Do Supplements Really Increase Deep vs REM Sleep? What the Evidence Shows

Sleep Architecture and Supplements: Can You Target Deep vs REM Sleep?

Sleep is not a single state. Across a typical night, adults rotate through 4–6 cycles of non-REM (N1, N2, N3/deep or slow‑wave sleep) and REM sleep, each cycle lasting roughly 90–110 minutes. N3 tends to dominate the first half of the night and is linked to physical restoration, glymphatic metabolite clearance, and immune support, while REM becomes more abundant toward morning and is associated with memory integration and emotional processing. Many supplements are marketed to “increase deep sleep” or “boost REM,” but how well does research support stage-specific effects?

This focused review examines what clinical studies suggest about popular sleep supplements and their impact on sleep stages and cycles, and contrasts them with cognitive behavioral therapy for insomnia (CBT‑I), the gold-standard non-drug approach.

How Researchers Measure Sleep Stages (and Why It Matters)

• Gold standard: Polysomnography (PSG) uses EEG, EOG, EMG, respiration, and other signals to score N1, N2, N3, and REM with high accuracy.
• Common in trials: Actigraphy and sleep diaries estimate sleep timing and efficiency but do not reliably quantify sleep stages. Claims about “more deep sleep” from actigraphy or consumer wearables should be viewed cautiously.

Key point (evidence level: strong): Only PSG can confirm changes in N3 or REM percentages. Many supplement trials rely on subjective outcomes, so stage-specific claims are often indirect.

Melatonin: Helps You Fall Asleep, Not a Stage Shifter

What it is: A chronobiotic hormone that helps align sleep timing with circadian rhythms.

Evidence: Meta-analyses and systematic reviews report that melatonin can shorten sleep onset latency and modestly increase total sleep time, especially in circadian-related difficulties (e.g., delayed sleep phase), with mixed effects in primary insomnia.

• A 2013 meta-analysis in PLoS One reported shorter sleep latency and slightly longer total sleep time, with no consistent changes in sleep architecture reported across trials (evidence level: strong).

Takeaway: Research suggests melatonin primarily supports sleep initiation and circadian timing. Consistent increases in N3 or REM have not been demonstrated (evidence level for stage changes: limited/strong against specific stage effects).

Magnesium (including glycinate): Calming Mechanisms, Sparse PSG Data

What it is: An essential mineral involved in GABAergic signaling and NMDA receptor modulation; glycinate is a well-tolerated chelate often marketed for “deep sleep.”

Evidence: Systematic reviews suggest small improvements in subjective sleep quality, particularly in older adults or those with low magnesium status. However, trials with polysomnography are scarce, and head-to-head comparisons of forms (glycinate vs others) on sleep stages are lacking.

• Reviews in Nutrients and related journals report modest benefits on overall sleep quality and latency, with highly variable methods and limited objective staging data (evidence level: moderate for sleep quality; emerging for deep sleep claims).

Takeaway: Magnesium may help overall sleep quality in some people, but research has not consistently shown increases in N3 or REM by PSG. Claims that magnesium glycinate “boosts deep sleep” remain preliminary (evidence level for stage effects: emerging).

Glycine: Thermoregulatory Support with Mixed Architecture Findings

What it is: A simple amino acid that may lower core body temperature and promote sleep propensity.

Evidence: Small randomized trials from Japan have reported better subjective sleep quality, less next‑day fatigue, and improved performance after poor sleep. Limited PSG data suggest no large, consistent changes in stage proportions; some small studies note trends toward improved slow‑wave continuity or shorter REM latency, but effects are inconsistent.

• Trials published in Sleep and Biological Rhythms and related journals show subjective benefits with minimal, variable architecture changes (evidence level: moderate for subjective sleep quality; emerging for stage-specific effects).

Takeaway: Glycine may help you feel more restored after sleep loss and may modestly ease sleep initiation, but reliable increases in N3 or REM have not been established (evidence level for stage effects: emerging).

Apigenin (Chamomile Flavone): Traditional Calm, Limited Human PSG

What it is: A bioactive in chamomile and parsley with GABA-A receptor activity in preclinical models.

Evidence: Systematic reviews of chamomile suggest small improvements in sleep quality and anxiety in some populations, but human trials rarely include PSG, and apigenin-specific data are sparse.

• Reviews in Phytotherapy Research and complementary medicine journals report modest subjective benefits with heterogeneous study quality (evidence level: emerging for sleep; lacking for architecture).

Takeaway: Apigenin-containing chamomile may help with relaxation and perceived sleep quality. Effects on N3/REM are not defined in human trials (evidence level for stage effects: emerging).

