Sleep Optimization: Evidence-Based Supplement Stack Protocol

Key Definitions

Sleep architecture refers to the cyclical structure of sleep, organized into 90–110-minute cycles across the night, each containing multiple stages: N1 (light transition), N2 (sleep spindles, K-complexes — the bulk of sleep), N3 (slow-wave/deep sleep, critical for physical restoration and memory consolidation), and REM (rapid eye movement, associated with emotional processing and procedural memory).

Sleep onset latency (SOL) — the time elapsed from “lights out” to verified sleep onset. Normal: <20 minutes. Chronic SOL >30 min indicates sleep onset insomnia.

Sleep efficiency (SE) — percentage of time in bed actually asleep. Target: ≥85%. Calculated as (Total Sleep Time / Time in Bed) × 100.

NREM slow-wave sleep (SWS / N3) — also called deep sleep or delta sleep. Characterized by high-amplitude, low-frequency delta waves (<2 Hz). Dominant in the first half of the night. Critical for growth hormone release, immune function, and declarative memory consolidation. Most supplements that “improve sleep quality” act primarily here.

REM sleep — occurs predominantly in the second half of the night, cycling every ~90 min with increasing duration. Critical for emotional regulation, creativity, and procedural memory. Generally less targetable by supplements; protecting sleep duration is the primary REM intervention.

Circadian rhythm — the ~24-hour endogenous biological clock governed by the suprachiasmatic nucleus (SCN) in the hypothalamus. Entrained primarily by light (the zeitgeber). Melatonin signals the onset of biological night; cortisol signals morning waking (the Cortisol Awakening Response, CAR).

Cortisol Awakening Response (CAR) — a 50–160% spike in cortisol occurring in the first 20–30 minutes after waking. Healthy CAR correlates with preparedness, executive function, and immune activity. Blunted CAR correlates with chronic stress, burnout, and poor sleep quality.

Adenosine — a sleep-pressure molecule that accumulates during wakefulness (sleep homeostat). Caffeine works by blocking adenosine receptors. Adequate adenosine accumulation requires avoiding caffeine in the 8–10 hours before bed.

GABA (gamma-aminobutyric acid) — the brain’s primary inhibitory neurotransmitter. Most sedative sleep pharmaceuticals (benzodiazepines, Z-drugs) work by potentiating GABA-A receptors. Several natural compounds (magnesium, L-theanine, apigenin, glycine) modulate GABAergic pathways through distinct mechanisms.

Chronotype — genetically-influenced preference for sleep timing (morning vs. evening). Supplements can optimize sleep quality within one’s chronotype but cannot meaningfully shift it in healthy adults. For chronotype shifts, timed light exposure and strategic melatonin use are more appropriate tools.

Key Findings

The strongest signals in the evidence base for sleep supplementation come from several converging lines of research:

1. Glycine 3 g before bed produces measurable polysomnographic improvements in sleep efficiency and reductions in sleep latency in volunteers with sleep complaints — the Yamadera et al. (2007) crossover RCT is the foundational reference. The mechanism (core body temperature reduction via NMDA receptors in the SCN) was confirmed in subsequent animal mechanistic work (Kawai et al. 2009; Bannai & Kawai 2012). This compound is probably the most underrated and underutilized sleep supplement relative to its evidence base.

2. Magnesium supplementation consistently improves subjective and objective sleep in populations with deficiency (common in Western diets). The Abbasi et al. (2012) RCT in elderly insomniacs (PMID: 23853635, PMC3703169) showed significant improvements in sleep efficiency, sleep time, early morning awakening, and ISI score. The form matters: glycinate and threonate have superior bioavailability and tolerability compared to oxide.

3. Low-dose melatonin (0.3–0.5 mg) is as effective as pharmacological doses (3–10 mg) for circadian reset, without the next-morning grogginess caused by supratherapeutic melatonin concentrations. Zhdanova et al. (1996, PMID: 8843534) and (2001, PMID: 11600532) provide the key dose-response data. Most commercial products contain 5–10 mg — a dose 10–20× higher than needed for circadian effects.

4. L-Theanine (100–200 mg) consistently promotes relaxed wakefulness and improved sleep quality via alpha-wave induction and GABA/glutamate modulation. Lyon et al. (2011, PMID: 22214254) demonstrated improved objective sleep quality in an RCT, particularly relevant for ADHD populations with sleep-onset difficulties.

5. Chamomile extract (containing apigenin) shows moderate evidence across meta-analyses for sleep quality and insomnia reduction. However, the specific contribution of isolated apigenin (popularized by Andrew Huberman as 50 mg) is mechanistically plausible but not directly proven in human RCTs using purified apigenin alone.

Sleep Architecture — Why It Matters

Understanding sleep architecture is essential for setting realistic expectations about what supplements can and cannot do.

