Repeated low-level red light (RLRL) therapy is the fastest-rising emerging modality in myopia management. Strong short-term RCT results, notable rebound on withdrawal, and ongoing safety questions — a balanced evidence review.
Repeated low-level red light (RLRL) therapy is a photobiomodulation intervention delivered via a desktop device that emits red light at 650nm wavelength. In the published RCT protocols, it is administered at home for 3 minutes per session, twice daily with a minimum 4-hour interval between sessions, typically 5–7 days per week. The devices used in clinical trials emit at approximately 1,600 lux and 0.29mW for a 4mm pupil — classified as Class I (lowest risk) laser devices.
The proposed mechanism involves light-induced stimulation of cytochrome c oxidase in mitochondria, enhancing cellular metabolism in retinal and choroidal tissues. Choroidal thickening has been consistently observed after RLRL treatment, and this choroidal response correlates with treatment efficacy — though researchers note the degree of choroidal thickening alone does not fully explain the observed axial length changes.
All published major RCTs have been conducted in China. The technology was developed and is regulated in China. It is not currently approved for clinical use in the USA, UK, Europe, or Australia, though devices are available commercially in some markets.
Jiang et al. 2022 (Ophthalmology) — The first multicenter RCT, 264 children aged 8–13 with −1.00D to −5.00D myopia, randomised to RLRL plus single-vision spectacles vs spectacles alone. At 12 months: mean axial elongation 0.13mm (95% CI 0.09–0.17mm) in the RLRL group vs 0.38mm (95% CI 0.34–0.42mm) in the control group. Spherical equivalent progression was −0.20D vs −0.79D. No severe adverse events (sudden vision loss ≥2 lines or scotoma) were observed.
Xiong et al. 2022 (Clin Exp Ophthalmol) — 2-year follow-up of the original cohort. Children who continued RLRL (RLRL-RLRL group) showed 0.16mm total axial elongation over 2 years. Children who stopped RLRL after year 1 and switched to spectacles (RLRL-SVS group) showed 0.42mm axial elongation in year 2 alone — significantly more than the 0.28mm in the ongoing control group, demonstrating a statistically significant rebound effect.
Xu et al. 2024 (Ophthalmology) — First RCT specifically in high myopia (≥−4.00D, n=192, ages 6–16). RLRL group showed a mean AL change of −0.06mm (i.e., actual axial shortening) vs +0.34mm in the control group at 12 months. 53.3% of treated eyes showed axial shortening exceeding 0.05mm. No adverse events reported in this trial.
Xiong et al. 2024 (Ophthalmology) — Multicenter RCT of RLRL combined with Ortho-K in children who were fast progressors on Ortho-K (≥0.50mm/yr). 48 children randomised 2:1 to combination vs Ortho-K alone. At 12 months: combination group −0.02mm axial change vs +0.27mm in the Ortho-K-only group, an adjusted difference of −0.29mm.
The IMI 2025 Interventions paper (Bullimore et al.) specifically noted that 5 of the 6 highest rebound values (≥0.14mm axial length increase after discontinuation) across all myopia control modalities were reported in atropine and red-light therapy studies — not in optical modalities like Ortho-K or MiSight. This is a clinically significant finding. If RLRL therapy is started, clinicians should plan for a structured withdrawal protocol and monitor closely for 6 months post-cessation.
Across all published RCTs, no severe adverse events (defined as sudden vision loss ≥2 lines or development of a scotoma) have been reported. Structural assessments including OCT have not shown damage to the macula, retinal nerve fiber layer, or photoreceptor layer in any published study.
However, the IMI 2025 Interventions paper notes that the evidence base for RLRL safety is limited to short-duration trials, primarily 12 months. Long-term retinal safety data for 5–10+ years of exposure — particularly during critical periods of retinal development in young children — does not yet exist. The IMI does not endorse RLRL for widespread clinical use outside of clinical trial settings in Western markets.
The Cochrane 2023 living systematic review (Lawrenson et al.) also notes the methodological limitations of existing RLRL trials, including variable compliance monitoring, the absence of placebo-controlled blinding in most studies, and the predominance of Chinese paediatric populations, limiting generalisability.
RLRL devices are not approved for myopia control in the United States, European Union, UK, or Australia. Clinicians in these markets who encounter patients using RLRL devices purchased online should be aware of the evidence gaps and rebound risk on cessation. Monitoring axial length closely in such patients is appropriate.
| Modality | 12-month AL benefit | Rebound on stop | Approval status |
|---|---|---|---|
| RLRL therapy | ~0.25–0.26mm vs control | Significant (Xiong 2022) | Not approved (West); China-developed |
| MiSight 1 day | ~0.10mm/yr (6-yr data) | Not significant (Chamberlain 2022) | FDA-approved (USA, ages 8–12) |
| HAL/Stellest | ~0.17mm/yr (2-yr data) | Not reported | CE-marked (Europe); not approved US |
| Ortho-K | ~0.17mm (1-yr median) | Modest (MD 0.10mm) | Off-label (most markets) |
| Atropine 0.05% | Strong efficacy | Higher (LAMP Phase 3) | Off-label (most Western markets) |