Affiliate disclosure: Some links in this article are affiliate links. We may earn a commission at no additional cost to you.

Medical disclaimer: This article reviews published research on glycine supplementation and collagen synthesis. It is not medical advice, diagnosis, or treatment. Not medical advice. Consult a qualified healthcare professional before changing your diet, supplement regimen, or any health-related decision.

Glycine is marketed as a collagen-supporting amino acid, a sleep aid, and a longevity compound — sometimes simultaneously, sometimes by the same product. The overlap is real but the evidence behind each claim has different strength. What the Japanese amino acid research actually shows is more specific and more useful than the combined marketing narrative suggests.

Ajinomoto, Japan’s largest amino acid producer, published the primary human clinical data on glycine and sleep through the 2000s and 2010s — trials that stand out in the supplement literature for using polysomnography and next-day performance measures rather than self-report alone. Separately, Japanese investigators contributed to the collagen peptide supplementation research that expanded in that same period, primarily through skin elasticity studies. Both bodies of evidence deserve examination on their own terms.

TL;DR

• Glycine constitutes roughly one-third of collagen’s amino acid composition; the Gly-X-Y repeat structure of collagen’s triple helix requires glycine at every third position without exception
• Japanese amino acid researchers (Ajinomoto group) published controlled trials showing 3g glycine before sleep was associated with reduced sleep latency and improved objective sleep quality measures in individuals with self-reported sleep difficulties
• The proposed mechanism — peripheral vasodilation driving core body temperature reduction at sleep onset — is distinct from sedative or GABAergic mechanisms used by sleep medications
• Meléndez-Hevia et al. (2009, Journal of Biosciences) proposed a “glycine gap” in which endogenous synthesis plus dietary intake may not meet total metabolic demand for collagen maintenance; subsequent analyses have questioned the estimated magnitude
• Collagen peptide supplementation trials (2.5–10g/day over 8–12 weeks) show associations with skin elasticity and hydration improvements in small RCTs; most are industry-funded; endpoints are surrogate biomarkers, not longevity measures
• Glycine at 3g/night appears well-tolerated in published human trials; long-term high-dose data in healthy adults remains limited

Glycine as collagen’s structural backbone

Collagen’s defining molecular feature is the triple helix — three polypeptide chains wound around each other in a right-handed superhelix. The structure is possible only because glycine, the smallest amino acid with no side chain, fits at every third position in the repeating Gly-X-Y sequence. Proline and hydroxyproline occupy the X and Y positions predominantly. Any substitution at the glycine position disrupts helix geometry; this is the mechanism underlying osteogenesis imperfecta and several connective tissue disorders caused by glycine-to-serine substitutions in collagen genes.

Collagen is the most abundant protein category in the human body by mass, comprising approximately 25-35% of total body protein. Type I collagen — found in skin, bone matrix, tendons, and cornea — is the most abundant single protein. Given that glycine is approximately one-third of collagen’s amino acid content by molar fraction, the body’s daily glycine demand for collagen maintenance is substantial.

The body synthesizes glycine endogenously, primarily through the serine hydroxymethyltransferase reaction and the glycine cleavage system in the liver, at roughly 3g/day in adults. Dietary protein contributes an additional 2-5g/day depending on intake composition — fish, meat, and legumes all carry glycine-containing protein, and Japanese dietary patterns with their emphasis on fish-based broths and dashi contribute meaningful amounts. Meléndez-Hevia et al. published a 2009 calculation in Journal of Biosciences estimating that total glycine demand from collagen maintenance and other metabolic pathways might significantly exceed what biosynthesis and dietary intake together supply — a “glycine gap” they put at roughly 10g/day. The calculation has been cited in the supplement literature as justification for glycine supplementation. Subsequent analyses have questioned both the accounting method and whether all identified demands represent net glycine consumption rather than cycling, so the gap should be treated as a testable hypothesis rather than a measured deficit.