The Sleep Cycle

A typical night contains 4–6 complete sleep cycles of ~90–110 minutes each:

Cycle 1:  N1 → N2 → N3 (long deep sleep) → brief REM
Cycle 2:  N1 → N2 → N3 (moderate) → longer REM
Cycle 3+: N2 → shorter N3 → increasingly long REM periods

Deep sleep (N3) is heavily front-loaded — most occurs in the first 3–4 hours. REM is back-loaded — most occurs in hours 5–8. This means:

• Sleeping 6 hours instead of 8 hours disproportionately cuts REM, not deep sleep
• Going to bed 1 hour late reduces N3 less than going to bed 2 hours late (which starts cutting into the deep sleep window)
• Alcohol suppresses REM and disrupts sleep architecture even when it appears to “help” falling asleep
• Cannabis similarly suppresses REM chronically

What Supplements Target

Chronotype and Supplementation

Supplements do not change your chronotype. An evening chronotype person who takes melatonin at 10 PM will not become a morning person. However, strategic use of melatonin (0.3 mg, 5–6 hours before desired sleep time) combined with bright light exposure in the morning can gradually shift circadian timing by 30–60 minutes over several weeks — but this is a separate intervention from sleep quality optimization.

Key Compounds — Evidence Review

Magnesium Glycinate

Evidence level: Strong (human RCT + mechanistic)

Magnesium is the fourth most abundant mineral in the body and a cofactor in over 300 enzymatic reactions. Approximately 45–68% of the US population is estimated to consume below the Estimated Average Requirement for magnesium, with similar trends in European populations.

Key study: Abbasi et al. (2012) — double-blind, placebo-controlled RCT in 46 elderly subjects with primary insomnia. Intervention: 500 mg magnesium (as oxide) daily for 8 weeks. Results: statistically significant improvements in ISI score, sleep efficiency, sleep time, sleep onset latency, and early morning awakening. Serum magnesium, renin, melatonin, and cortisol levels all improved favorably. PMID: 23853635; PMC: PMC3703169.

Mechanisms:

• GABA-A receptor co-agonist (potentiates inhibitory neurotransmission)
• NMDA receptor antagonist (reduces excitatory glutamate activity, calming effect)
• Suppression of the HPA axis → lower cortisol → easier sleep initiation
• Regulation of the circadian clock via magnesium-dependent enzyme activity

Why glycinate specifically: Magnesium oxide (used in the Abbasi trial) has poor bioavailability (~4%). Magnesium glycinate binds the mineral to the amino acid glycine, improving absorption to ~25–40% and eliminating the laxative effect seen with oxide and citrate at sleep-relevant doses.

Dosing: 300–400 mg elemental magnesium as glycinate, taken 60–90 minutes before bed. Note: product labels may show total weight (e.g., 400 mg magnesium glycinate) which provides only ~50–70 mg elemental magnesium — read labels carefully and look for the elemental amount.

Caveats: Most deficient individuals will notice improved sleep quality within 1–2 weeks. Those with adequate magnesium status may see smaller benefits. Magnesium is consistently safe at recommended doses.

Magnesium L-Threonate

Evidence level: Moderate (animal RCT + limited human data)

Magnesium L-threonate (MgT) is a patented form developed at MIT specifically for its blood-brain barrier (BBB) crossing ability. While standard magnesium forms raise serum magnesium, they do not reliably increase cerebrospinal fluid (CSF) magnesium concentrations. MgT does.

Key study: Slutsky et al. (2010) — foundational animal study in rats demonstrating that dietary MgT elevates brain magnesium by ~15% compared to other forms and produces significant enhancements in synaptic plasticity, working memory, and long-term memory. Published in Neuron. PMID: 20152124.

Human data is more limited. A preliminary trial (Sun et al., 2016, small N) suggested improvements in sleep quality scores and cognitive performance in older adults with cognitive decline, but larger placebo-controlled human trials are still lacking as of 2026.

Mechanisms:

• Crosses BBB → elevates CSF magnesium → NMDA receptor modulation in hippocampus and cortex
• Enhanced synaptic density and NMDAR-mediated LTP
• Relevant sleep mechanism: reduces nocturnal hyperarousal and anxiety-mediated sleep disruption through central magnesium action

Who benefits most: Individuals with anxiety-driven insomnia or cognitive concerns alongside sleep issues. The cognitive + sleep combination makes MgT more appropriate than glycinate in this subset.

Dosing: Typical commercial MgT products (e.g., Magtein): 2 g total weight providing ~144 mg elemental magnesium. Take 60–90 minutes before bed. Can be used instead of or in combination with magnesium glycinate (different mechanisms, different bioavailability profiles).

Honest caveat: MgT is significantly more expensive than glycinate, and the human sleep-specific evidence is weaker than for magnesium glycinate. The cognitive benefits from animal data are compelling, but translating rodent neuroscience to humans requires caution.

L-Theanine

Evidence level: Moderate-Strong (multiple RCTs)

L-Theanine is an amino acid found almost exclusively in tea (Camellia sinensis), typically present at 6–50 mg per cup of green tea. It is the primary reason green tea, despite containing caffeine, produces a qualitatively different alertness state compared to coffee.

