Sleep Metrics Explained: Duration, Efficiency, Latency, WASO, HRV, SpO2 and More

Sleep trackers have made previously clinical sleep metrics accessible to millions of people. Whether you use an Oura Ring, Apple Watch, Fitbit, Garmin, Whoop, or another wearable, you're probably generating data you don't fully understand — and possibly worrying about numbers that don't mean what you think they mean.

This guide covers each major sleep metric: what it measures, what's considered healthy, how wearables measure it, and how much you should care about it.

1. Sleep Duration

Definition: Total time spent asleep, typically measured from sleep onset to final waking. Distinct from "time in bed" (TIB).

Healthy range: 7–9 hours for most adults (see our full guide to sleep duration recommendations). Below 7 hours per night consistently is associated with significantly worse health outcomes in large prospective studies. Above 9 hours in healthy adults may be a sign of underlying illness.

How wearables measure it: Most devices detect sleep onset using a combination of reduced movement (accelerometer data) and physiological signals like reduced heart rate. They're reasonably accurate at detecting total sleep time — typically within 30–45 minutes of polysomnography-measured values in most studies.

What to watch: Focus on your weekly average rather than individual nights. One short night is not cause for alarm; chronic short sleep is.

2. Sleep Efficiency

Definition: The percentage of time in bed that is actually spent asleep. Calculated as: (Total Sleep Time / Time in Bed) × 100.

Healthy range: 85% or higher. Clinical sleep medicine defines insomnia as partially characterized by sleep efficiency consistently below 85%. Elite sleepers often achieve 90–95%.

Example: If you're in bed for 8 hours and sleep for 7 hours, your sleep efficiency is 87.5% — healthy.

Why it matters: Low sleep efficiency means you're spending significant time in bed awake — which can be a sign of insomnia, anxiety, sleep apnea, or simply spending too much time in bed (which paradoxically worsens insomnia). A core component of Cognitive Behavioral Therapy for Insomnia (CBT-I) is sleep restriction — temporarily reducing time in bed to drive efficiency up, then gradually extending it.

How wearables measure it: Reasonably well, though they may misclassify quiet wakefulness as sleep, inflating efficiency estimates.

3. Sleep Latency

Definition: The time it takes to fall asleep after getting into bed and intending to sleep. Measured from "lights out" to sleep onset.

Healthy range: 10–20 minutes. Under 8 minutes is considered by the AASM as a sign of significant sleepiness (and potentially sleep debt or a sleep disorder). Over 30 minutes regularly may indicate insomnia or anxiety-related arousal at bedtime.

Why it matters: Sleep latency is a sensitive marker of sleepiness/sleep debt (short latency = very sleepy) and of hyperarousal at bedtime (long latency = too alert/anxious to sleep). Very short sleep latency is not a sign of being a "great sleeper" — it's more often a sign of being chronically under-slept. A healthy person who is getting adequate sleep should take some time to fall asleep — they're not desperate for it.

How wearables measure it: Moderately well. Devices may over-estimate latency (classifying early N1 as wakefulness) or underestimate it (classifying still wakefulness as sleep).

4. WASO (Wake After Sleep Onset)

Definition: Total time spent awake during the night after initially falling asleep. Includes all nighttime awakenings, from normal brief arousals between cycles to longer periods of wakefulness.

Healthy range: Under 30 minutes total WASO per night is generally considered healthy for adults under 60. Brief arousals (30 seconds to 2 minutes) between cycles are normal and expected — you have 10–15 of them per night. Longer awakenings that accumulate to 30+ minutes total suggest disrupted sleep.

WASO increases with age: Older adults commonly have higher WASO due to lighter, more fragmented sleep architecture — values up to 45–60 minutes may be less concerning in adults over 65.

Common causes of high WASO:

• Sleep apnea (arousals triggered by breathing events)
• Anxiety or stress (hyperarousal preventing return to sleep after normal arousals)
• Alcohol (increases wakefulness in the second half of the night as it metabolizes)
• Caffeine (reduces sleep continuity, particularly later in the night)
• Pain, discomfort, or medical conditions
• Bladder urgency (nocturia) — itself often a sign of sleep apnea

How wearables measure it: Variable accuracy. They may underestimate WASO by classifying brief awakenings as sleep, or overestimate it by flagging periods of quiet stillness in a light stage as wakefulness.

5. Sleep Debt

Definition: The cumulative deficit between the sleep you need and the sleep you've actually obtained. Not directly measured by any wearable — it must be calculated from known sleep need and tracked sleep duration.

Healthy range: Ideally zero — consistently meeting your biological sleep need. Most people in modern societies carry some chronic sleep debt.

How wearables approach it: Some devices (Whoop, Oura) offer "sleep debt" estimates based on comparing your actual sleep to their estimate of your need. These estimates are imprecise but directionally useful.

How to use it: Focus on cumulative trends. A week of significantly short sleep requires a week of slightly longer sleep (or several nights of extended sleep) to meaningfully reduce debt. Use our Sleep Debt Tracker for a personalized estimate.

6. Resting Heart Rate (RHR) During Sleep

Definition: Your heart rate during sleep — typically lowest in the middle of the night (during N3 deep sleep) and rising slightly toward the end of the night as cortisol increases in anticipation of waking.

Healthy range: Highly individual. Athletic individuals may have RHR of 40–50 bpm during sleep; sedentary individuals may be in the 60s. What matters more than the absolute number is your personal baseline and deviations from it.

