The Science of Sleep: What Your Smartwatch Data Is Actually Telling You

Every morning your COLMI smartwatch or smart ring gives you a sleep score. Maybe it's a 74, maybe it's an 88. Light sleep: 3 hours 12 minutes. Deep sleep: 1 hour 5 minutes. REM: 1 hour 40 minutes. These numbers appear in the Da Fit app with impressive precision — but what do they actually mean? Are they accurate? And most importantly, what should you do with them?

This guide goes deep into the science behind sleep, explains what each stage does for your body and mind, covers what your wearable can and can't accurately measure, and — most usefully — gives you practical ways to use your sleep data to genuinely improve your health and daily performance.

Why Sleep Is the Foundation of Health

Before the specifics, it's worth establishing why sleep deserves serious attention. Sleep is not a passive state of rest — it's an intensely active process during which your brain and body perform tasks that are impossible during wakefulness:

• Cellular repair and protein synthesis: Growth hormone is primarily secreted during deep sleep, driving tissue repair, muscle rebuilding after exercise, and cellular maintenance throughout the body
• Immune system reinforcement: Cytokines (immune system signalling proteins) are produced during sleep. Chronic sleep deprivation measurably suppresses immune function and increases susceptibility to infection
• Memory consolidation: During REM sleep, the brain transfers information from short-term to long-term memory and integrates new learning with existing knowledge. Sleep deprivation impairs learning retention dramatically
• Metabolic regulation: Sleep deprivation disrupts the hormones ghrelin (hunger) and leptin (satiety), increasing appetite, reducing the ability to feel full, and promoting calorie intake of 200 to 500 calories above normal
• Emotional processing: REM sleep processes emotional experiences and stress, which is why sleep-deprived people are more emotionally reactive, anxious, and prone to poor decision-making
• Cardiovascular maintenance: Blood pressure drops during sleep, giving the cardiovascular system essential recovery time. Chronic poor sleep is independently associated with elevated cardiovascular disease risk
• Brain waste clearance: The glymphatic system — the brain's waste clearance mechanism — is most active during deep sleep, clearing metabolic byproducts including amyloid beta (associated with Alzheimer's disease) from brain tissue

Sleep is not optional recovery time. It is the most important thing your body does every day, and the data your COLMI device provides about your sleep is potentially the most health-relevant information in the entire Da Fit app.

The Sleep Architecture: What Actually Happens During a Night

A full night of sleep consists of several complete cycles, each lasting approximately 90 minutes. Each cycle passes through different sleep stages in a characteristic progression:

Stage N1: Light Sleep Onset (Transition)

N1 is the very brief transition from wakefulness into sleep, typically lasting only 1 to 7 minutes. You're easy to wake at this stage and may experience hypnic jerks — the sudden muscle twitches that sometimes wake you just as you're falling asleep. N1 is not counted as meaningful sleep in most wearable sleep summaries; it's generally included in light sleep totals or not counted separately.

Stage N2: True Light Sleep

N2 is the most abundant stage in a typical night, making up approximately 40 to 60% of total sleep time. Your heart rate slows, body temperature drops slightly, and the brain produces characteristic sleep spindles (bursts of neural activity thought to protect sleep from disturbance) and K-complexes (large neural responses to external stimuli). N2 is the stage of sleep your wearable is most likely to label simply as "light sleep."

N2 sleep is not "junk" sleep — it plays important roles in memory consolidation, motor learning, and physical recovery, and sleep spindle density during N2 is associated with intelligence and learning ability. However, insufficient N3 (deep sleep) at the expense of excess N2 indicates sleep that isn't as restorative as it could be.

Stage N3: Deep Sleep (Slow-Wave Sleep)

N3 — also called slow-wave sleep, delta sleep, or deep sleep — is the most physically restorative stage. The brain produces large, synchronised delta waves, heart rate and breathing reach their lowest points of the night, and the body's repair processes are most active. Growth hormone secretion is maximised, tissue repair is prioritised, and immune function is enhanced.

Deep sleep is disproportionately concentrated in the first half of the night. If you go to bed at 11pm and wake at 7am, most of your deep sleep occurs between 11pm and 2am. This is why poor early sleep quality (disrupted by alcohol, late meals, or stress) is so costly even if you sleep through the morning hours — the most restorative sleep has already been lost.