Key study: Lyon et al. (2011) — randomized, double-blind, placebo-controlled crossover trial in 98 boys aged 8–12 with ADHD. Intervention: 400 mg Suntheanine® daily (200 mg twice daily) for 6 weeks. Results: significant improvements in sleep efficiency ratio and sleep maintenance as measured by actigraphy. PMID: 22214254; published in Alternative Medicine Review.

Additional supporting evidence: Multiple smaller RCTs and mechanistic studies confirm alpha-wave induction (8–13 Hz EEG activity, associated with relaxed wakefulness) beginning within 30–40 minutes of ingestion. Ozeki et al. demonstrated L-theanine’s ability to counter caffeine-induced arousal while preserving attention.

Mechanisms:

• Increases alpha-wave activity in frontal cortex
• Modulates GABA and glutamate neurotransmission (partial GABA agonism, reduced excitatory glutamatergic signaling)
• Structurally similar to glutamate — crosses BBB and acts as a glutamate receptor partial agonist/antagonist
• Does NOT directly cause sedation — promotes calm wakefulness that transitions more easily into sleep

Dosing: 100–200 mg, 30–60 minutes before bed. The 400 mg used in Lyon et al. was split across the day; evening-only dosing at 100–200 mg is effective for sleep applications.

Notes: L-Theanine pairs synergistically with low-dose melatonin and glycine. Safe for long-term use. No dependency, tolerance, or withdrawal observed in the literature.

Apigenin

Evidence level: Moderate for chamomile extract; Weak for isolated apigenin

⚠️ Honest caveat — Huberman vs. the evidence:

Andrew Huberman (Stanford) has popularized 50 mg of isolated apigenin as a sleep supplement. Apigenin is a flavonoid found at high concentrations in chamomile (Matricaria chamomilla), parsley, and celery. It is a partial GABA-A receptor agonist (benzodiazepine site) with demonstrable anxiolytic and mild sedative properties in animal models.

However, the human RCT evidence base is for chamomile extract, not purified apigenin. The meta-analyses that support sleep and anxiety benefits (see Hieu et al., 2019 below) used standardized chamomile extracts containing multiple bioactive compounds, of which apigenin is the primary candidate — but extraction ratios, synergistic flavonoids, and dose-response data for isolated apigenin specifically are not well-established in human trials.

Key study: Hieu et al. (2019) — systematic review and meta-analysis of randomized and quasi-randomized trials examining chamomile for state anxiety, GAD, sleep quality, and insomnia. Significant improvements in sleep quality and anxiety found across trials. PMID: 31006899.

A more recent systematic review and meta-analysis (Kazemi et al., 2024, PMID: 39106912, published in Complementary Therapies in Medicine) specifically on chamomile and sleep included 10 studies (772 participants) and confirmed beneficial effects on sleep quality outcomes.

Mechanisms (plausible for apigenin):

• GABA-A receptor positive allosteric modulator (benzodiazepine site — weaker than pharmaceuticals)
• Adenosine receptor agonism (A1 and A2A)
• Inhibition of CYP2C9 (enzyme relevant to estrogen metabolism — relevant caveat for women)

Dosing: 50 mg isolated apigenin is the Huberman-cited dose; this is mechanistically reasonable but not directly RCT-validated for isolated apigenin. Chamomile extract standardized to ≥1.2% apigenin at 270–540 mg total is better supported by the actual trial literature.

Practical recommendation: Use a chamomile extract product rather than isolated apigenin for better alignment with the evidence base. If using isolated apigenin, 50 mg is a reasonable starting dose with an acceptable safety profile.

Glycine

Evidence level: Strong (multiple human RCTs + established mechanism)

Glycine is a non-essential amino acid and the smallest of all amino acids. It functions as both an inhibitory neurotransmitter (glycine receptors) and as a co-agonist at excitatory NMDA receptors. The sleep effects of oral glycine involve a specific, elegant mechanism: vasodilation of peripheral blood vessels → heat dissipation → reduction in core body temperature → accelerated sleep onset and increased slow-wave sleep.

Key study: Yamadera et al. (2007) — crossover RCT in volunteers with chronic sleep complaints. Intervention: 3 g glycine taken 1 hour before bedtime. Results: significant improvements in subjective sleep quality, fatigue, liveliness, and clarity on the morning after. Polysomnographic data showed shortened sleep latency and increased slow-wave sleep. Published in Sleep and Biological Rhythms (DOI: 10.1111/j.1479-8425.2007.00262.x).

Follow-up study: Bannai et al. (2012) — RCT in partially sleep-restricted healthy volunteers. Intervention: 3 g glycine before bed. Results: significant improvements in subjective daytime sleepiness, fatigue, and performance measures on the following day, compared to placebo. Published in Frontiers in Neurology. PMC: PMC3328957.

