Lavender for Sleep Quality
Inhaled lavender shows a consistent but small-to-moderate signal for better self-reported sleep, yet the trials are mostly small and unblindable, so placebo and expectation effects cannot be ruled out. The one rigorous oral trial found nearly all of lavender's sleep benefit came indirectly from easing anxiety rather than from any direct sedative effect.
Why this grade7-layer evidence engine
The grade is Preliminary because the human data point in a positive direction but are methodologically weak. Two meta-analyses found lavender improved sleep on validated scales: a 2026 review of 11 RCTs (PMID 40600743, n=628) reported a small-to-moderate effect (SMD -0.56, p=0.005), and a 2024 meta-analysis in older adults (PMID 39654196) found a stronger lavender-only subgroup effect (SMD -1.39). A small open-label RCT in stroke patients (PMID 38241803, n=40) also favored lavender.
The central problem is blinding. Lavender's scent cannot be masked, and most aromatherapy trials are small, single-blind, or open-label, so gains on subjective questionnaires are vulnerable to expectation and placebo effects, and heterogeneity between studies is high. Tellingly, the only double-blind oral trial (Silexan 80 mg; PMID 31121394, n=212) found 98.4% of the sleep benefit was mediated by lavender's anti-anxiety action rather than direct sedation. That trial was industry-funded and treated sleep as a secondary endpoint.
Authorities are cautious. The US FDA recognizes lavender oil only as a GRAS flavoring agent, not a treatment, and the American Academy of Sleep Medicine does not endorse aromatherapy. Cleveland Clinic notes the calming scent may help and cites a 2019 study suggesting it can raise melatonin, while the Sleep Foundation calls it a non-standard, possibly short-lived aid. Lavender also carries an additive CNS-depression risk when combined with sedatives, benzodiazepines, or alcohol.
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 = 0.569
- tier_from_score — 依分數區間映射至 tier letter
- apply_hec_rules — 高品質 SR/MA 顯示 positive (2 篇 > 0 negative)
- tier_strict_requirement_check — Tier 條件達標，未降階
- detect_disputes — 偵測到 0 個 hard + 0 個 soft dispute
- decide_status — 依 tier + dispute 結果決定 status