There are two ways to notice that deep sleep is not where it should be.
The first is through a device: you open Garmin or Apple Watch in the morning and see Deep sleep: 28 min. A small number, no clear context for what it means, but enough to make you wonder.
The second way requires no device at all. You sleep a full 7 to 8 hours, but wake up heavy-headed, eyes struggling to open, needing half an hour before you feel properly awake. Getting out of bed feels like it has not actually rested you. Some people describe it as "sleeping without sleeping": the body was there all night, but the brain never recharged. The tiredness is not from too few hours. It is from too little depth.
Both signals are worth paying attention to. A device gives you a number with its own margin of error. Your body gives you a sensation with its own validity. When both point in the same direction, the signal is stronger.
This article focuses on one specific question: why does deep sleep (NREM N3) get compressed, what factors are taking away N3 time each night, and what role do micronutrients play in recovering it.
If you are not yet familiar with the structure of sleep and what each phase does, Sleep Phases and Micronutrients is a useful starting point. If you are unsure which type of poor sleep you are experiencing, The 3 Types of Poor Sleep will help you identify that first.
What wearables measure, and what your body tells you
If you use a wearable, it helps to understand where the number comes from.
Apple Watch and Garmin track sleep primarily through two inputs: wrist movement (accelerometry) and heart rate. An algorithm then classifies sleep states. The clinical gold standard is polysomnography (PSG), which measures brain electrical activity (EEG) directly in a sleep lab.
A 2025 study in PMC comparing six wearable devices against PSG found that accuracy for four-stage classification (wake, light sleep, deep sleep, REM) ranged between 50 and 65%, and devices like Garmin tended to underestimate actual N3 time (PMC12038347). A separate 2024 study confirmed mean absolute errors for Garmin's deep sleep estimate of up to 64.94 minutes (PMC10820351).
This does not mean the data is useless. It does mean you should not draw conclusions from a single night's number. Trends over weeks, comparing your good nights against your worse ones, carry far more meaning than any one absolute figure.
If you do not use a wearable, your morning experience is a rough but real signal. Waking after 7 to 8 hours and still feeling like you cannot quite start the day, with a heavy body and a slow brain, suggests slow-wave sleep may be insufficient. That feeling is harder to quantify than minutes on a screen, but it should not be dismissed simply because there is no number attached.
What N3 actually does, and why the deficit matters
N3, also called slow-wave sleep (SWS), is the deepest stage of NREM sleep, defined by high-amplitude delta waves on the EEG. In a healthy adult sleeping 8 hours, N3 typically accounts for 80 to 120 minutes per night, concentrated in the first half of sleep.
This is not passive rest. Three significant biological processes run specifically during N3:
First: the glymphatic system clears the brain. The glymphatic network is a channel system surrounding the brain's blood vessels, most active during deep sleep. It pumps cerebrospinal fluid through brain tissue, flushing metabolic waste accumulated during the day, including beta-amyloid and tau, two proteins associated with Alzheimer's disease. A 2025 study in Nature confirmed that norepinephrine fluctuations in the brain during NREM sleep directly drive this clearance cycle (PMID 39788123). Less N3 means an incomplete clean-up.
Second: growth hormone is released in a concentrated pulse. Around 70 to 80% of the day's GH output is delivered during the first slow-wave episode, shortly after sleep onset. GH drives tissue repair, muscle recovery, fat metabolism, and cellular renewal (PMID 8627466). Athletes noticing poor recovery, or older adults seeing gradual loss of muscle mass, are often experiencing the consequence of declining N3 over time.
Third: the immune system runs targeted maintenance. During N3, the body synthesises cytokines, signalling molecules that coordinate inflammatory responses and infection defence. This is why people who sleep poorly get sick faster and recover more slowly.
When N3 is compressed, all three processes are cut short. The next day begins with a brain that has not been fully cleared, muscles that have not fully recovered, and an immune system running at reduced capacity.
How much is enough?
Reference ranges by age, based on PSG data from large population studies:
| Age | Average N3 per night (8 hours of sleep) |
|---|
| 18–30 | 80–120 minutes (roughly 17–20% of total sleep) |
| 30–50 | 60–90 minutes |
| 50–65 | 30–60 minutes |
| Over 65 | 10–40 minutes |
The age-related decline is physiologically normal and cannot be fully reversed. But if you are 30 years old and your device consistently reports under 30 minutes across multiple nights, that is a signal worth investigating.
Why N3 gets compressed: causes in order of impact
1. Sleep timing and total duration
N3 is most concentrated in the first 3 to 4 hours after falling asleep. Going to bed significantly later than usual pushes that window toward early morning, when the circadian biology is already moving away from slow-wave production. Sleeping fewer than 6 hours cuts more N3 than any other single factor, simply because there is not enough time to complete the cycles.
