Alzheimer disease (AD) is a primary progressive neurodegenerative disease. It is characterized by progressive deterioration of memory and executive function, impairing activities of daily living. It affects approximately 4.7% of people in their 60s and accounts for more than 50% of all dementia cases. It is estimated to affect about 5.5 million people in the United States and up to 35 million worldwide, with an incidence rate of 11 per 1,000 person-years.
AD progresses as a continuum from normal cognitive function (preclinical AD) to mild cognitive impairment (MCI) to AD dementia 1). Although there is currently no cure, treatments exist that can slow the progression of cognitive decline (with limited effectiveness).
AD affects many structures of the eye. Pathological changes have been reported in the retina, optic nerve, lens, tear fluid, cornea, pupil, and choroid. The retina is embryologically derived from the diencephalon and connects to the brain via the optic nerve. Due to the structural and functional similarities between the blood-retinal barrier and the blood-brain barrier, as well as the commonality of the neurovascular unit (NVU), the retina is considered a “window to the brain” 1). It is noted for allowing observation of CNS changes in a low-cost, non-invasive manner compared to PET or CSF testing 1).
QWhy does Alzheimer's disease cause changes in the eye?
A
The retina is embryologically part of the brain (diencephalon), and the blood-retinal barrier and blood-brain barrier share structural and functional similarities. Since the pathology of AD (amyloid β deposition, tau pathology, neuroinflammation) extends to the retina as it does to the brain, various ocular changes occur1).
Clinical findings of AD are distributed across various structures of the eye.
Retinal Findings
RNFL thinning: Meta-analysis shows that peripapillary RNFL is significantly thinner in AD dementia patients compared to controls (SMD=−0.67) 1).
GC-IPL thinning: SMD=−0.46. AUROC may be higher than that of RNFL (0.685 vs 0.601) 1).
RGC reduction: Reduction of retinal ganglion cells and degeneration of melanopsin RGCs have been reported in postmortem studies 1).
Retinal Aβ and tau deposition: Deposition of tau, amyloid beta (Aβ), and phosphorylated tau in the retina has been reported.
Optic nerve and other
Optic nerve thinning: Axonal degeneration and loss lead to reduced optic nerve thickness and optic disc pallor.
Lens changes: An association between AD and supranuclear cataract has been reported. Aβ deposition is observed in the lens and may precede brain MRI findings and clinical symptoms by up to 10 years.
Tears, cornea, and pupil: Increased tear flow and protein levels, decreased corneal sensitivity, and reduced pupillary light reflex amplitude have been reported.
Choroid: A decrease in choroidal thickness is observed.
Retinal vascular findings (detected by OCTA and dynamic vessel analysis):
Decreased vascular density: Reduced retinal vascular density and enlargement of the foveal avascular zone (FAZ)1)
Delayed vascular response: Time to reach 30% of maximum arterial dilation in response to flicker stimulation is prolonged in AD (7.0 seconds vs. 5.0 seconds in controls, AUROC 0.853)1)
Reduced fractal dimension: Arterial 1.201 vs. 1.235 (p=0.008), venous 1.171 vs. 1.210 (p<0.001)1)
Increased retinal oxygen saturation: Moderate AD arterial 94.2%±5.4% vs. healthy 90.5%±3.1% (p=0.028)1)
This is a subtype of AD in which visual symptoms are the initial manifestation due to focal atrophy of the parieto-occipital lobes. Homonymous hemianopia or cortical visual impairment with negative structural imaging or only posterior cortical atrophy may suggest AD.
QWhat is visual variant Alzheimer disease (VVAD)?
A
VVAD, also called posterior cortical atrophy (PCA), is a subtype of AD in which visual symptoms are the initial manifestation due to focal atrophy of the parieto-occipital lobes. Reading difficulty, visuospatial agnosia, and problems with visual information processing are the main symptoms, and visual symptoms are prominent from the early stage, even when memory impairment is not obvious.
The main etiology of AD is based on the amyloid hypothesis. Imbalance between Aβ production and clearance leads to Aβ deposition and aggregation, and abnormal aggregation of tau protein results in the formation of neurofibrillary tangles (NFTs). The number of NFTs correlates with disease severity. Misfolded proteins induce oxidative stress and inflammatory damage, leading to neuronal loss and brain atrophy through impaired synaptic and neuronal activity.
