Glycine improves sleep quality

Glycine, an amino acid, has been reported to help sleep in a few small trials at about 3 grams taken before bed, but there is little research and almost all of it comes from one research group. A 2026 review that searched the medical literature found three trials of glycine itself with sleep outcomes, each reporting a benefit on at least one measure. All three were small, all were in adults, and all came from researchers connected to a company that makes amino acids. The one sleep-laboratory test by an unconnected group gave glycine inside a collagen supplement and did not improve either of its two main sleep measures: how quickly people fell asleep and how much of their time in bed they spent asleep.

In the trial that measured sleep with sensors (polysomnography), 11 adults with ongoing poor sleep took 3 grams of glycine or a placebo before bed in a single-blind crossover. On glycine they reached light sleep and deep slow-wave sleep sooner, the share of time in each sleep stage did not change, and they rated their sleep better the next morning. Daytime sleepiness was lower at one morning time point on one of two scales, and a memory task was better at the noon test only. A second trial, in 7 healthy men whose time in bed was cut by a quarter for three nights, found lower fatigue ratings than placebo on the first day but not the third and quicker reactions on a vigilance test, while sleepiness ratings did not differ significantly. A 2024 systematic review counts these same studies rather than independent ones, and it rates them as small and at high risk of bias.

These trials are one research line, not several groups arriving at the same answer. The two glycine trials cited here and the animal study behind the proposed mechanism share an author, several of their authors worked for an amino-acid manufacturer, and the sleep-restriction trial was run in that company's own employees. The review's third glycine trial is an earlier trial by the same group, described in a 2025 review as a placebo-controlled crossover in 15 women with sleep complaints that found better next-morning fatigue ratings. The 2026 review that counted these trials was funded by Nestlé Health Science, which sells collagen supplements.

The one sleep-laboratory test from an unconnected group gave 15 grams of glycine-rich collagen peptides or a placebo to 13 athletic young men with sleep complaints each night for a week, in a double-blind crossover. Its two main measures, time to fall asleep and sleep efficiency on the sensors, did not change. It did record fewer night-time awakenings and better accuracy on one cognitive test, both secondary measures. The 2026 review found one other placebo-controlled collagen trial, which reported lower fatigue scores, and several comparisons in which collagen did no better for sleep than a whey protein.

Every trial here was in adults: healthy volunteers or people reporting poor sleep. None included children, teenagers, pregnant or breastfeeding women, or people with kidney or liver disease, and the one trial that lists its exclusions left out people taking sleep medication or with a diagnosed sleep disorder. No trial has tested glycine at the sleep dose together with any medication. At far higher doses, glycine has been added to the antipsychotic clozapine in small trials in schizophrenia. At 30 grams a day, people taking glycine did not show the improvement in symptoms seen with clozapine alone, and the authors suggested glycine may interfere with clozapine. At 60 grams a day, glycine added no benefit and no one's symptoms worsened.

The proposed mechanism comes from rats. In rats, glycine acted on NMDA receptors in the brain's master clock, the suprachiasmatic nucleus, increasing blood flow to the skin and lowering core body temperature, and a falling core temperature is one of the body's signals for sleep. That route has not been measured with glycine itself in people, and the collagen trial, which measured core temperature, found no difference from placebo. The glycine trials point in the same direction, but they are few, small and from one group, and the one independent sleep-laboratory check did not reproduce the main result.