Mechanistic confirmation: Kawai et al. (2009) and the Neuropsychopharmacology study (2015, PMID: 25533534) confirmed that glycine’s sleep-promoting and hypothermic effects are mediated by NMDA receptors in the suprachiasmatic nucleus (SCN), with downstream vasodilation causing the core body temperature drop.

Core body temperature and sleep: Sleep onset requires a 0.5–1°C drop in core temperature. This is why warm baths/showers 60–90 minutes before bed (counterintuitively) help sleep — the subsequent cooling accelerates sleep onset. Glycine mimics this effect from the inside.

Dosing: 3 g, taken 30–60 minutes before bed. Can be taken as powder in water (flavorless, mildly sweet). Capsules are available but the dose is bulky (typically 6+ capsules at 500 mg each).

Safety: Exceptionally safe. Glycine is an endogenous amino acid with no known toxicity at standard supplemental doses. Well tolerated chronically.

Bottom line: Glycine is arguably the most underrated sleep supplement in the evidence-based space. Strong mechanistic rationale, multiple RCTs, accessible cost, excellent safety. Consider making it the cornerstone of any sleep stack.

Low-Dose Melatonin

Evidence level: Strong (multiple RCTs + robust mechanistic data)

Melatonin is perhaps the most widely misused supplement in the sleep category. Over-the-counter dosing in the United States (1–10 mg, often 5 mg as the standard unit) is dramatically higher than the dose required for circadian effects.

Key studies:

• Zhdanova et al. (1996) — foundational dose-response study demonstrating that melatonin doses as low as 0.3 mg (physiological) are sufficient to raise serum melatonin to nocturnal levels and facilitate sleep onset as measured by polysomnography. Doses above this produce supraphysiological levels without added benefit for sleep onset. PMID: 8843534.

• Zhdanova et al. (2001) — RCT in age-related insomnia subjects. Low-dose melatonin (0.3 mg and 1.0 mg) significantly reduced sleep onset latency and restored normal sleep architecture. Published in J Clin Endocrinol Metab. PMID: 11600532.

The dose problem: A standard 5 mg melatonin tablet produces blood melatonin levels 10–20× higher than the physiological nocturnal peak. This causes:

1. Morning grogginess (melatonin’s half-life is 30–60 minutes, but sustained elevation from large doses extends into morning)
2. Disruption of natural sleep architecture
3. Potential for tolerance over time (receptor downregulation with chronic supraphysiological dosing)
4. Suppression of endogenous melatonin production with prolonged high-dose use (debated in the literature, but a reasonable precaution)

Melatonin’s role is circadian, not sedative: Melatonin does not cause sedation directly. It signals “biological night” to the SCN and peripheral tissues. It is most useful for:

• Circadian misalignment (shift work, jet lag, delayed sleep phase)
• Advancing sleep timing in evening chronotypes
• Age-related melatonin decline (significant after age 50)
• Transition support when changing sleep schedules

Dosing: 0.3–0.5 mg (physiological dose), taken 90–120 minutes before desired bedtime. Use the lowest effective dose. Do not use nightly if sleep timing is not a primary issue — reserve for circadian reset applications.

Note on commercial availability: 0.3 mg tablets are available (e.g., Life Extension 0.3 mg Melatonin). Alternatively, a 1 mg tablet can be split, or sublingual 0.5 mg products can be used.

GABA (Oral Supplement)

Evidence level: Weak-Moderate (limited human data, BBB controversy)

Oral GABA supplementation is frequently marketed for sleep, but its mechanism is genuinely controversial. The central question is whether exogenous GABA can cross the blood-brain barrier in sufficient quantities to exert CNS effects.

The BBB problem: GABA is a charged molecule at physiological pH and lacks a known high-affinity transporter across the BBB in adults. Most neuropharmacologists are skeptical that oral GABA has direct central effects.

The enteric pathway hypothesis: An alternative mechanism — proposed based on Yamatsu et al. (2016, Food Science and Biotechnology, DOI: 10.1007/s10068-016-0076-9) — suggests that oral GABA may act on GABA-B receptors in the enteric nervous system, indirectly modulating the vagus nerve and reducing sympathetic arousal. This gut-brain axis mechanism bypasses the BBB question.

Key study: Yamatsu et al. (2016) — small controlled study showing that 100 mg oral GABA reduced sleep latency and improved subjective sleep quality. The mechanism was not definitively established.

Additional supporting data: A 2020 systematic review of oral GABA and sleep (PMC7527439) concluded: “There is only very limited supportive evidence regarding the role of oral GABA intake on objective sleep improvement.” Not sufficient to strongly recommend as a primary sleep supplement.

Practical verdict: GABA may have modest sleep-onset benefits via the enteric pathway. If included in a stack, 100 mg 30–60 min before bed. However, better evidence exists for the other compounds listed here. Do not prioritize GABA over glycine, theanine, or magnesium.

Timing Protocol (Critical)

Timing is as important as dose. Most supplement failures occur because compounds are taken immediately before bed rather than at the appropriate pre-sleep window.

