penguin Books.
[4]. Ohayon, M. M., et al. (2004). Meta-analysis of quantitative sleep parameters from childhood to old age in healthy individuals: developing normative sleep values across the human lifespan. Sleep, 27(7), 1255-1273.
[5]. Kondo M. et al. (2007). Physiological significance of cyclic changes in room temperature around dusk and dawn for circadian rhythms of core and skin temperature. Journal of Physiological Anthropology, 26(4), 429-436.
[6]. Cajochen, C. et al. (2024). Ultradian sleep cycles: Frequency, duration, and associations with individual and environmental factors—A retrospective study. Sleep Health, 10(1), S52-S62.
[7]. Bonnet, M. H., & Arand, D. L. (1997). Heart rate variability: sleep stage, time of night, and arousal influences. Electroencephalography and clinical neurophysiology, 102(5),
390-396.
References
References
Sleep data should feel actionable, not academic.
By learning to read the nightly rhythm hidden inside your sleep ‘timeline’—i.e. spotting the heart-rate spikes that signal each 90-minute cycle—you unlock some practical wins:
Wake up when your body is at its lightest sleep stage. Set alarms to coincide with the natural cycle end and reduce grogginess, and get happy mornings.
Target improvements, not perfection. If REM dips below ≈25% or deep sleep under ≈20%, look at the low-hanging habit fixes: consistent bedtimes, cooler bedrooms, and alcohol-free evenings. Simple changes to unlock a better life.
Subjective feeling beyond the numbers: You can take home the patterns from the data, and you need not worry about the score every time. Listening to your body can be above all else.
Sleep data should feel actionable, not academic.
By learning to read the nightly rhythm hidden inside your sleep ‘timeline’—i.e. spotting the heart-rate spikes that signal each 90-minute cycle—you unlock some practical wins:
Wake up when your body is at its lightest sleep stage. Set alarms to coincide with the natural cycle end and reduce grogginess, and get happy mornings.
Target improvements, not perfection. If REM dips below ≈25% or deep sleep under ≈20%, look at the low-hanging habit fixes: consistent bedtimes, cooler bedrooms, and alcohol-free evenings. Simple changes to unlock a better life.
Subjective feeling beyond the numbers: You can take home the patterns from the data, and you need not worry about the score every time. Listening to your body can be above all else.
The first cycle begins with sleep onset and contains a fair amount of Light & Deep sleep, ending on a brief REM period, summing up to 96 min in total.
The second cycle is only 70 minutes long. However, I’ve slipped into a longer Deep sleep period. Notice how this is the last deep sleep period for the night- a natural pattern we have evolved towards. Most deep sleep occurs in the first part of the night, and most REM occurs in the latter.
The third and fourth cycles are similar, with duration of 88 and 120 minutes, containing mostly Light and REM stages.
The fifth cycle here shows an incomplete Ultradian cycle, having woken up 21 minutes into the cycle, by 7:47 am. It’s a clear sign- showing potential for alarm time optimisation in my life. This also contributes to reduced %REM, as during this morning hour, REM sleep dominates.
But, how do you identify the cycle start/finish?
At the end of each Ultradian sleep cycle, there is a marked spike in HR. This increase in heart rate is, in fact, a nervous system response to sleep stage transitions [7]. A brief awakening is visible at this point, where you’re closest to natural awakening. This marks the transition from one cycle to another, sometimes with awake periods.
The first cycle begins with sleep onset and contains a fair amount of Light & Deep sleep, ending on a brief REM period, summing up to 96 min in total.
The second cycle is only 70 minutes long. However, I’ve slipped into a longer Deep sleep period. Notice how this is the last deep sleep period for the night- a natural pattern we have evolved towards. Most deep sleep occurs in the first part of the night, and most REM occurs in the latter.
The third and fourth cycles are similar, with duration of 88 and 120 minutes, containing mostly Light and REM stages.
The fifth cycle here shows an incomplete Ultradian cycle, having woken up 21 minutes into the cycle, by 7:47 am. It’s a clear sign- showing potential for alarm time optimisation in my life. This also contributes to reduced %REM, as during this morning hour, REM sleep dominates.
But, how do you identify the cycle start/finish?
At the end of each Ultradian sleep cycle, there is a marked spike in HR. This increase in heart rate is, in fact, a nervous system response to sleep stage transitions [7]. A brief awakening is visible at this point, where you’re closest to natural awakening. This marks the transition from one cycle to another, sometimes with awake periods.
Why does it matter?
If you learn about your Ultradian pattern, you can now set a morning alarm such that you wake up at the end of a cycle rather than in between. This makes all the difference between starting your day groggy and waking up fresh and alert. Estimate wake-up time in multiples of 90-minute blocks. However, this is hard to perfect manually and can take a few attempts, which is precisely why we’re building the ‘RiseWise’ feature at Remster. It automatically detects the end of a sleep cycle closest to your alarm time, and wakes you up gently with warmth, helping you wake up fresh every day.
Embrace consistent sleep and wake timing to keep the pattern more predictable, as our bodies love regularity.
