**Background:** Dementia affects 46.8 million people globally, with Alzheimer's disease (AD) being the most common form (60%–80% of cases). Current diagnostic methods like MRI are costly and limited, while body fluid biomarkers lack diagnostic specificity. The retina, as an extension of the central nervous system, offers a rapid, low-cost, and non-invasive alternative via optical coherence tomography (OCT), which provides high-resolution images of retinal layers and vasculature. Previous studies have shown associations between retinal measures and cognitive impairment, but large-scale longitudinal data are needed to establish causal relationships. The UK Biobank, a prospective cohort of over 500,000 participants aged 40–69 recruited between 2006 and 2010, provides a unique resource with extensive genetic, lifestyle, and health data. This paper outlines the rationale and methods for adding eye measures to the UK Biobank Repeat Imaging study to investigate dementia.
**Methods:** The UK Biobank Repeat Imaging study will invite up to 60,000 participants who previously underwent baseline imaging (brain, heart, abdomen MRI, DEXA, carotid ultrasound) to undergo repeat multimodal imaging between 2022 and 2028. Eye measures, including OCT and color fundus photography, will be collected using the Topcon Triton swept-source OCT platform (central wavelength 1050 nm), which enables visualization of the choroid and vasculature. The protocol includes wide-angle 12 mm × 9 mm scans covering the optic disc and macula. Candidate biomarkers include total macular retinal thickness, macular inner retinal sublayer thicknesses (RNFL, GC-IPL, GCC), peripapillary RNFL thickness, choroidal vascularity index, retinal arteriolar and venular calibers, retinal vascular fractal dimension, and retinal vascular tortuosity. Image processing will use FDA-approved algorithms for layer segmentation and the QUARTZ system for automated vessel segmentation and quantification (with 87% accuracy for arteriole/venule classification). The VAMPIRE system will provide additional vasculometry summaries, including fractal analyses. Recruitment will occur at four dedicated imaging centers across the UK (Newcastle, Stockport, Reading, Bristol) with standardized equipment and protocols to minimize variability. Participants will be invited via email/postal invitation and telephone booking, with an expected acceptance rate of ~60% based on pilot studies. The study will also collect cognitive function tests (e.g., fluid intelligence, pairs matching, reaction time) and leverage existing data on APOE genotypes, comorbidities, and medication use.
**Key Results:** Baseline data from 48,998 participants attending the initial imaging assessment (mean age 55.2 years, 52% female) show that 13,732 (28%) had retinal imaging. The repeat imaging study will oversample those with baseline retinal imaging, aiming for >16,800 participants with overlapping data. Previous UK Biobank analyses demonstrated that individuals in the lowest quintile of RNFL thickness were 11% (95% CI 2% to 21%) more likely to fail at least one of four cognitive tests. Additionally, macular ganglion cell-inner plexiform layer thickness, ganglion cell complex thickness, and total macular thickness were significantly associated with smaller total brain volume (p<0.001), grey matter volume (p<0.01), and hippocampal volume (p<0.02). The study is powered to detect a 0.03 SD change in cognitive score or brain measures per 1 SD increase in retinal biomarker with 99% power (alpha=0.001). With an annual dementia incidence of ~2.5 per 1000 person-years in those aged ≥60, approximately 250 dementia cases are expected within 2 years of retinal imaging among 45,000–55,000 participants with good-quality images.
**Clinical Implications:** The addition of OCT and retinal imaging to the UK Biobank Repeat Imaging study will create a major resource for dementia diagnostic discovery. Retinal biomarkers may enable early identification of individuals at high risk for cognitive decline and dementia, facilitating lifestyle interventions or enrollment in drug trials. The longitudinal nature of the data will allow assessment of causal relationships between retinal changes and cognitive outcomes, potentially leading to non-invasive, low-cost screening tools. The open-access policy will promote global research into ophthalmic diseases and other health outcomes beyond dementia.