2. Alcohol
This is the most underestimated factor. Alcohol induces drowsiness and speeds up sleep onset, but it disrupts sleep architecture in the hours that follow. A 2024 meta-analysis of 27 studies (PMID 39631226) confirmed that even low doses significantly fragment REM, and higher doses reduce N3 in the second half of the night. This explains why a night of drinking often leads to waking earlier than intended and feeling unrested despite enough hours in bed.
3. Chronic stress and cortisol
Cortisol directly opposes deep sleep. When the HPA (hypothalamic-pituitary-adrenal) axis is chronically overactive from prolonged stress, evening cortisol does not fall low enough for the brain to shift into delta-wave activity. The article The 3 Types of Poor Sleep covers this mechanism in more detail, particularly in relation to mid-night waking and early-morning awakening.
4. Screens and blue light before bed
Blue light from phones and screens suppresses endogenous melatonin, delaying the circadian signal that initiates sleep onset. A 2024 study in Brain Communications (PMC11154150) found that using a phone without a blue-light filter before bed measurably reduced N3 in the first quarter of the night. That first quarter is the most important window for slow-wave sleep.
5. Room temperature
To enter N3, core body temperature needs to drop by roughly 1 to 2°C. A warm room slows this process. The clinically recommended bedroom temperature for deep sleep is 18 to 20°C. This is not a comfort preference. It is a physiological requirement.
6. Late intense exercise
High-intensity training raises core body temperature and activates the sympathetic nervous system, both of which work against N3 entry. A gap of at least 3 hours before bed is recommended. That said, regular exercise earlier in the day, morning or early afternoon, has a documented positive effect on N3 quality over time.
Micronutrients and deep sleep: evidence by nutrient
This section does not repeat the general mechanisms already covered in Sleep Phases and Micronutrients. The focus here is on the specific evidence for each micronutrient's effect on N3 and slow-wave sleep.
Magnesium: the strongest evidence for N3
A 2012 RCT (PMC3703169) in 46 older adults with primary insomnia found that 500 mg of magnesium daily for 8 weeks not only improved overall sleep quality but increased slow-wave sleep time from 10.1 to 16.5 minutes and raised delta power, the EEG signature of N3. A 2025 review (PMC12535714) confirmed the mechanism: magnesium inhibits NMDA receptors and activates GABA-A receptors, two direct points of action in the brain's delta-wave generation process.
The forms with the best absorption for sleep: bisglycinate and L-threonate. Magnesium oxide absorbs poorly and acts primarily in the gut rather than the nervous system.
Food sources high in magnesium: Pumpkin seeds (~150 mg per 30g), almonds (~80 mg per 30g), cooked spinach (~78 mg per 100g), black beans (~60 mg per 100g), mackerel and salmon (~35 to 50 mg per 100g), dark chocolate above 70% (~50 mg per 30g). The adult daily requirement ranges from 310 to 420 mg depending on sex and age.
If your diet already includes a consistent variety of leafy greens, nuts, seeds, and fatty fish, magnesium needs are likely being met through food. If those groups are missing regularly, or if you have tracked your intake and found magnesium consistently below your requirement, bisglycinate or L-threonate supplementation is a practical option with direct clinical evidence.
Glycine: the amino acid that works through body temperature
Glycine is a non-essential amino acid with a mechanism that is specific to N3. At the suprachiasmatic nucleus (the brain's circadian clock), glycine activates NMDA receptors in a way that promotes peripheral vasodilation, helping core body temperature drop faster at the start of the night. A lower core temperature is a necessary condition for entering N3 (PMC4397399).
In a small clinical trial (PMID 22293292), 3g of glycine taken before bed increased NREM proportion, shortened N3 latency, and improved subjective ratings of morning freshness. Glycine is an example of a micronutrient acting through the thermal regulation pathway rather than directly on neurotransmitters.
Food sources high in glycine: Glycine is concentrated in connective tissue and collagen. Chicken skin and pork skin (braised or slow-cooked) are among the richest sources. Bone broth provides roughly 1 to 2g of glycine per 250ml. Red meat, especially beef and lamb, contains approximately 1.5 to 2g per 100g. Fish and soy foods contain glycine at lower levels.
Precise dose control from whole foods is difficult. If your diet is low in animal protein or rarely includes collagen-rich cuts, glycine powder (3g before bed) is a practical approach with direct evidence for N3 support.
Zinc: co-regulator of the circadian clock
Zinc is involved in regulating the expression of clock genes (CLOCK, BMAL1), the molecular framework controlling the 24-hour circadian rhythm. When the circadian rhythm drifts, the early-night N3 window shrinks or shifts. A 2024 systematic review (PMC11456512) confirmed improved sleep quality with zinc supplementation, particularly in groups with documented deficiency.
Food sources high in zinc: Oysters are by far the richest source (~74 mg per 100g). Beef and lamb provide 4 to 7 mg per 100g. Pumpkin seeds contain roughly 2 to 4 mg per 30g. Lentils, mung beans, and cashews are moderate sources. The adult daily requirement is 8 to 11 mg.