Typical healthy adults spend 15 to 25% of total sleep time in N3, equivalent to 60 to 120 minutes for a 7 to 8 hour sleep period. Deep sleep decreases with age — older adults typically spend 5 to 15% of sleep in N3, which is one reason recovery and immune function become more challenging with advancing age.

Stage REM: Rapid Eye Movement Sleep

REM sleep is the dreaming stage. The brain is highly active — in some respects more active than during wakefulness — the eyes move rapidly beneath closed lids, and most voluntary muscles are temporarily paralysed by the brainstem (which prevents you from acting out dreams). Heart rate and breathing become more variable than in NREM sleep.

REM sleep is critical for emotional processing, creativity, memory integration, and mood regulation. Insufficient REM is strongly associated with mood disorders, poor emotional regulation, impaired creativity, and learning difficulties. Alcohol is a particularly potent REM suppressor — even moderate drinking before bed can cut REM sleep by 20 to 25%, which is why drinking alcohol in the evening often produces a night of poor-quality sleep despite falling asleep easily.

REM periods lengthen throughout the night, with most REM sleep concentrated in the early morning hours. The final 2 hours before waking (say, 5am to 7am for a 7am wake time) are disproportionately rich in REM sleep. Cutting sleep short by even an hour in the morning can significantly reduce your total REM sleep for the night.

Healthy adults typically spend 20 to 25% of total sleep time in REM, equivalent to 80 to 120 minutes per night for a 7 to 8 hour sleep period.

What Your Wearable Can and Can't Measure

Sleep stage detection without EEG (brain wave measurement) is fundamentally an estimation problem. Consumer wearables use a combination of movement data and heart rate patterns as proxies for brain activity. Here's what that means in practice:

What Your COLMI Device Measures

• Movement via accelerometer: Deep sleep produces minimal body movement; light sleep and wakefulness involve more. The accelerometer provides the primary input for sleep stage classification.
• Heart rate patterns: Heart rate follows predictable patterns across sleep stages (lowest and most regular in deep sleep, slightly higher and more variable in REM, intermediate in light sleep). Your COLMI's optical sensor provides this input.
• Heart rate variability (HRV): HRV also varies across sleep stages and adds a second physiological dimension to stage classification beyond simple heart rate level.

What It Cannot Measure

• EEG brain waves — the gold standard for sleep stage classification. Only a clinical sleep laboratory (polysomnography, PSG) can measure these directly.
• Eye movement (used in clinical settings to confirm REM sleep)
• EMG (muscle tone) used to confirm REM sleep paralysis in clinical settings

Accuracy of Consumer Sleep Tracking

Research comparing consumer wearables to clinical PSG finds:

• Total sleep time: Accurate within 15 to 20 minutes for most users on most nights
• Sleep onset detection: Generally accurate within 5 to 15 minutes
• Wake detection: Reasonably accurate for wakes lasting more than 2 to 3 minutes; brief arousals are commonly missed
• Light sleep (N2) identification: Moderately accurate; most time in light sleep is correctly classified
• Deep sleep (N3) identification: Accuracy is lower; consumer devices tend to overestimate deep sleep in some users and underestimate it in others. Deep sleep percentage should be treated as a relative trend indicator rather than an absolute measurement.
• REM sleep identification: The most challenging stage to detect without EEG. Consumer accuracy for REM is the most variable, with studies showing agreement rates of 65 to 80% depending on the device and individual.

The key insight is this: your sleep score and stage data are most valuable as trend indicators over time, not as precise nightly measurements. If your deep sleep has been declining for two weeks, that's meaningful regardless of whether the absolute percentages are precisely accurate. If your REM proportion drops significantly after nights when you drink alcohol, that's a reliable and actionable signal. Individual nightly variations should be taken with more scepticism than week-over-week trends.

How to Interpret Your COLMI Sleep Score

The sleep score in Da Fit (typically 0 to 100) combines several factors:

• Total sleep duration (compared to age-appropriate recommendations)
• Sleep efficiency (proportion of time in bed actually spent asleep)
• Deep sleep duration relative to total sleep
• REM sleep duration relative to total sleep
• Number and duration of wake periods during the night
• Consistency with your previous sleep pattern

Rough interpretation guide:

• 85 to 100: Excellent sleep. Well-rested and likely to perform well the following day.
• 70 to 84: Good sleep. Adequate recovery for most people.
• 55 to 69: Fair sleep. You may notice some fatigue or reduced performance the next day.
• Below 55: Poor sleep. Recovery is compromised and performance, mood, and health are likely affected.