It also means if you wish to go back to bed in the morning, you’ll need a ~90-minute sleep window before waking up again.
How do you spot these?
Open up your sleep timeline along with the Heart Rate (HR) plot. Here’s an example of my sleep timeline on a typical night, captured using a Polar Verity Sense device with the Sleep² app. I’ve split them with green dashed lines to identify the 5 cycles easily.
Why does it matter?
If you learn about your Ultradian pattern, you can now set a morning alarm such that you wake up at the end of a cycle rather than in between. This makes all the difference between starting your day groggy and waking up fresh and alert. Estimate wake-up time in multiples of 90-minute blocks. However, this is hard to perfect manually and can take a few attempts, which is precisely why we’re building the ‘RiseWise’ feature at Remster. It automatically detects the end of a sleep cycle closest to your alarm time, and wakes you up gently with warmth, helping you wake up fresh every day.
Embrace consistent sleep and wake timing to keep the pattern more predictable, as our bodies love regularity.
It also means if you wish to go back to bed in the morning, you’ll need a ~90-minute sleep window before waking up again.
How do you spot these?
Open up your sleep timeline along with the Heart Rate (HR) plot. Here’s an example of my sleep timeline on a typical night, captured using a Polar Verity Sense device with the Sleep² app. I’ve split them with green dashed lines to identify the 5 cycles easily.
Your sleep occurs in natural cycles called ‘Ultradian Rhythms’, each lasting about 90 minutes. These cycles consist of different sleep stages: Light, Deep/REM, and brief awake moments, before starting again.
The term ‘Ultradian’ refers to a repeating pattern in the order of a few hours, and in this case, a predictable 90-min cycle driven by human physiology, due to which it does vary from person to person [6]. A healthy person experiences about 4-5 such cycles each night, summing up to 7-8 hours of sleep.
Your sleep occurs in natural cycles called ‘Ultradian Rhythms’, each lasting about 90 minutes. These cycles consist of different sleep stages: Light, Deep/REM, and brief awake moments, before starting again.
The term ‘Ultradian’ refers to a repeating pattern in the order of a few hours, and in this case, a predictable 90-min cycle driven by human physiology, due to which it does vary from person to person [6]. A healthy person experiences about 4-5 such cycles each night, summing up to 7-8 hours of sleep.
To improve % Deep sleep, on the other hand, adding physical activity to your day and taking time to wind down are proven techniques. As seen above in my report from July 4th, I did put in a workout, which, despite a stressful day, contributed to a decent overall score of 72%.
Apart from this, a cool bedroom environment in the early parts of the night, to assist in reducing your core body temperature, is seen to be crucial for higher durations of deep, restful sleep [5]. This is the core philosophy of what we do at Remster: we’re building a smart bed cover with active temperature control. It intelligently reduces temperature 30 minutes before bedtime and actively helps you slip into deep sleep.
To improve % Deep sleep, on the other hand, adding physical activity to your day and taking time to wind down are proven techniques. As seen above in my report from July 4th, I did put in a workout, which, despite a stressful day, contributed to a decent overall score of 72%.
Apart from this, a cool bedroom environment in the early parts of the night, to assist in reducing your core body temperature, is seen to be crucial for higher durations of deep, restful sleep [5]. This is the core philosophy of what we do at Remster: we’re building a smart bed cover with active temperature control. It intelligently reduces temperature 30 minutes before bedtime and actively helps you slip into deep sleep.
What’s interesting is that the logical regions of the brain are inactive during REM sleep- the prefrontal cortex, which usually sends instructions and rational thoughts to other brain regions, is temporarily at rest. Motor functions are also blocked, i.e. muscles frozen solid, which is evolution’s way of preventing us from enacting our dreams in bed.
So, what’s better for us, REM or Deep sleep?
The answer is simple: one cannot replace the other. While deep sleep is crucial for recovery, REM has a great role in storing memory in permanent storage and in enhancing problem-solving capability. This is why being told to ‘sleep on it’ is actually good advice when making a big decision.
Today, your data shows you the percentage of sleep duration spent in each stage. A common metric, straightforward, but what’s the right amount? Many research studies show that 25% REM sleep and 20% Deep sleep are healthy amounts, varying with age, sex, genetics, etc. [4]
If that’s not the case in your sleep report, a common reason for low REM% % could include alcohol consumption- a known suppressant of REM periods. In this case, however, the report shows inconsistent bedtime, and simply the total duration of sleep (<7 hr), are areas to focus on for improving % REM duration.
What’s interesting is that the logical regions of the brain are inactive during REM sleep- the prefrontal cortex, which usually sends instructions and rational thoughts to other brain regions, is temporarily at rest. Motor functions are also blocked, i.e. muscles frozen solid, which is evolution’s way of preventing us from enacting our dreams in bed.
So, what’s better for us, REM or Deep sleep?
The answer is simple: one cannot replace the other. While deep sleep is crucial for recovery, REM has a great role in storing memory in permanent storage and in enhancing problem-solving capability. This is why being told to ‘sleep on it’ is actually good advice when making a big decision.