Collagen decline with age

Collagen density in human skin begins declining from roughly the mid-20s, with published estimates of 1-1.5% annual decline in skin collagen content after age 30. With chronic UV exposure the measured loss accelerates substantially. The age-associated change is not limited to quantity: aberrant cross-linking accumulates through glycation reactions — collagen fibers reacting with glucose and other reducing sugars to form advanced glycation end-products — and through oxidative modification. These cross-linking changes stiffen collagen networks independently of their quantity, contributing to increased arterial rigidity, reduced skin elasticity, and changes in cartilage mechanics.

The cross-linking dimension matters for interpreting glycine supplementation claims. Dietary glycine availability primarily addresses the synthesis supply side; it does not resolve existing cross-linked structures. Whether maintaining glycine availability in older adults with declining endogenous synthesis and dietary intake would support collagen synthesis rates is a biologically coherent hypothesis, but human trial data on this specific question — as distinct from the skin elasticity marker studies discussed below — has not been published in the same volume or quality as the sleep data.

Sleep: the Ajinomoto research program

The glycine and sleep evidence is the strongest clinical case for dietary glycine supplementation, and it comes primarily from Ajinomoto’s amino acid research group.

Inagawa and colleagues published initial nap observations in Sleep and Biological Rhythms in 2006, finding that 3g glycine administered before a scheduled daytime nap was associated with reduced subjective sleepiness following the nap compared to placebo in a small crossover design.

The more substantive evidence came through two 2012 publications from Bannai and colleagues. The first, by Bannai and Kawai in Journal of Pharmacological Sciences (PMID 22293292), reported a randomized placebo-controlled crossover trial in 11 subjects with self-reported sleep difficulties. Participants received 3g glycine or placebo before bed. Key findings from polysomnography and standardized questionnaires:

• Sleep latency: shorter time to sleep onset in the glycine condition
• Sleep architecture: reduced wakefulness during the first sleep cycle; more slow-wave sleep in the glycine condition in follow-up intervals
• Subjective quality: morning ratings of sleep satisfaction improved in the glycine arm relative to placebo

The second 2012 publication by Bannai, Kawai, and colleagues in Frontiers in Neurology (PMID 22529837) examined next-day consequences with a broader sample. Participants taking 3g glycine the previous night showed significant reductions in next-day fatigue and sleepiness on standardized rating scales compared to those taking placebo, alongside improvements on a psychomotor vigilance task used to assess reaction-time-based alertness.

The proposed mechanism is physiologically coherent. Glycine receptors in peripheral vasculature, when activated, produce vasodilation in cutaneous vessels — increasing skin blood flow and facilitating the core body temperature drop that naturally precedes and supports sleep onset. Core body temperature lowering is a well-established physiological correlate of sleep onset; it is the same principle underlying the Japanese practice of warm bathing before sleep, where initial bath-mediated warming produces rebound heat dissipation that supports sleep timing. Glycine also serves as a required co-agonist at the NMDA receptor’s GluN1 subunit alongside glutamate, and modulation of specific sleep-wake neural circuits through this pathway has been proposed as a secondary mechanism, though the vasodilation pathway is the better-supported route for the specific effect observed in the trials.

The standard caveats apply: sample sizes across the published Ajinomoto glycine sleep trials total fewer than 100 subjects; the research group has a commercial interest in glycine product sales; and no long-term trials have examined whether the sleep effects persist over months of consistent use or whether tolerance develops. These limitations do not invalidate the controlled trial evidence, but they define the calibrated interpretation: 3g glycine before sleep appears associated with measurable short-term sleep quality improvements in trials conducted by a commercially affiliated group and published in peer-reviewed journals. Independent replication at larger scale would strengthen the evidence considerably.

Collagen peptide supplementation: skin marker evidence

The collagen peptide supplementation literature developed alongside the glycine sleep research, though the supplements differ — hydrolyzed collagen peptides contain a full amino acid profile weighted toward glycine, proline, and hydroxyproline, not isolated glycine.

