Glycine is the smallest amino acid there is — a single hydrogen atom where every other amino acid carries a side chain — and it is one of the cheapest compounds in the supplement aisle. That combination usually signals a nothing-burger. Glycine is the exception worth taking seriously, because it is doing four unrelated jobs at once (building collagen, supplying glutathione, disposing of excess methionine, and acting at two different neuroreceptors), and because a large, well-run mouse study found it extended lifespan. The interesting question is not whether glycine matters — it obviously does — but whether the amount you already have is enough, and whether taking more changes anything you can feel or measure. Those are two different questions, and the evidence answers them very differently.
The shortfall that the whole case rests on
Textbooks call glycine non-essential because you can synthesise it, mostly from serine via serine hydroxymethyltransferase. The problem is a stoichiometric ceiling: that reaction cannot produce more glycine than the one-carbon units it generates alongside it, which caps output. Working through every reported production and consumption flux in a 70 kg adult, Meléndez-Hevia and colleagues concluded that synthesis from serine supplies about 3 g/day — more than 85% of all endogenous glycine — while a normal diet contributes another 1.5–3.0 g/day. Against total metabolic demand, collagen turnover included, they estimated a deficit of roughly 10 g/day, and argued glycine should be reclassified as semi-essential.[1]
That is the single most important number on this page, and it deserves to be labelled honestly: it is a modelled shortfall, not a measured requirement, and the flux figures it is built from are estimates with wide error bars. Nobody has run the obvious confirmatory experiment — feeding graded glycine doses and measuring collagen turnover, glutathione synthesis and one-carbon flux simultaneously in people. The framing is not fringe, though. An independent metabolic review reaches the softer version of the same conclusion: glycine “has also been suggested as a conditionally essential amino acid,” and circulating glycine is consistently lower in obesity, type 2 diabetes and fatty liver disease.[11] So: a plausible gap, not a demonstrated one. Everything below is an attempt to see whether filling it does anything.
Collagen: the strongest mechanism, the weakest outcome data
Collagen is roughly one-third glycine by residue count — the repeating Gly-X-Y motif puts a glycine in every third position, because it is the only residue small enough to sit at the crowded centre of the triple helix. Nothing else can substitute. That makes glycine a hard structural requirement for connective tissue in a way that no other single ingredient in the joint-and-skin aisle is, and it is the reason the collagen demand dominates the shortfall calculation above.[1]
It also makes “take glycine for collagen” a more defensible claim in principle than most collagen marketing. If you swallow hydrolysed collagen peptides, they are digested and absorbed as free amino acids and small di- and tripeptides; the ones your body then reassembles into new collagen are, disproportionately, glycine and proline. Glycine powder supplies the rate-limiting one directly and for a tiny fraction of the price. That argument is sound as far as it goes — and it stops sharply at the point where anyone asks for a human outcome.
The best supporting experiment is in a dish. Bovine articular chondrocytes cultured across a range of amino-acid concentrations increased type II collagen synthesis by 60–75% as glycine rose above 1.0 mM, and kept increasing, while the effects of proline and lysine plateaued below that — which the authors read as direct evidence of glycine deficiency for collagen synthesis.[2] That is a real finding in a real model. It is also cultured cattle cartilage, published by the same group that produced the shortfall calculation and that has applied for a patent on using glycine in osteoarthritis. Searching the question from the other side — for a randomised trial of glycine supplementation with a joint, skin or tendon outcome — returns nothing. There is no human evidence that glycine powder improves cartilage, skin or connective tissue. If you want the trial-level picture on the ingredient people actually buy for this, see collagen peptides: the evidence.
Sleep: what is actually behind the 3-gram habit
This is the reason most people buy glycine, and the evidence is more interesting — and more limited — than either the marketing or the sceptics suggest.
The mechanism is unusually well worked out for a supplement. In rats, oral glycine shortened the latency to non-REM sleep and lowered core body temperature; oral and intracerebroventricular glycine raised cutaneous blood flow at the plantar surface dose-dependently, dumping heat. Blocking NMDA receptors abolished that vasodilation, while the glycine-receptor antagonist strychnine did not, and microinjecting glycine directly into the suprachiasmatic nucleus reproduced the effect. Ablate the SCN and the sleep-promoting and hypothermic effects vanish entirely.[5] That is a coherent story: glycine acts as an NMDA co-agonist in the circadian pacemaker, triggers peripheral vasodilation, and the resulting drop in core temperature is the physiological signal that normally precedes sleep onset.[3]
The human trials are where the caution starts. The two studies everyone cites — Inagawa 2006 and Yamadera 2007, reporting that 3 g of glycine before bed improved subjective sleep quality and shortened polysomnographically measured sleep latency in people with persistent sleep complaints — were published in Sleep and Biological Rhythms, which is not indexed in PubMed. What is verifiable is the authors’ own account of them in a review and in a later paper.[3][4] We could not read those primary trials, so we cannot vouch for their sample sizes or methods, and neither can anyone quoting them at you.