Today, your data shows you the percentage of sleep duration spent in each stage. A common metric, straightforward, but what’s the right amount? Many research studies show that 25% REM sleep and 20% Deep sleep are healthy amounts, varying with age, sex, genetics, etc. [4]
If that’s not the case in your sleep report, a common reason for low REM% % could include alcohol consumption- a known suppressant of REM periods. In this case, however, the report shows inconsistent bedtime, and simply the total duration of sleep (<7 hr), are areas to focus on for improving % REM duration.
Deep sleep is the “restorative” part of your sleep duration. This is where our bodies heal, immunity improves, and muscle repair occurs [2]. It is also called as “Slow Wave Sleep” (SWS), as your heartbeat slows down, breathing comes to a calm, relaxed pace, and awakening becomes difficult at this stage.
Whereas, the REM stage is very different. It involves Rapid Eye Movements (thus the name), and brain activity is seen to ‘light up’ on MRI scans. Researchers have even found 30% more activity in emotional regions of the brain than while we are awake [3]. The conscious experience of dreaming occurs here, leading to the rich psychotic experience that we recall upon waking.
Deep sleep is the “restorative” part of your sleep duration. This is where our bodies heal, immunity improves, and muscle repair occurs [2]. It is also called as “Slow Wave Sleep” (SWS), as your heartbeat slows down, breathing comes to a calm, relaxed pace, and awakening becomes difficult at this stage.
Whereas, the REM stage is very different. It involves Rapid Eye Movements (thus the name), and brain activity is seen to ‘light up’ on MRI scans. Researchers have even found 30% more activity in emotional regions of the brain than while we are awake [3]. The conscious experience of dreaming occurs here, leading to the rich psychotic experience that we recall upon waking.
When I first bought my Garmin FR 255, in fascination with the tech, I tried to figure out what was going on. I used to admire how it detected exactly when I’d woken up the previous night. But soon, once the excitement faded, it wasn’t all that useful practically. How can this be made practical? Read on.
When I first bought my Garmin FR 255, in fascination with the tech, I tried to figure out what was going on. I used to admire how it detected exactly when I’d woken up the previous night. But soon, once the excitement faded, it wasn’t all that useful practically. How can this be made practical? Read on.
Notice how I say “predict” —> That’s because these devices use algorithms that translate the vital measurements into estimates of sleep stages, rather than directly “detect” your sleep stage. Clinically, sleep stages are actually determined by monitoring brain activity in lab setups using intrusive probes, with a technique called Polysomnography (PSG). So, the predictions made by your watch will typically be 80-85% accurate to PSG results [1], the gold standard in the industry.
Now, when you wake up to see a timeline like this, what do you notice?
Notice how I say “predict” —> That’s because these devices use algorithms that translate the vital measurements into estimates of sleep stages, rather than directly “detect” your sleep stage. Clinically, sleep stages are actually determined by monitoring brain activity in lab setups using intrusive probes, with a technique called Polysomnography (PSG). So, the predictions made by your watch will typically be 80-85% accurate to PSG results [1], the gold standard in the industry.
Now, when you wake up to see a timeline like this, what do you notice?
The overall score provided is called the sleep score, usually scored out of 100, which tries to encapsulate all the data into a single parameter. Many times we feel very well rested, but somehow, the score will be < 80. It can be due to various reasons, so while this score itself doesn’t have to mean so much, the long-term timeline can still be useful.
The overall score provided is called the sleep score, usually scored out of 100, which tries to encapsulate all the data into a single parameter. Many times we feel very well rested, but somehow, the score will be < 80. It can be due to various reasons, so while this score itself doesn’t have to mean so much, the long-term timeline can still be useful.
Smartwatches (and rings) today measure a few of your vital signs, i.e. how fast your heart is beating (Heart Rate), how many breaths you’re taking each minute (Respiration Rate), etc. They do so at regular intervals to predict the stage of sleep you’re in.
Smartwatches (and rings) today measure a few of your vital signs, i.e. how fast your heart is beating (Heart Rate), how many breaths you’re taking each minute (Respiration Rate), etc. They do so at regular intervals to predict the stage of sleep you’re in.
The sleep timeline is the most dominant plot on the dashboard- it talks about when you fell asleep, what stages of sleep you were at, and when you woke up. Simply put, it’s a visual representation of your brain activity & body function over the course of the night.
The sleep timeline is the most dominant plot on the dashboard- it talks about when you fell asleep, what stages of sleep you were at, and when you woke up. Simply put, it’s a visual representation of your brain activity & body function over the course of the night.
I get it. Sleep tracking is boring- you have to wear your smartwatch to bed, and the report in the morning is a few graphs and metrics that aren’t exactly easy to interpret.
Let’s fix that.
I get it. Sleep tracking is boring- you have to wear your smartwatch to bed, and the report in the morning is a few graphs and metrics that aren’t exactly easy to interpret.
Let’s fix that.
REM-Deep Balance:
About 90-minute sleep cycles
About 90-minute sleep cycles