Don't obsess over individual nights. A single poor night is normal and happens to everyone. A consistent pattern of scores below 70 warrants attention and investigation into the underlying causes.

Practical Actions Based on What You See in the Data

If Your Deep Sleep Is Consistently Low (Less Than 45 Minutes)

• Eliminate alcohol consumption on weeknights — alcohol suppresses deep sleep even in small amounts
• Lower your bedroom temperature to 16 to 19°C. Body temperature must drop for deep sleep to initiate; a cool bedroom facilitates this
• Begin regular aerobic exercise. Cardio training is the most evidence-based intervention for increasing deep sleep duration and quality
• Establish a consistent bedtime within a 30-minute window. Irregular sleep timing reduces deep sleep proportion

If Your REM Sleep Is Consistently Low (Less Than 70 Minutes)

• Stop drinking alcohol, especially in the evening. This is the single most impactful intervention for REM suppression
• Avoid cutting sleep short in the morning. The last 1 to 2 hours before your natural wake time are the richest in REM. Setting an alarm earlier than needed routinely sacrifices REM sleep
• Address anxiety or psychological stress if present. Stress hormones suppress REM sleep and excessive REM suppression is associated with mood disorders
• Avoid cannabis use before bed. Cannabis significantly suppresses REM sleep

If You Wake Frequently During the Night

• Limit fluid intake in the 2 hours before bed to reduce nocturia (waking to use the toilet)
• Investigate sleep apnoea if you wake repeatedly without obvious cause, particularly if your COLMI device's SpO2 data shows overnight dips below 90%. Discuss with your GP.
• Maintain a consistent sleep schedule. Irregular timing promotes lighter, more fragmented sleep
• Review caffeine timing. Caffeine's half-life is approximately 5 hours; a 4pm coffee has 25% of its effect still active at 2am

If Your Sleep Duration Is Consistently Under 7 Hours

• Prioritise sleep as a health behaviour equivalent to exercise and nutrition. Most UK adults are chronically sleep-deprived by choice.
• Move your bedtime 15 minutes earlier each week until reaching your target duration, rather than making a sudden large change that's harder to maintain
• Review evening screen time and blue light exposure in the 2 hours before bed, which suppresses melatonin and delays sleep onset

Why Smart Rings Outperform Smartwatches for Sleep Data

COLMI smart rings (R02 through R12) have a structural advantage over smartwatches for sleep tracking that goes beyond sensor position. The ring form factor is dramatically more comfortable to sleep in than a wrist watch — no display, no bulk, no pressure on the underside of the wrist when lying with arms in various sleeping positions. Because rings are more comfortable, wearers are significantly more likely to actually wear them overnight consistently.

Consistent overnight wearing is the single most important factor in generating useful sleep data. A smartwatch that's worn 60% of nights due to discomfort produces an incomplete, often unrepresentative picture of your sleep. A ring that's worn 95% of nights produces a comprehensive, reliable baseline from which meaningful trends emerge quickly.

The COLMI R07 (30-day battery) eliminates the risk of the ring running out of charge during a sleep cycle entirely. The R09 (lightest concave design) maximises overnight comfort for users who are particularly sensitive to ring presence. For sleep tracking specifically, these two models are COLMI's strongest recommendations.

The Long Game: Using Your Data Over Months

The richest sleep insights from your COLMI device emerge not from individual nights but from months of accumulated data. With 30, 60, and 90 days of consistent sleep data in Da Fit, you can observe:

• Whether your sleep quality is trending better or worse over time
• How fitness improvements (starting an exercise program, for example) have affected your deep sleep
• Seasonal patterns in sleep duration and quality
• The impact of lifestyle changes you make deliberately (reducing alcohol, improving sleep timing, installing blackout curtains)
• Whether your sleep has been consistently insufficient long-term (revealing chronic sleep debt that explains persistent fatigue)

This longitudinal perspective is the most powerful and most underutilised aspect of wearable sleep tracking. Wear your COLMI every night, review your data weekly, and over months you'll build a picture of your sleep health that no doctor's appointment or one-night sleep study can provide.