The one glycine sleep study you can read end to end is a companion trial in healthy volunteers whose sleep was cut to 75% of normal for three consecutive nights. Three grams of flavoured glycine before bed significantly reduced next-day fatigue on a visual analogue scale, produced a non-significant trend toward less sleepiness, and significantly improved psychomotor vigilance test performance. Melatonin levels and the clock genes Bmal1 and Per2 were unchanged.[4] The honest description of that trial: ten men enrolled, three excluded or withdrawn, seven analysed; randomised but single-blinded; and all participants were employees of the company that funded and ran it.
Glutathione — and what GlyNAC adds that glycine alone does not
Glutathione, the cell’s principal intracellular antioxidant, is a tripeptide of glutamate, cysteine and glycine. Older adults are short of the raw material for it. In a stable-isotope study, elderly subjects had red-cell glycine concentrations of 218 µmol/L versus 487 µmol/L in younger controls — less than half — alongside lower cysteine, lower glutathione, roughly half the fractional synthesis rate, and higher F2-isoprostanes. Two weeks of supplementation with the precursors fully restored glutathione synthesis and concentration and lowered markers of oxidant damage to the point where supplemented elderly subjects were indistinguishable from the young controls.[6]
Here is the part that matters for anyone choosing between products, and the reason this page hands the story off rather than retelling it. Those precursors were cysteine and glycine, given together. Glycine supplies one of the three residues; cysteine supplies the one that is rate-limiting for glutathione synthesis in most tissues most of the time, and it is the one hardest to get from food in a usable form — which is what N-acetylcysteine is for. Searching the question from the opposite direction — for a trial giving glycine alone and measuring glutathione synthesis in older adults — returns nothing. The human glutathione result belongs to the pair, not to glycine.
That pair is GlyNAC, and it is a genuinely different proposition: randomised trials reporting improvement across a broad sweep of aging markers, a replication problem, and an independent trial that missed its primary endpoint. We cover all of it, including the parts that do not flatter it, in GlyNAC: impressive trials, an unsettled replication question. The practical translation: if glutathione is your goal, glycine on its own is the wrong purchase — add cysteine. If sleep, collagen or the methylation story is your goal, the NAC adds cost and a sulfur burden you may not need.
Longevity: the ITP result, stated at its actual size
Glycine is one of a small number of compounds tested by the National Institute on Aging’s Interventions Testing Program — the only lifespan programme that runs the same protocol in genetically heterogeneous mice at three independent sites, precisely because single-lab longevity results so often fail to replicate. An 8% glycine diet produced a 4–6% increase in median lifespan, statistically significant in both males (p = 0.002) and females (p < 0.001), with maximum lifespan also increased. Pooling sexes, the effect went the same direction at all three sites (p = 0.01, 0.053, 0.03). Glycine-fed mice were less likely to die of pulmonary adenocarcinoma, and none of the 40 pathologies scored at necropsy were significantly increased.[7] By ITP standards that is a clean, modest, real positive — and most ITP compounds fail outright.
Now the honest deflation. Eight percent of the diet by weight is an enormous dose. A mouse eating roughly 4 g of chow a day takes in about 0.3 g of glycine, which is on the order of 10 g per kilogram of body weight per day. Even after the conventional body-surface-area scaling used to translate rodent doses, the human equivalent lands near 0.8–1 g/kg/day — on the order of 60 g a day for a 70 kg adult, or about twenty times a 3 g bedtime capsule. Nobody has taken glycine at that intake for years, and no human lifespan or healthspan trial of glycine exists or is likely to. Two further caveats belong in the same breath: one co-author of the ITP paper discloses a company that makes and sells a glycine supplement, and mouse lifespan results, even good ones, have a poor record of translating.
