Glycine for Sleep Quality
A handful of small trials hint that 3 g of glycine before bed may modestly improve subjective sleep quality and next-day alertness, but the evidence is too thin and conflicted to call it a proven sleep aid. No major health authority or clinic endorses it for sleep.
Why this grade7-layer evidence engine
The grade is weak (Tier C) because the entire human case rests on a small cluster of low-quality trials. In randomized studies (PMID 22293292), 3 g of glycine taken about an hour before bed improved subjective sleep quality, shortened the time to reach deep slow-wave sleep, and eased daytime sleepiness, and a double-blind trial (PMID 22529837) found less fatigue and better performance after a short night. But all of this comes from a single Japanese research group, the samples are tiny (n=11 to ~19), outcomes are mostly questionnaires, and every study was industry-funded (Ajinomoto, a glycine maker), a clear conflict of interest.
Crucially, the findings have never been replicated by a large independent trial or pooled meta-analysis, so the positive signal cannot yet be trusted as a reliable effect.
The institutional picture is unsupportive. EU regulators (EFSA) concluded that a cause-and-effect relationship has not been established and approved no sleep claim for glycine. Major clinical sources, including NIH's Office of Dietary Supplements, Mayo Clinic, Cleveland Clinic, and Harvard Health, do not list glycine as a sleep treatment, and the leading sleep-medicine guideline does not recommend it. Only the Sleep Foundation comments, cautiously noting the three small studies 'may hold promise' while flagging their limits. Bottom line: plausible and possibly helpful, but unproven, not harmful at typical doses, and not a substitute for first-line measures like CBT-I.
Scoring transparency
All scores computed by a 7-layer evidence engine — fully auditable▸View the full decision path (audit trail)
- compute_raw_score — 加權公式: L2×0.30 + L3×0.25 + L5×0.25 + L11×0.10 + L1×0.10 = 0.445
- tier_from_score — 依分數區間映射至 tier letter
- apply_hec_rules — 無高階證據可裁決
- tier_strict_requirement_check — Tier 條件達標，未降階
- detect_disputes — 偵測到 0 個 hard + 1 個 soft dispute
- decide_status — 依 tier + dispute 結果決定 status