Legend: 🔴 = High priority (strongest evidence) | 🟡 = Medium priority (good evidence, additive) | 🟢 = Situational (use when appropriate) | 🔵 = Optional (weaker evidence)

Practical Timing Example (10:30 PM target bedtime)

• 8:30–9:00 PM: Melatonin 0.3 mg (if using for circadian reset)
• 9:00–9:30 PM: Magnesium glycinate 300–400 mg elemental
• 9:30–10:00 PM: Glycine 3 g + L-Theanine 200 mg + Apigenin 50 mg (or chamomile extract)
• 10:30 PM: Lights out

Stack Combinations

Minimum Stack — 2 Compounds (Starting Point)

Best first combination: Magnesium glycinate + Glycine

Rationale: These two have the strongest individual evidence bases, complementary mechanisms (GABA modulation + core temperature reduction), and address both sleep onset and deep sleep quality. Cost-effective. Well tolerated.

• Magnesium glycinate: 300 mg elemental, 60–90 min before bed
• Glycine: 3 g, 30–60 min before bed

Expected effects: Improved sleep onset, longer slow-wave sleep, reduced early morning awakening (especially in magnesium-deficient individuals). Subjective morning freshness improvement within 1–2 weeks.

Alternative 2-compound stack: L-Theanine + Melatonin (if circadian timing is the primary issue)

Core Stack — 4 Compounds

Magnesium glycinate + Glycine + L-Theanine + Apigenin/Chamomile

This combination addresses the full spectrum: deep sleep quality, sleep onset, anxiolytic relaxation, and GABA modulation via multiple pathways simultaneously.

Melatonin is deliberately omitted from the core stack — use it separately only when circadian misalignment is present.

Expected effects: Meaningful improvement in sleep onset latency, sleep efficiency, and subjective sleep quality. Most users report noticeable effects within 3–7 days. Full adaptation takes 2–4 weeks.

Advanced Stack — 5+ Compounds

For individuals with persistent sleep difficulties who have already optimized sleep hygiene and seen partial benefit from the core stack.

Note on magnesium stacking: Using both glycinate and threonate simultaneously provides different elemental amounts and different mechanistic targets. Total elemental magnesium should remain below 500–600 mg/day from all supplemental sources to avoid GI discomfort. The Tolerable Upper Intake Level (UL) for supplemental magnesium is 350 mg/day from the AGES/DRI guidelines — above this, laxative effects become likely with poorly absorbed forms, but less so with highly bioavailable forms.

Sleep Hygiene First (Non-Supplement)

No supplement stack can overcome poor sleep fundamentals. These interventions are prerequisites, not optional add-ons.

Temperature

Core body temperature must drop 0.5–1°C for sleep onset. Optimal bedroom temperature: 16–19°C (60–67°F). A warm shower or bath 60–90 min before bed paradoxically accelerates sleep onset by pulling blood to the periphery, triggering the post-shower cooling response.

Light Exposure

• Morning: Bright light (natural sunlight or 10,000 lux light box) within 30–60 minutes of waking. This anchors the circadian clock and advances the timing of evening melatonin onset.
• Evening: Avoid bright overhead light and blue-light-rich screens in the 60–90 minutes before bed. Candlelight-level illumination is ideal. Blue light blocking glasses (orange-tinted) are a pragmatic middle ground if screen elimination is unrealistic.
• Night: Total darkness during sleep. Even low-level light through closed eyelids (particularly blue/green wavelengths) measurably suppresses melatonin and disrupts sleep architecture.

Consistency

Sleep timing consistency is the single most powerful variable for sleep quality. A regular bedtime ±30 minutes, even on weekends, dramatically improves all sleep metrics. The “social jet lag” caused by sleeping in on weekends shifts the circadian clock and degrades Monday–Tuesday sleep quality.

Caffeine Cutoff

Caffeine’s half-life is approximately 5–7 hours (individual variation: 3–12 hours based on CYP1A2 genetics). A standard recommendation is no caffeine after 12 PM for a 10 PM bedtime. In slow caffeine metabolizers, even morning coffee can measurably impair slow-wave sleep architecture measured objectively that night.

Alcohol

Alcohol reduces sleep onset latency (it feels sedating) but significantly fragments sleep in the second half of the night, suppresses REM, and causes earlier awakening. There is no safe dose of alcohol for sleep quality — even one drink measurably degrades sleep architecture.

Tracking & Optimization

Wearable Devices

Consumer wearables cannot precisely measure individual sleep stages (only polysomnography provides ground truth), but they are useful for tracking trends over time.