The mechanism, though, is where this gets genuinely interesting, and it is not antioxidant hand-waving. Methionine restriction reliably extends lifespan in rodents. Glycine is the acceptor substrate for glycine N-methyltransferase (GNMT), the enzyme that disposes of excess methionine by stripping a methyl group from S-adenosylmethionine and methylating glycine to sarcosine. Supplemental glycine, on this account, does not mimic methionine restriction metaphorically — it drives the actual disposal reaction, pulling systemic SAM down.[8] The genetic evidence is direct: in Drosophila, overexpressing Gnmt alone suppressed the age-dependent rise in SAM and extended lifespan, and both dietary restriction and reduced insulin signalling depend at least partly on Gnmt to deliver their lifespan effect.[9] So the strongest version of the glycine longevity hypothesis is a one-carbon story, not an antioxidant one — which also explains why the effect requires a dose big enough to move methionine flux.
Metabolic and cardiovascular: genetics and one small trial
People with higher circulating glycine have less coronary heart disease and less type 2 diabetes. The obvious question is whether that is cause or consequence, and here the field did something better than another observational cohort. A genome-wide meta-analysis in 80,003 participants identified 27 loci for circulating glycine and used them as instruments. The result splits cleanly: genetically higher glycine was associated with lower coronary heart disease risk, partly mediated through blood pressure — but the genetic evidence for type 2 diabetes was weak, and the analysis instead found a strong inverse genetic effect of hyperinsulinaemia on glycine.[10] In plain terms: the diabetes association looks largely like insulin resistance lowering your glycine, not low glycine causing your diabetes. That is the single most useful correction on this page, because the reverse claim is repeated constantly in supplement copy.
Mendelian randomisation is still genetics, not a trial. It tests a lifelong small difference in glycine exposure, which is not the same intervention as a scoop of powder for six months. The direct supplementation evidence is one randomised trial: 60 people with metabolic syndrome, 30 given 15 g/day of glycine and 30 placebo for three months. Lipid peroxidation (TBARS) fell 25% versus placebo, erythrocyte superoxide dismutase activity fell 20% in step with SOD2 expression, and systolic blood pressure fell significantly — in the men only (p = 0.043).[12] Encouraging, entirely surrogate endpoints, one site, never replicated. Low circulating glycine is a consistent feature of obesity and metabolic disease, so the target is real; the treatment claim is not yet.[11]
What glycine has actually been tested for
| Use | Best human evidence | How much weight it carries |
|---|---|---|
| Sleep quality and latency | Small crossover trials, 3 g at bedtime; the one readable in full analysed 7 subjects, single-blind | Plausible and cheap — but one corporate lab, no independent replication |
| Next-day fatigue after short sleep | Same trial: VAS fatigue down, psychomotor vigilance improved | Real endpoints, tiny n, same source |
| Glutathione restoration in aging | Stable-isotope study, 8 elderly adults — synthesis fully restored | Strong result, but glycine was given with cysteine — that is GlyNAC |
| Collagen, joints, skin | None. Supporting data is cultured bovine chondrocytes | Best mechanism on the page, zero human outcomes |
| Metabolic syndrome markers | One 3-month RCT, 15 g/day, n = 60: TBARS −25%, systolic BP down in men | Single small trial, surrogate endpoints |
| Coronary heart disease | Mendelian randomisation, 80,003 people: higher glycine → lower CHD | Genetic evidence of causality, not a treatment trial |
| Type 2 diabetes | Same analysis: genetic evidence weak; insulin resistance appears to lower glycine | The association is largely backwards — don't buy the causal version |
| Lifespan | NIA ITP: 8% glycine diet, +4–6% median lifespan, both sexes, 3 sites | A real positive — in mice, at roughly 20× a human bedtime dose |
How to actually take it: dose, taste, timing, cost
For sleep: 3 g, taken shortly before bed. That is the dose used in every human sleep study, and there is no evidence more works better.[3][4] Timing matters more than it does for most supplements, because the proposed mechanism is a transient drop in core body temperature — taking it in the morning is taking it at the wrong end of the physiology.[5]
For metabolic or shortfall-filling use: 10–15 g/day, usually split. Fifteen grams daily is what the metabolic syndrome trial used for three months.[12] It is also roughly the amount the flux analysis says is missing.[1] GlyNAC protocols use weight-based doses of both amino acids together and are a separate regimen — see the GlyNAC page for those.
Taste is a genuine, underrated advantage. Glycine is sweet — it was named for it (Greek glykys, sweet) — at somewhere around three-quarters the sweetness of sucrose, with a clean finish and no bitterness. It dissolves readily in water or tea. That sounds trivial until you compare it to swallowing capsules of NAC, creatine or berberine every night; palatability is adherence, and adherence is the whole game for anything you take daily.
