General characteristics of the study population.

Background
Biomarkers are widely recognized as crucial breakthroughs in tackling amyotrophic lateral sclerosis (ALS). Among them, retina markers may hold promise due to the close retina-brain connection and non-invasive, portable detection methods. Thus, using optical coherence tomography (OCT), we investigated the link between baseline cell-level retinal features and future ALS risk.
Methods and findings
Participants from the UK Biobank underwent OCT scans to assess retinal layers, macula, and optic disc parameters. Follow-up commenced two years after the baseline period (2006–2010), during which ALS cases were identified using International Classification of Diseases (ICD) codes from medical and assessment records. Cox proportional hazards models were applied to examine the relationship between retinal markers and incident ALS. Over a median follow-up of 14.11 years, 70 ALS cases occurred among 53,824 participants (incidence 10.58 per 100,000 person-years). Most participants were White (94.6%), 44.8% male, with a median age of 58 years. After adjusting for demographics and comorbidities affecting the retina, a standard deviation (SD) decrease of 15.19 µm in photoreceptor layer (PRL) thickness was associated with a 19% (95% confidence interval [7, 29]; p = 0.002) increased risk of ALS, while a SD increase of 26.11 µm in retinal pigment epithelium (RPE) thickness corresponded to a 20% (95% CI [7, 34]; p = 0.002) higher risk. Sensitivity analyses excluding follow-ups of less than 4 and 6 years yielded consistent results. Subgroup analyses showed these findings were more pronounced in smokers. The main limitation of this study is its single time point observational design.
Conclusion
A thinner PRL and thicker RPE may precede the clinical diagnosis of ALS, offering potential clues for early diagnosis and insights into the disease’s pathogenesis.