Interventions to reduce short-wavelength ("blue") light exposure at night and their effects on sleep: a systematic review and meta-analysis
Systematic review and meta-analysis (PROSPERO CRD42018105854) of studies using color-tinted lenses (e.g. orange or amber) to filter short-wavelength light before sleep: pooled effects were small-to-medium for objective sleep efficiency and total sleep time, large for PSQI in a small subset of self-report studies, with inconsistent individual trials overall (k = 12).
Scope
Intervention: spectral filters worn in frames (e.g. orange / amber lenses) intended to reduce short-wavelength (“blue”) retinal exposure in the pre-sleep interval. Outcomes: self-report and objective (e.g. actigraphy) nocturnal sleep measures.
Meta-analytic estimates (authors’ random-effects models; read full text for study tables)
- Sleep efficiency (objective): Hedges’ g = 0.31 (95% CI −0.05 to 0.66); k = 7; I² ≈ 38%
- Total sleep time (objective): g = 0.32 (95% CI 0.01 to 0.63); k = 6; I² ≈ 12%
- Pittsburgh Sleep Quality Index (self-report): g = −1.25 (95% CI −2.39 to −0.11); k = 3; I² ≈ 36% — direction interpreted as better global sleep quality vs control in pooled self-report direction
- Total sleep time (self-report): g = 0.51 (95% CI 0.18 to 0.84); k = 3; I² = 0%
Interpretation
Abstract stresses inconsistent single-study directions despite some pooled signals; authors highlight populations where benefits are more often discussed (insomnia, bipolar disorder, delayed sleep phase, ADHD) while calling for more controlled trials given ubiquitous evening short-wavelength–enriched light sources.
Evidence hygiene
- Not interchangeable with room-only dimming RCTs or bright morning light phase-response work—this row is ocular short-wavelength reduction before bed.
- Pair with Figueiro et al. 2019 (
figueiro-2019-evening-light-sleep) for older-adult amber-room / lens trial context versus this broader pooled lens.
Outcomes
- Effect Size (Cohen's d / SMD)Pooled Hedges' g (short-wavelength–reducing lenses before sleep): objective sleep efficiency g=0.31 (95% CI −0.05 to 0.66; k=7; I²≈38%); objective total sleep time g=0.32 (0.01–0.63; k=6); PSQI self-report g=−1.25 (−2.39 to −0.11; k=3); self-reported total sleep time g=0.51 (0.18–0.84; k=3). Individual studies inconsistent overall (k=12 per abstract).