Tart Cherry Juice: Melatonin-Containing, Duration-Oriented

What it is: Montmorency tart cherries contain phytochemicals and small amounts of melatonin.

Evidence: Small randomized trials in older adults with insomnia and in athletes report longer total sleep time and improved sleep efficiency by actigraphy and diaries; PSG-based staging is rarely reported. The mechanism may involve melatonin and tryptophan metabolism.

• Trials in Journal of Medicinal Food and European Journal of Nutrition show modest increases in sleep duration, but architecture changes are uncertain (evidence level: moderate for total sleep time; limited for stage effects).

Takeaway: Tart cherry may extend sleep duration in some groups, but evidence does not show consistent shifts in N3 or REM (evidence level for stage effects: limited).

Traditional Botanicals: Valerian, Passionflower, and Jujube Seed

Valerian (Valeriana officinalis)

• Evidence: Meta-analyses are mixed; some show small improvements in overall sleep quality, with variable methodology. PSG evidence for N3/REM changes is inconsistent or absent.
• Evidence level: moderate for subjective quality; limited for architecture.

Passionflower (Passiflora incarnata)

• Evidence: Small RCTs report improved subjective sleep quality and reduced anxiety; objective staging data are limited.
• Evidence level: emerging for sleep quality and anxiety; lacking for architecture.

Jujube seed (Ziziphus spinosa, Suan Zao Ren)

• Traditional use: A core herb in East Asian formulas for calming the spirit and reducing nighttime awakenings.
• Evidence: Systematic reviews of Suan Zao Ren–containing formulas report improvements in validated insomnia scales versus controls; PSG reporting is rare, and herb combinations vary.
• Evidence level: traditional/emerging for sleep quality; lacking for architecture.

Bridging perspectives: Traditional systems emphasize calming the mind and consolidating sleep rather than targeting specific EEG stages. Western trials similarly find that many botanicals may improve perceived sleep continuity without demonstrating reliable increases in N3 or REM on PSG.

How This Compares to CBT‑I

CBT‑I is the first‑line behavioral therapy for chronic insomnia and includes sleep restriction, stimulus control, circadian regularity, and cognitive strategies.

• Meta-analyses (e.g., Annals of Internal Medicine) show robust, durable improvements in insomnia severity, shorter sleep latency, reduced wake after sleep onset, and better sleep efficiency (evidence level: strong).
• Architecture: While CBT‑I primarily consolidates sleep opportunity rather than directly increasing N3 or REM, some studies suggest improved slow‑wave continuity as sleep becomes more consolidated; consistent stage percentage shifts are not its main mechanism (evidence level: moderate for indirect architecture normalization).

What to Watch For in Supplement Studies Aiming at “Deep Sleep”

• Objective measurement: Prefer trials with PSG if stage changes are claimed (evidence level: strong rationale).
• Population: Effects may differ in older adults, those with circadian delay, anxiety, or nutrient insufficiency (evidence level: moderate).
• Outcomes: Improvements in sleep latency, efficiency, or next‑day alertness are meaningful even without stage shifts (evidence level: strong for clinical relevance).
• Interactions: Botanicals and nutrients may interact with medications or conditions. Discuss changes with a clinician, especially if pregnant, nursing, or managing chronic illness.

Bottom Line

• Melatonin: Research suggests it helps you fall asleep and align circadian timing but does not reliably change N3 or REM proportions (evidence level: strong for onset; limited for architecture).
• Magnesium (including glycinate): May modestly improve overall sleep quality in some groups; evidence that it specifically boosts deep sleep is preliminary and inconsistent (evidence level: moderate for quality; emerging for N3 effects).
• Glycine: May improve subjective sleep quality and next‑day functioning after poor sleep; consistent stage changes have not been shown (evidence level: moderate for subjective benefits; emerging for architecture).
• Apigenin/chamomile: May help relaxation and perceived sleep quality; human PSG data are limited (evidence level: emerging).
• Tart cherry: May increase total sleep time in select groups; architecture effects remain unclear (evidence level: moderate for duration; limited for stages).
• Valerian, passionflower, jujube seed: Traditional calming remedies with some supportive trials for sleep quality; little evidence for altering N3/REM on PSG (evidence level: moderate to emerging, traditional use strong).
• Compared with supplements, CBT‑I has the strongest and most durable evidence for improving insomnia and consolidating sleep, with any architecture normalization likely secondary to better sleep continuity.

Overall, current science does not strongly support the idea that over‑the‑counter supplements can reliably “increase deep sleep” or “boost REM” on polysomnography. Several options may help you fall asleep faster, sleep longer, or feel more restored—valuable outcomes even without stage-specific changes. Pairing any supplement approach with behavioral strategies aligned with CBT‑I principles may offer the most robust, sustainable gains.