Oura Ring: Generally considered the most accurate consumer sleep tracker for sleep staging and HRV. Track:

• Sleep efficiency (target ≥85%)
• Total deep sleep (N3) duration (target: ≥20% of total sleep time)
• REM sleep duration (target: ≥20% of total sleep time)
• HRV (Heart Rate Variability) — higher nighttime HRV correlates with parasympathetic dominance and sleep quality; look for trend improvements over 2–4 weeks with stack introduction
• Resting heart rate — lower values generally indicate better recovery

Garmin (Fenix/MARQ/Forerunner series): Comparable sleep staging with Body Battery metric. Less accurate for sleep stage distinction but reliable for sleep duration and HRV.

Apple Watch: Sleep data is basic; primarily useful for total sleep duration. Third-party apps (AutoSleep, Pillow) add limited staging.

Interpreting Data

• Introduce one supplement at a time for 7–10 days before adding another. This allows identification of which compound produces the most benefit for your individual biology.
• Track subjective morning refresh score (0–10) alongside wearable data — sometimes subjective improvement precedes measurable wearable changes.
• Do not over-optimize: sleep data variability is high night to night. Look for 2–4 week trend lines, not individual nights.
• HRV improvement with sleep stack is a meaningful signal — it indicates the nervous system is better recovered and the stack is working at a physiological level.

Special Populations

ADHD

Sleep difficulties are highly prevalent in ADHD — sleep onset insomnia affects 70–80% of adults with ADHD, driven by hyperarousal, delayed circadian phase (evening chronotype predominance), and stimulant medication timing.

Best evidence for ADHD + sleep:

• L-Theanine: Lyon et al. (2011) specifically studied ADHD boys. Improves sleep efficiency without requiring dose reduction of stimulant medications. Generally safe with stimulants.
• Melatonin: Multiple RCTs in pediatric ADHD (Van der Heijden et al., Weiss et al.) demonstrate meaningful reductions in sleep onset latency. Low dose (0.5–1 mg) at appropriate circadian timing is the standard approach.
• Magnesium: ADHD populations often show higher rates of magnesium deficiency. Magnesium supplementation addresses both sleep and may have modest effects on hyperactivity symptoms.

Stimulant timing note: Afternoon/evening stimulant doses are a major contributor to ADHD-related sleep onset insomnia. The most impactful intervention is often medication timing adjustment rather than supplementation.

Pregnancy

• Melatonin: Avoid during pregnancy. Melatonin receptors are present in placental tissue and the fetal brain; exogenous melatonin during pregnancy has not been adequately studied for safety. The endogenous melatonin profile during pregnancy serves developmental signaling functions.
• Magnesium glycinate: Generally considered safe and often beneficial during pregnancy (common deficiency). Consult OB/midwife for dose.
• L-Theanine: Insufficient human safety data during pregnancy. Caution advised.
• Glycine: Endogenous amino acid; generally safe. Glycine demand actually increases during pregnancy (placental and fetal development). Low-dose supplementation (3 g) is unlikely to be harmful, but consult healthcare provider.
• Chamomile/Apigenin: Chamomile has traditionally been avoided in pregnancy due to potential uterine stimulant effects in high doses. Do not use during pregnancy.

Elderly (65+)

• Melatonin production declines significantly with age — this is a primary driver of age-related sleep quality decline. Low-dose melatonin (0.3–1 mg) may be particularly beneficial in this population. Abbasi et al. (2012) specifically studied elderly insomniacs.
• Magnesium absorption also decreases with age; supplement priority is higher.
• Reduced kidney function in some elderly individuals warrants attention to total supplemental loads — consult physician if any renal concerns.
• Glycine’s safety profile is appropriate for elderly populations and may additionally support glucose metabolism and collagen synthesis.

Safety & Interactions

Melatonin

• + Hormonal contraceptives (estrogen-containing): Estrogen may inhibit melatonin metabolism via CYP1A2 inhibition, potentially elevating melatonin levels. Monitor for next-day grogginess.
• + Diabetes medications / insulin: Melatonin impairs glucose tolerance and can reduce insulin sensitivity acutely. Timing of melatonin relative to eating is important. Not recommended close to meals.
• + Immunosuppressants: Melatonin has immune-modulating properties; use with caution in transplant patients on immunosuppression.
• + Anticoagulants (warfarin): Some evidence of additive anticoagulant effects. Monitor INR if combining.
• + Caffeine: Caffeine interferes with melatonin onset. The primary interaction is behavioral (avoid caffeine before bed), not pharmacological.

L-Theanine

• + Sedatives / benzodiazepines / Z-drugs: Additive CNS depression possible. Reduce sedative dose if combining; do not use with alcohol and sedatives simultaneously.
• + Stimulant medications (methylphenidate, amphetamines): L-Theanine does not antagonize stimulant effects meaningfully at standard doses. This combination (theanine + stimulant) is actually studied and generally safe.
• + Antihypertensives: Theanine has mild blood pressure-lowering effects; additive effect possible with antihypertensive medications.