An important note: zinc from plant foods is bound to phytates and is significantly less bioavailable than zinc from meat and shellfish. People eating vegetarian or vegan diets have a higher risk of zinc insufficiency even when total food intake appears adequate. In those cases, supplementing with zinc bisglycinate (better absorbed than zinc oxide) is a practical way to meet actual requirements, including the sleep-related ones.
A practical framework for prioritising
The evidence is clear: behavioural interventions come first. No micronutrient compensates for nights of alcohol, an overheated room, or two hours of screen time right before bed.
A practical self-assessment framework:
Step 1: Check the external factors first. In the last two weeks, has there been alcohol on any evenings? Regular late bedtimes past midnight? A bedroom consistently above 22°C? Is a phone the last thing you look at before closing your eyes?
Step 2: If the environment is in order, look inward. Has work stress been sustained for months? Is there a period of chronic pressure that is not resolving? These connect to the HPA axis and cortisol, and they are not problems that micronutrients fix.
Step 3: Assess the diet honestly. Do daily meals include regular servings of magnesium-rich foods, zinc-rich animal proteins, and collagen-containing cuts? If you are not sure, tracking food intake with an app for one to two weeks is the most reliable way to find out what you are actually eating. Many people are surprised to find their magnesium intake is around half the daily requirement, despite a subjective sense of eating reasonably well.
Step 4: Consider supplementation when diet falls short or is inconsistent. If the environment is managed, stress is not a dominant factor, but the diet consistently fails to reach target amounts of magnesium or zinc due to schedule, food preferences, or higher physiological demand, supplementation is a reasonable and evidence-supported step. Prioritise high-bioavailability forms (bisglycinate for both magnesium and zinc). Blood testing before long-term supplementation is useful if you want to confirm a true deficiency.
The bottom line
Deep sleep (N3) is the most restorative phase of the night: it clears the brain via the glymphatic system, triggers the main growth hormone pulse, and runs critical immune maintenance. In healthy younger adults, 80 to 120 minutes per night is the target range. The signal that N3 is insufficient may come from a wearable device, or simply from the feeling of waking unrefreshed after a full night.
The factors that most reliably compress N3, in order of impact: late or short sleep, alcohol, chronic stress, screen light before bed, and a warm bedroom. These need to be addressed first. No micronutrient substitutes for them.
Magnesium, glycine, and zinc have evidence specifically supporting N3. The starting point is food: seeds, dark leafy greens, meat, shellfish, and collagen-rich cuts provide all three. When diet is not consistently sufficient, supplementation is a practical option with a clear rationale.
Wearables give useful directional data with meaningful error margins. Use them to track your own trends, not to hit an absolute target number.
Content is educational and not a substitute for medical advice. Consult a healthcare professional before changing your supplement regimen.
References
- Hablitz LM, Nedergaard M. The Glymphatic System: A Beginner's Guide. Neurochem Res. 2021. PMC11835678
- Yamamoto K et al. Norepinephrine-mediated slow vasomotion drives glymphatic clearance during sleep. Cell. 2025. PMID 39788123
- Van Cauter E et al. Physiology of growth hormone secretion during sleep. J Pediatr. 1996. PMID 8627466
- Sletten TL et al. A performance validation of six commercial wrist-worn wearable sleep-tracking devices for sleep stage scoring compared to polysomnography. Sleep. 2025. PMC12038347
- Liang Z et al. Evaluating Accuracy in Five Commercial Sleep-Tracking Devices Compared to Research-Grade Actigraphy and Polysomnography. Nat Sci Sleep. 2024. PMC10820351
- Gardiner C et al. The effect of alcohol on subsequent sleep in healthy adults: A systematic review and meta-analysis. Sleep Med Rev. 2024. PMID 39631226
- Harding EC et al. Altered sleep architecture following consecutive nights of presleep alcohol. J Sleep Res. 2024. PMID 38205895
- Hysing M et al. Effects of evening smartphone use on sleep and declarative memory consolidation in male adolescents and young adults. Brain Commun. 2024. PMC11154150
- Abbasi B et al. The effect of magnesium supplementation on primary insomnia in elderly: A double-blind placebo-controlled clinical trial. J Res Med Sci. 2012. PMC3703169
- Peuhkuri K et al. The Mechanisms of Magnesium in Sleep Disorders. Nat Sci Sleep. 2025. PMC12535714
- Kawai N et al. The sleep-promoting and hypothermic effects of glycine are mediated by NMDA receptors in the suprachiasmatic nucleus. Neuropsychopharmacology. 2015. PMC4397399
- Bannai M, Kawai N. New therapeutic strategy for amino acid medicine: glycine improves the quality of sleep. J Pharmacol Sci. 2012. PMID 22293292
- da Silva J et al. Effects of zinc supplementation on sleep quality in humans: A systematic review of RCTs. Biol Trace Elem Res. 2024. PMC11456512
- Dijk DJ, Czeisler CA. Regulation and Functional Correlates of Slow Wave Sleep. J Clin Sleep Med. PMC2824213