What deviations indicate:

• Elevated RHR during sleep — may indicate illness (often precedes symptom onset by 1–2 days), alcohol consumption (which raises RHR in the second half of the night), high stress or cortisol, or overtraining in athletes
• Declining RHR trend over weeks — typically indicates improving cardiovascular fitness
• Rising RHR trend — may indicate accumulating physiological stress from training, illness, or sleep deprivation

How wearables measure it: Modern photoplethysmography (PPG) sensors in wearables are reasonably accurate for heart rate — typically within a few beats per minute of medical-grade measurements during sleep.

7. Heart Rate Variability (HRV)

Definition: The variation in time between consecutive heartbeats (measured in milliseconds). Not the same as heart rate — HRV measures how consistent or variable the intervals between beats are. High HRV indicates the autonomic nervous system is flexible and responsive; low HRV indicates a stressed, rigid system.

Healthy range: Highly individual and age-dependent. Young adults may average 50–100ms (RMSSD). Middle-aged adults typically range 30–60ms. Values below 20ms in adults consistently suggest elevated autonomic stress. More important than absolute values is your personal baseline — a deviation of 20%+ from your own baseline is more meaningful than any absolute number.

Why it matters for sleep: HRV during sleep reflects sleep quality and physiological recovery. High HRV during sleep indicates the parasympathetic nervous system is dominant (appropriate for recovery). Low HRV during sleep suggests sympathetic activation — typically meaning the body is under stress. Alcohol reliably reduces sleep HRV. Good sleep consistently improves next-day HRV.

How wearables measure it: There is significant variation in HRV measurement methodology between devices, making cross-device comparisons meaningless. Even within the same device, HRV measurements can vary significantly night to night due to sensor placement, movement, and algorithm differences. Track trends and personal baseline, not absolutes.

8. Blood Oxygen Saturation (SpO2)

Definition: The percentage of hemoglobin in the blood that is carrying oxygen, measured by pulse oximetry. Normally abbreviated SpO2 ("peripheral capillary oxygen saturation").

Healthy range: 95–100% during sleep. Values consistently below 90% are clinically significant and may indicate sleep-disordered breathing (sleep apnea), COPD, or other respiratory conditions. Values below 88% are considered potentially dangerous and warrant medical evaluation.

Why it matters for sleep: The primary clinical application is screening for sleep apnea. Obstructive sleep apnea causes repeated drops in blood oxygen (desaturations) as breathing stops and restarts throughout the night. Even moderate sleep apnea (15–30 events per hour) can produce many desaturations per night.

How wearables measure it: Consumer SpO2 sensors are less accurate than clinical pulse oximeters — typically ±2–3% in ideal conditions, more in motion or with poor skin contact. They're useful for identifying patterns of significant desaturation but should not be used for clinical diagnosis. Persistent readings below 94% during sleep should prompt a clinical evaluation.

Important note: Skin tone affects the accuracy of optical SpO2 sensors. Multiple studies have found that consumer and even some clinical pulse oximeters overestimate SpO2 in individuals with darker skin tones, potentially masking real hypoxia. This is an ongoing issue in device design that users with darker skin should be aware of.

9. Respiratory Rate

Definition: The number of breaths per minute during sleep. Measured by some advanced wearables using accelerometer data or optical sensors detecting chest movement.

Healthy range: 12–20 breaths per minute (12–16 during deep sleep). Elevated respiratory rate during sleep can indicate illness (fever, infection), respiratory conditions, or anxiety. Some wearables use respiratory rate as a component of their readiness/recovery scores.

How wearables measure it: Accuracy varies significantly between devices. Oura and some Garmin devices derive it from PPG signal variations; others use accelerometry. Less reliable than heart rate or SpO2 measurements.

Consumer Tracker Accuracy: What the Research Says

Multiple independent studies comparing consumer sleep trackers to polysomnography (PSG) — the clinical gold standard — have found:

• Total sleep time: Reasonable accuracy (typically within 30–45 minutes of PSG); wearables tend to overestimate total sleep time
• Sleep vs. wake classification: Good performance (~85–90% agreement with PSG for broad sleep/wake classification)
• REM sleep: Moderate — devices perform better on REM than NREM staging; typical accuracy 60–75%
• N3 deep sleep: Most variable — highly inconsistent between devices and individuals; should not be relied on for clinical decision-making
• N1 and N2 distinction: Generally poor — most devices don't distinguish reliably between light sleep stages

The bottom line: consumer trackers are useful for tracking trends in your own sleep over weeks and months. They're not accurate enough for individual-night clinical decision-making, and sleep stage percentages (especially deep sleep and REM) should be treated as rough approximations rather than precise measurements.

The Risk of Orthosomnia

"Orthosomnia" — a term coined by researchers at Rush University Medical Center — describes the paradoxical situation where someone becomes so preoccupied with optimizing their sleep tracker metrics that the anxiety this creates actually worsens their sleep. Signs of orthosomnia include:

• Changing behavior based on individual night readings rather than trends
• Feeling anxious before bed about what the tracker will show
• Feeling fine until checking the tracker, then feeling tired because it showed low sleep
• Disrupting your day based on a "poor sleep score"

If your sleep tracker is causing anxiety or changing your behavior in ways that feel compulsive, consider taking a break from checking it daily. Use the data weekly or monthly for trend analysis rather than night-by-night optimization.

Frequently Asked Questions