Magnesium

• + Antibiotics (fluoroquinolones, tetracyclines): Magnesium chelates these antibiotics, reducing absorption by 40–90%. Separate administration by at least 2–4 hours.
• + Bisphosphonates (osteoporosis drugs): Similar chelation issue. Take medications 2+ hours before magnesium.
• + Diuretics (thiazide-type): Increase magnesium excretion — supplementation may be particularly beneficial.
• + Potassium-sparing diuretics: Increase magnesium retention — risk of hypermagnesemia (rare at standard supplement doses but monitor).
• + Proton pump inhibitors (Omeprazole, etc.): Chronic use can cause hypomagnesemia; supplementation is appropriate.

Apigenin (Chamomile)

• + CYP2C9 substrates: Apigenin inhibits CYP2C9. This enzyme metabolizes warfarin, NSAIDs (diclofenac), some antidiabetics (glipizide), and phenytoin. Elevated drug levels possible with concurrent chamomile supplementation.
• + Hormonal medications: Apigenin has weak estrogenic activity and inhibits aromatase. In estrogen-sensitive conditions (hormone-receptor positive breast cancer, endometriosis), use with caution and medical guidance.
• + Sedatives: Additive CNS depression at high doses.

Glycine

• + Clozapine: Glycine has been studied as an adjunct in schizophrenia (NMDA enhancement strategy). Do not use high-dose glycine (>15 g/day) with clozapine without psychiatric guidance — though 3 g sleep doses are unlikely to be clinically significant.
• General safety: Glycine is one of the safest compounds on this list. No significant drug interactions at 3 g doses are well-documented.

Limitations & Caveats

The Apigenin Evidence Gap

The most significant honest caveat in this protocol is the apigenin evidence gap. The mechanistic basis for apigenin’s GABA-A agonism is well-established in preclinical studies. The clinical evidence, however, is for chamomile extract — a complex botanical preparation containing apigenin alongside apigenin-7-glucoside, luteolin, quercetin, and other bioactive compounds.

No human RCT to our knowledge has isolated 50 mg apigenin against placebo with sleep as a primary endpoint using objective polysomnographic measurement. The Huberman recommendation of 50 mg isolated apigenin is mechanistically defensible but not directly proven by human trial. Users should understand they are extrapolating from (a) animal pharmacology, (b) chamomile extract RCTs, and (c) mechanistic plausibility — not from a direct apigenin RCT.

Population Generalizability

Several key studies were conducted in populations with compromised sleep or specific diagnoses:

• Abbasi et al. (2012): elderly with primary insomnia
• Lyon et al. (2011): boys with ADHD
• Yamadera et al. (2007): volunteers with sleep complaints

Extrapolation to healthy young adults without sleep disorders is reasonable for some compounds (glycine, theanine) but the effect size may be smaller in already-good sleepers.

Measurement Validity

Many sleep supplement trials use subjective self-report measures (PSQI, ISI, VAS scales) rather than polysomnography. Subjective improvement does not always correlate with objective sleep architecture changes. Where polysomnographic data exists (Yamadera 2007 for glycine; Zhdanova 1996/2001 for melatonin; Lyon 2011 using actigraphy), the evidence is more compelling.

Individual Variation

Response to sleep supplements varies significantly based on:

• Baseline magnesium status (higher baseline deficiency → stronger magnesium response)
• Chronotype and circadian alignment (determines melatonin utility)
• Genetic CYP1A2 polymorphisms (affects caffeine/melatonin metabolism)
• Baseline anxiety levels (higher anxiety → greater theanine and apigenin benefit)
• Gut microbiome composition (may affect GABA production and oral GABA effects)

No single stack works identically for everyone. The protocol presented here is a starting framework requiring individual calibration.

What These Supplements Cannot Do

• Compensate for insufficient sleep duration. If you are sleeping 5 hours when you need 8, no supplement stack will fix cognitive impairment or health consequences.
• Treat clinical sleep disorders. Obstructive sleep apnea, restless legs syndrome, narcolepsy, and clinical insomnia disorder require medical evaluation and treatment. Supplements are adjunctive, not therapeutic.
• Permanently reset a misaligned circadian rhythm without behavioral changes (consistent sleep timing, light management).

The Bottom Line

The evidence-based sleep supplement stack, ranked by evidence quality and practical impact:

1. Glycine (3 g, 30–60 min before bed) — Most underrated. Strong RCT evidence. Unique mechanism (core temperature lowering). Cost-effective. Start here if you want one compound.

2. Magnesium glycinate (300–400 mg elemental, 60–90 min before bed) — Most broadly applicable. Addresses the widespread dietary deficiency. GABA modulation + cortisol lowering + circadian enzyme support. The cornerstone of any sleep stack.

3. L-Theanine (100–200 mg, 30–60 min before bed) — Particularly useful for anxiety-driven sleep onset insomnia and ADHD-related hyperarousal. Non-sedating. Synergizes well with everything else on this list.

4. Apigenin/Chamomile (50 mg apigenin or 400–500 mg chamomile extract standardized to ≥1.2% apigenin, 30–60 min before bed) — Reasonable addition for anxiolytic and GABA-A modulation. Use chamomile extract for better evidence alignment.