Cost is the strongest practical argument for trying it. Glycine is a bulk commodity amino acid made at industrial scale. A kilogram of plain powder typically sells for the price of a single month of a branded longevity supplement, which puts a 3 g nightly dose in the range of a few cents a day and even a 15 g/day protocol at well under half a dollar. Buy the plain powder. Capsules of glycine cost several times more per gram, and a 500 mg capsule means six capsules to reach a sleep dose and thirty to reach a metabolic one. Sleep-formula blends that list glycine alongside a dozen other ingredients at undisclosed doses are the worst value in the category.
Safety, and the two cautions worth knowing
Glycine is well tolerated, including at doses far above anything a supplement label suggests. In a double-blind crossover trial, 22 patients took 0.8 g/kg/day — roughly 56 g/day for a 70 kg adult — for six weeks, with clinical laboratory values and serum amino acids monitored, and the treatment was reported as well tolerated.[13] Fifteen grams a day for three months was likewise uneventful.[12] The realistic side effects at supplement doses are mild and gastrointestinal — nausea or loose stools if you take a large amount at once on an empty stomach.
- Clozapine — the one interaction that is not theoretical. In a 12-week double-blind trial, 19 patients with treatment-resistant schizophrenia on clozapine received either 30 g/day of glycine or placebo. Glycine did not help — and the patients on clozapine without glycine had a 35% reduction in positive symptoms that the glycine group did not get, leading the authors to conclude that glycine may interfere with the antipsychotic efficacy of clozapine.[14] That is a small trial, but it is a directionally alarming one, and glycine is an NMDA co-agonist, so the mechanism is plausible. If you take clozapine, do not add glycine without your prescriber’s involvement.
- Kidney stones — a real but small effect, and only for some people. Glycine is a precursor to oxalate through glyoxylate, and this has been quantified properly. Using labelled infusions in healthy adults, glycine catabolism at normal flux accounted for less than 5% of urinary oxalate. Pushed to a high infusion rate that raised whole-body glycine flux by 72%, its contribution rose to 16.0% ± 1.6%.[15] So a large daily dose measurably increases oxalate production, from a small base. That is not a general warning — it is a specific one for recurrent calcium-oxalate stone formers and anyone with primary hyperoxaluria, who should not take gram-scale glycine without their nephrologist or urologist weighing in.
- Kidney and liver disease generally. There is no evidence that supplemental glycine harms healthy kidneys, and we could not find a trial showing renal toxicity at any tested dose. But amino-acid loads are handled by the kidney and liver, and no glycine supplementation trial has enrolled people with significant chronic kidney disease. Absence of evidence, in that population, is not reassurance.
The honest bottom line
Glycine sits in an unusual place. Most cheap supplements have a weak mechanism and weak evidence; most interesting longevity compounds are expensive and unproven. Glycine has an excellent mechanism, a real if modest mouse lifespan result, genuinely encouraging human genetics for cardiovascular risk, and almost no proper human outcome trials at all — while costing pennies. The gap between how well-founded the biology is and how little anyone has bothered to test it in people is the story of this molecule, and it is exactly what you would expect for something nobody can patent.
A fair way to hold it: if you want help sleeping, 3 g at night is a cheap, palatable, well-tolerated experiment with a coherent mechanism and evidence you should discount heavily for its single source — you will know within two weeks whether it does anything for you. If you want glutathione, buy the combination instead, not glycine alone. If you want collagen or joint benefit, the mechanism is better than the marketing but the outcome evidence is empty. And if you are taking it for the ITP result, be clear-eyed that you are extrapolating from an 8% mouse diet, and weigh it against the rest of the field in our longevity evidence matrix.
This article is research information, not medical advice. Glycine is sold as a dietary supplement and is not approved by the FDA to treat, prevent or manage insomnia, osteoarthritis, metabolic syndrome, cardiovascular disease or aging. Do not add glycine to clozapine without your prescriber’s involvement — a controlled trial found it may reduce clozapine’s antipsychotic efficacy. If you form calcium-oxalate kidney stones, have primary hyperoxaluria, or have chronic kidney or liver disease, discuss gram-scale amino-acid supplementation with your clinician first; people with those conditions have not been enrolled in the trials described here. Persistent insomnia deserves a proper clinical evaluation rather than a supplement.