Deep sleep duration is the total minutes spent in slow-wave sleep (stage N3) during a night, the stage most strongly tied to physical restoration, memory consolidation, and growth-hormone release. Healthy adults typically land between 60 and 110 minutes per night, concentrated in the first third of sleep. Because it front-loads, deep sleep is the stage most vulnerable to late alcohol, late workouts, warm bedrooms, and caffeine that lingered past its half-life.
Why it matters
If an evening stack is meant to improve recovery, deep sleep minutes are the clearest place to look for the signal. It is also the sleep-architecture metric most cited on the Huberman and Attia podcasts and, correspondingly, the one users most want to move. In practice, environment and timing shift it more than any single compound, which is why evaluating a new supplement for deep sleep without first fixing sleep hygiene usually produces a false negative.
Sleep stage reference
| Stage | Share of a healthy 7–8 hour night | What it does |
|---|---|---|
| N1 (light) | 2–5% | Transition from wake to sleep |
| N2 (light) | 45–55% | Sleep spindles, motor-learning consolidation |
| N3 (deep) | 13–23% (roughly 60–110 min) | Physical restoration, immune function, GH release |
| REM | 20–25% | Emotional processing, memory consolidation |
Use these as orientation, not targets. A night with 70 minutes of N3 and 95 minutes of REM at 88% efficiency is objectively a strong night for most adults.
Interpreting the number
- > 90 min — strong for an adult.
- 60–89 min — typical, protect it.
- 30–59 min — worth reviewing alcohol, late workouts, room temperature, and bedtime dose window content.
- < 30 min — a real signal; review stack overlap and environment before layering any new compound.
What moves the number
Deep sleep is suppressed by alcohol (even one drink within 3 hours), THC after several weeks of regular use, late caffeine, bedroom temperatures above 20 °C, eating within 2 hours of bed, and under-recovery. It is supported by cool rooms (18–19 °C / 65–67 °F), consistent bedtime, morning light exposure, and zone-2 cardio earlier in the day. Supplementation plays a smaller role than environment, and claims that magnesium glycinate, glycine, or apigenin reliably increase PSG-confirmed N3 minutes should be treated as unproven unless direct sleep-stage evidence is present.
Uncertainty
Wearable sleep staging is noisier than polysomnography, and device-to-device variance can run 15–25 minutes on the same night. Treat the number as a within-device trend rather than a portable figure.
- Do not compare raw minutes across Oura, Whoop, Apple Watch, Garmin, or Eight Sleep.
- Track the delta from a rolling 28-day moving average window rather than a single night.
- Single-night swings of 10–20 minutes are normal; look for trend shifts of 15% or more across 2–3 weeks.
- Growth-hormone claims attached to specific compounds are almost always overstated at consumer doses. The more defensible target is "protect the first sleep cycle," where most N3 occurs.
How this appears in Unfair
Deep sleep duration is a default objective proxy alongside sleep efficiency whenever a stack has a recovery, longevity, or training-quality goal. Unfair ingests stage data through wearable sync, normalizes against the user's own 28-day baseline establishment, and flags trend changes of more than 20% versus the prior window. Stage charts sit next to the nightly stack on the review screen so evening compounds and environment tags (late workout, travel, drink) are visible against the N3 line — one of the advanced capabilities that makes the signal usable.
Clinical safety note
Persistently very low deep sleep (under 20 minutes per night) despite corrected sleep hygiene and timing is a sleep-medicine question, not a supplement one. A formal sleep study is the right next step, particularly when paired with loud snoring, daytime fatigue, or witnessed apneas.