Proksch et al. (2014, Skin Pharmacology and Physiology, PMID 23949208) enrolled 69 women aged 35–55 in an 8-week double-blind placebo-controlled trial of 2.5g or 5g bioactive collagen peptides daily. Cutometer-measured skin elasticity showed improvement in both active groups versus placebo at 4 weeks, with the effect persisting at the 8-week time point. Effect sizes were in the range of 7-15% improvement relative to baseline on the elasticity parameter.

Asserin et al. (2015, Journal of Cosmetic Dermatology, PMID 26362110) examined 105 women over 12 weeks with collagen hydrolysate supplementation, finding improvements in skin hydration and dermis density measured by ultrasound compared to placebo.

Japanese investigator-led trials and industry-supported studies examining collagen peptide supplementation in Japanese female cohorts have produced broadly consistent directional findings on skin moisture and viscoelasticity measures, contributing to the relatively large consumer market for collagen supplement products in Japan.

Industry funding is common across this research category, and the endpoints — skin elasticity and hydration biomarkers — are surrogate measures. The biological chain from oral collagen peptides to measurable skin collagen change is plausible: hydroxyproline-containing di- and tripeptides are detectable in blood following oral intake in human studies, and in vitro work shows these peptides can stimulate fibroblast collagen synthesis. But the mechanism from surrogate endpoint to a longevity-relevant clinical outcome has not been established.

Sourcing and dose context

For the sleep application, the dose used consistently across the Japanese trials is 3g taken 30-60 minutes before bed. Glycine powder is essentially tasteless and dissolves readily in water; capsule forms are available but less economical at this dose. Glycine supplements from established amino acid suppliers are available on Amazon.

For collagen peptides, the dose range in published trials runs from 2.5g to 10g/day. Hydrolyzed collagen peptide products from bovine and marine sources are available on Amazon. Marine collagen peptides (typically Type I, derived from fish skin and scales) have been proposed to have marginally higher bioavailability based on lower mean peptide molecular weight, but head-to-head comparative bioavailability data in humans remains limited. Both bovine and marine Type I collagen products are appropriate for the skin elasticity application studied in the published trials.

What the evidence does not establish

The sleep research is specific to 3g pre-sleep doses and short-term trial durations. Whether the observed effects persist with months of consistent use, whether effective dose varies substantially between individuals, or whether glycine supplementation compares favorably to standard sleep hygiene measures (consistent sleep timing, light management, temperature control) has not been addressed in published data.

The collagen synthesis pathway from dietary glycine supplementation to increased tissue collagen in aging humans — the central longevity claim in most glycine supplement marketing — has not been tested in a well-powered human clinical trial. The skin elasticity data from collagen peptide trials involves hydrolyzed collagen as the supplement form, not isolated glycine, and active peptide components include proline and hydroxyproline residues alongside glycine. Attributing those effects specifically to the glycine fraction is not supported by the trial designs. Whether the glycine gap hypothesis, if accurate, translates to a clinically meaningful deficit in older adults that supplementation could address remains an open research question.

For those exploring dietary approaches to sleep quality, glycine at 3g before bed represents one of the better-evidenced options in the dietary supplement category — mechanistically coherent, tested in controlled human trials, and with an acceptable short-term tolerability profile based on published data. Discussing it with a healthcare professional is appropriate before use, particularly for individuals managing neurological conditions given glycine’s role as a co-agonist at NMDA receptors.

For those primarily interested in the skin aging application, collagen peptides are among the more studied dietary supplement interventions in this category, with expectations appropriately set: surrogate marker improvements in short RCTs, not established longevity outcomes.

---

Research cluster: Green Tea and Mortality: What the Ohsaki and JPHC Cohorts Actually Found | Cellular Senescence and Senolytics: p16/SASP Mechanisms and Japan’s Research Contribution | CoQ10 and Ubiquinol: Mitochondrial Aging Evidence and Japan’s Research Context