5. Melatonin (0.3–0.5 mg, 90–120 min before bed) — Use only when circadian misalignment is the primary issue. Not a nightly supplement for most people. The dose sold in most stores is 10–20× too high.

The non-negotiables: No supplement stack outperforms consistent sleep timing, dark and cool sleeping environment, morning light exposure, and caffeine cutoff before noon. Address these first. Add supplements as enhancements, not replacements.

Sources

1. Yamadera W et al. (2007). Glycine ingestion improves subjective sleep quality in human volunteers, correlating with polysomnographic changes. Sleep and Biological Rhythms, 5(2), 126–131. DOI: 10.1111/j.1479-8425.2007.00262.x

2. Bannai M, Kawai N, Ono K, Nakahara K, Murakami N. (2012). The effects of glycine on subjective daytime performance in partially sleep-restricted healthy volunteers. Frontiers in Neurology, 3, 61. PMC: PMC3328957. PMID: 22529837

3. Kawai N et al. (2015). The sleep-promoting and hypothermic effects of glycine are mediated by NMDA receptors in the suprachiasmatic nucleus. Neuropsychopharmacology, 40(6), 1405–1416. PMID: 25533534

4. Abbasi B, Kimiagar M, Sadeghniiat K, Shirazi MM, Hedayati M, Rashidkhani B. (2012). The effect of magnesium supplementation on primary insomnia in elderly: A double-blind placebo-controlled clinical trial. Journal of Research in Medical Sciences, 17(12), 1161–1169. PMID: 23853635; PMC: PMC3703169

5. Lyon MR, Kapoor MP, Juneja LR. (2011). The effects of L-theanine (Suntheanine®) on objective sleep quality in boys with attention deficit hyperactivity disorder (ADHD): a randomized, double-blind, placebo-controlled clinical trial. Alternative Medicine Review, 16(4), 348–354. PMID: 22214254

6. Zhdanova IV, Wurtman RJ, Morabito C, Piotrovska VR, Lynch HJ. (1996). Effects of low oral doses of melatonin, given 2–4 hours before habitual bedtime, on sleep in normal young humans. Sleep, 19(5), 423–431. PMID: 8843534

7. Zhdanova IV, Wurtman RJ, Regan MM, Taylor JA, Shi JP, Leclair OU. (2001). Melatonin treatment for age-related insomnia. Journal of Clinical Endocrinology & Metabolism, 86(10), 4727–4730. PMID: 11600532. DOI: 10.1210/jcem.86.10.7901

8. Hieu TH, Dibas M, Surya Dila KA, et al. (2019). Therapeutic efficacy and safety of chamomile for state anxiety, generalized anxiety disorder, insomnia, and sleep quality: A systematic review and meta-analysis of randomized trials and quasi-randomized trials. Phytotherapy Research, 33(6), 1604–1615. PMID: 31006899

9. Kazemi A et al. (2024). Effects of chamomile (Matricaria chamomilla L.) on sleep: A systematic review and meta-analysis of clinical trials. Complementary Therapies in Medicine, 84, 103071. PMID: 39106912

10. Slutsky I, Abumaria N, Wu LJ, Huang C, Zhang L, Li B, Zhao X, Govindarajan A, Zhao MG, Bhaskaran M, Bhansali S, Bhansali P. (2010). Enhancement of learning and memory by elevating brain magnesium. Neuron, 65(2), 165–177. PMID: 20152124. DOI: 10.1016/j.neuron.2009.12.026

11. Yamatsu A, Yamashita Y, Pandharipande TN, Maru I, Kim M. (2016). Effect of oral γ-aminobutyric acid (GABA) administration on sleep and its absorption in humans. Food Science and Biotechnology, 25(2), 547–551. DOI: 10.1007/s10068-016-0076-9

12. Bannai M, Kawai N. (2012). New therapeutic strategy for amino acid medicine: glycine improves the quality of sleep. Journal of Pharmacological Sciences, 118(2), 145–148. PMID: 22293292

13. Okamoto-Mizuno K, Mizuno K. (2012). Effects of thermal environment on sleep and circadian rhythm. Journal of Physiological Anthropology, 31, 14. PMC: PMC3427038

14. Leproult R, Van Cauter E. (2010). Role of sleep and sleep loss in hormonal release and metabolism. Endocrine Development, 17, 11–21. PMID: 19955752

15. Boyle NB, Lawton C, Dye L. (2017). The effects of magnesium supplementation on subjective anxiety and stress — A systematic review. Nutrients, 9(5), 429. PMC: PMC5452159

16. Kimura K, Ozeki M, Juneja LR, Ohira H. (2007). L-Theanine reduces psychological and physiological stress responses. Biological Psychology, 74(1), 39–45. PMID: 16930802

17. Rao TP, Ozeki M, Juneja LR. (2015). In search of a safe natural sleep aid. Journal of the American College of Nutrition, 34(5), 436–447. PMID: 25759004

Revision History