The Eye as a Diagnostic Tool for Alzheimer’s Disease

“

TL;DR

This narrative review summarizes current evidence for using ocular tissues—including retina, vitreous, aqueous humor, lens, pupil, cornea, and tear fluid—as noninvasive biomarkers for Alzheimer's disease (AD). Ocular biomarkers such as retinal amyloid-β deposits, retinal nerve fiber layer thinning on OCT, and altered tear protein profiles show promise for early AD detection, but most findings require validation in larger prospective studies before clinical implementation. The eye's accessibility as a CNS extension makes it a compelling target for developing affordable, noninvasive screening tools that could address the current underdiagnosis of AD.

Full summary

5,068 CHARS

**Background:** Alzheimer's disease (AD) is a progressive neurodegenerative disorder and the most common cause of cognitive impairment among the elderly, accounting for 60–70% of all dementias. AD affects approximately 50 million people worldwide, a number projected to triple by 2050, and carries an annual cost exceeding one trillion dollars. The disease follows a continuum from preclinical AD (asymptomatic with biomarker evidence) through mild cognitive impairment (MCI) to AD dementia. Current gold-standard diagnostic biomarkers—PET imaging and CSF levels of Aβ and tau—are invasive, expensive, or not widely available, leading to routine underdiagnosis and late recognition. The eye, as an embryologic and anatomic extension of the central nervous system, offers a uniquely accessible window for noninvasive biomarker detection. This review examines the evidence for AD biomarkers across multiple ocular structures.

**Methods:** This is a narrative review summarizing published literature on ocular biomarkers in Alzheimer's disease. The authors discuss findings from histopathologic studies, in vivo imaging studies (OCT, OCTA, corneal confocal microscopy), proteomic analyses of ocular fluids, and clinical assessments of pupillary function and tear composition. No systematic search strategy or inclusion/exclusion criteria are reported.

**Key Results:** - *Retina and optic nerve:* Histopathologic studies report 36.4% overall neuronal loss in AD retinas compared to controls. Aβ retinal deposits are significantly higher in AD, with mid-peripheral levels exceeding central retina. OCT studies consistently demonstrate thinning of the peripapillary retinal nerve fiber layer (RNFL) and ganglion cell–inner plexiform layer (GC-IPL), along with reduced total macular volume in AD. A 2017 meta-analysis of 25 studies confirmed statistically significant reductions in mean peripapillary RNFL and macular thickness in MCI and AD. Choroidal thickness is also reduced in AD, with subfoveal choroidal thickness identified as a strong predictor. OCTA studies show a larger foveal avascular zone (FAZ) in Aβ-positive preclinical patients and reduced retinal capillary vessel density in AD. - *Vitreous humor:* Aβ40, Aβ42, and t-tau are detectable in vitreous, with low levels associated with poorer MMSE scores. Neurofilament light chain (NfL) in vitreous is significantly associated with increased Aβ40, Aβ42, and t-tau. - *Aqueous humor:* Aβ40 and Aβ42 are present at concentrations comparable to CSF. Animal studies show Aβ42 injected into CSF promptly migrates to aqueous humor. - *Lens:* The FLESS (fluorescent ligand eye-scanning) system demonstrated 95% specificity and 85% sensitivity for detecting AD. Supranuclear cataracts colocalising with Aβ immunoreactivity were reported in AD and Down syndrome patients. However, several subsequent studies failed to replicate findings of Aβ aggregates or supranuclear opacities in AD lenses. - *Pupil:* Increased pupillary dilation during cognitive tasks is observed in AD and MCI. Higher genetic risk individuals show significantly greater pupil dilation during high cognitive demand. Baseline pupil size findings are inconsistent across studies. - *Cornea:* Corneal confocal microscopy reveals progressive decreases in corneal nerve fiber density (CNFD), branch density (CNBD), and fiber length (CNFL) in MCI and dementia. Mean corneal sensitivity is significantly lower in AD. APP and Aβ degradation enzymes are expressed in corneal fibroblasts and epithelium. - *Tear fluid:* Tear flow rate increased from 6±2 µL/min in controls to 12±2 µL/min in AD; tear protein concentration increased from 4.4±1.4 µg/µL to 8.8±2.9 µg/µL. A four-protein panel (lipocalin 1, dermcidin, lysozyme C, lacritin) showed 81% sensitivity and 77% specificity for AD. Tear t-tau is elevated in neurodegeneration and negatively correlated with CSF Aβ42. MicroRNA-200b-5p is increased in AD tear samples.

**Clinical Implications:** Ocular biomarkers offer the potential for noninvasive, cost-effective, and accessible screening for AD, addressing a critical gap where current diagnostic tools are too invasive or expensive for population-level screening. Retinal imaging (OCT/OCTA) is particularly promising due to its established clinical infrastructure and ability to detect structural changes that may precede cognitive symptoms by years. However, significant limitations remain: high-quality imaging is difficult in advanced dementia; coexisting ocular diseases (AMD, glaucoma, diabetic retinopathy) common in the elderly may confound results; tear fluid collection yields limited sample volumes that diminish with age; and vitreous/aqueous sampling remains invasive. Crucially, several key findings—particularly lens Aβ aggregates and supranuclear cataracts—have not been consistently replicated. The field requires large-scale, longitudinal validation studies with standardized protocols, defined cutoff values, and assessment of preclinical AD populations before any ocular biomarker can be incorporated into clinical practice.