Eyerising International has responded to a recent JAMA Ophthalmology publication suggesting structural changes in the retina following repeated low-level red light (RLRL) treatment for myopia, highlighting several study limitations.

The study, ‘Cone density changes after repeated low-level red-light treatment in children with myopia’ by Liao et al.,1 used adaptive optics scanning laser ophthalmoscopy (AOSLO) to investigate the potential retinal effects of RLRL, with the authors reporting a statistically significant reduction in cone density within 0.5 mm of the foveal centre in children treated with RLRL compared to controls.

In response, Professor Mingguang He, Chief Medical Officer of Eyerising International and inventor of the RLRL technology in question, in an official statement and webinar, has warned against drawing firm conclusions from the data. Some of his concerns were echoed by the accompanying invited commentary published in JAMA Ophthalmology by Professor Jacque Duncan of the University of California, San Francisco.2

KEY METHODOLOGICAL CONCERNS

Prof He’s official statement identified multiple limitations in the JAMA study that may undermine its reliability.

Foremost among these is the study’s retrospective design, which lacked baseline cone density measurements. Without these, it is impossible to determine whether the differences observed predated the therapy itself. The reliance on parental recall to determine RLRL exposure further risks recall bias, and the small sample size and wide confidence intervals cast doubt on the robustness of the results.

From a statistical standpoint, Prof He noted that the inclusion of both eyes from participants without adjusting for intra-subject correlation may also have inflated statistical significance and is not standard practice, nor are the 32 statistical comparisons completed without any Bonferroni or similar correction to avoid false positives.

TECHNICAL AND CLINICAL LIMITATIONS

In terms of imaging, Prof He stated that AOSLO data acquisition is well known to be challenging, even more so in children, due to the long capture times (15–20 minutes) required, which need precise fixation to ensure accurate image quality. The study also did not include repeatability assessments to confirm reliability, he said.

Furthermore, the reported 6% reduction in cone density within 0.5 mm of the fovea was highlighted by Prof He as potentially unreliable, given that this region is densely packed with cone cells, making it difficult to delineate individual cone boundaries accurately and introducing potential measurement bias. Most studies assess cone density outside this central 0.5 mm zone, where cone boundaries can be more clearly visualised. A 6% reduction also falls within the range of normal measurement variability and is well below typically significant thresholds. For example, studies on retinal degenerative diseases have suggested that cone density must drop by over 35% before visual acuity is noticeably affected. Finally, these changes were not accompanied by any visual symptoms or functional vision testing, leaving their clinical relevance uncertain, Prof He said.

CONTRARY EMERGING EVIDENCE

Prof He then highlighted two ongoing independent studies using AOSLO – one randomised controlled trial (RCT) and one prospective cross-sectional study – in which preliminary findings oppose the conclusions of the JAMA study.

Prof He said an RCT conducted at Beijing Tongren Hospital in China of 336 children identified no statistically significant difference in cone density between RLRL and controls at 33 months. Separately, in a cross-sectional study at Chongqing Medical University, higher cone density was seen in the RLRL myopic group versus the untreated myopic group, although this difference did not reach statistical significance, with the highest cone density observed in the non-myopic group, he said.

“These results will soon be published and are expected to provide critical data to further clarify the safety profile of RLRL,” said Prof He.

A CALL FOR BETTER EVIDENCE

Eyerising International continues to advocate for high-quality research to determine the long-term safety and efficacy of RLRL, including these prospective studies underway. Meanwhile, Prof He stressed the importance of contextualising results such as those reported in this study, particularly for clinicians who are less familiar with AOSLO interpretation.

“We welcome rigorous scientific investigations,” said Prof He. “But it is important for clinicians and the public to understand that the small changes observed in this study do not necessarily indicate retinal damage or visual impairment, especially in light of its design limitations.”

Prof He’s full statement can be found at: bit. ly/45tCl3f. The Eyerising International webinar recording of Prof He and authors from both ongoing AOSLO studies can be found at: bit.ly/43Lv5OW.