The association with vascular disease is also important. Vascular risk factors are associated with increased brain Aβ burden, and the coexistence of cerebrovascular disease and Aβ accelerates cognitive decline and neurodegeneration 1). Autopsy studies have shown that microvascular pathology is present even in AD patients without clinical evidence of mixed dementia 1).
The main risk factors and protective factors are shown below.
Category
Main factors
Modifiable risk factors
Type 2 diabetes, dyslipidemia, obesity, cardiovascular disease, smoking, anticholinergic drug use
Protective factors
Higher education level, bilingualism, social interaction, marriage, physical activity
Approximately one-third of AD cases worldwide are attributed to modifiable risk factors. Additionally, the existence of individuals who maintained cognitive function despite significant NFT and Aβ burden suggests that genetic susceptibility acts as a risk modifier.
The diagnosis of AD is primarily made clinically. It combines cognitive function assessment, physical examination, and brain imaging (MRI, CT, PET) for confirmation. Definitive diagnosis is only possible through postmortem histological examination to visualize NFTs and Aβ, in addition to clinical symptoms and signs. AD is diagnosed only after excluding contributions from alcohol use history, trauma, and vascular disorders.
Ophthalmic diagnostic methods are described below. Although ocular findings in AD are being studied as potential auxiliary biomarkers, they are not yet established diagnostic methods for AD at present.
Adaptive optics SLO: Directly visualizes nerve fiber bundles with ultra-high resolution of approximately 2 μm. Patients with MCI have significantly more hyperreflective granular deposits on the retinal nerve fiber layer (possibly reflecting inner retinal gliosis)1).
Dynamic Vessel Analysis (DVA): Evaluates retinal vessel response to flicker light stimulation. Arterial dilation in the AD dementia group was significantly lower than in the control and MCI groups (0.77% vs 3.53% vs 2.84%), AUROC 0.8531).
Retinal Oximetry: Noninvasive measurement using two wavelengths (570 nm and 600 nm). Arterial and venous oxygen saturation are increased in AD and MCI1).
FLIO (Fluorescence Lifetime Imaging Ophthalmoscopy): Preclinical AD patients show longer fluorescence lifetime than controls (593.9±93.3 vs 454.4±38.6 ps, p=0.036). Correlates with GC-IPL thickness and CSF Aβ/tau levels1).
Biomarker detection:
Retinal amyloid plaque imaging: SLO using curcumin fluorescent dye. In AD patients, fluorescence intensity increased 2.1-fold compared to baseline. Aβ deposits are densely clustered in the superotemporal quadrant and distributed along blood vessels.
Hyperspectral retinal imaging: Identifies biomolecules across 225 continuous spectral bands. Machine learning models can distinguish Aβ-positive from Aβ-negative cases1).
Tear biomarkers: A combination of lipocalin-1, dermcidin, lysozyme C, and lactritin has been reported to have 81% sensitivity and 77% specificity.
Peripheral retinal drusen: AD patients have more drusen deposits in the peripheral retina (25.4% vs 4.2%, p=0.04), particularly in the superonasal quadrant1).
Eye movements: Delayed saccades and impaired smooth pursuit eye movements have been reported (NCT01434940).
Diagnostic limitations:
RNFL thinning is not specific to AD and can also be seen in Parkinson’s disease and dementia with Lewy bodies1).
It is difficult to differentiate from glaucoma, and glaucoma also shows inferior and superior RNFL thinning patterns 1).
OCT and OCTA can detect AD-related retinal changes, but their disease specificity is low, and they cannot be used for definitive diagnosis of AD at present. Since similar findings occur in glaucoma and other neurodegenerative diseases, ophthalmic findings remain at the research stage as auxiliary biomarkers 1).
There is currently no curative treatment for AD. Treatments aimed at slowing the progression of cognitive decline exist, but their effects are limited. AD is not a disease to be treated solely by ophthalmology; interdisciplinary collaboration with neurology and psychiatry is essential.
Management of ocular symptoms in AD:
Vision correction: Appropriate refractive correction and provision of optical aids
Environmental adjustments: Consideration of low-luminance environments, lighting adjustments according to color vision and contrast sensitivity decline
Rehabilitation referral: Early referral to visual rehabilitation and counseling is important for patients with VVAD/PCA
Regular ophthalmology follow-up: Management of comorbid eye diseases such as cataracts and glaucoma
The retinal pathology of AD primarily involves Aβ deposition and tau aggregation.
Aβ plaques in the retina have been detected in histological studies, most frequently in the GC-IPL, with a tendency to cluster around blood vessels. Retinal Aβ accumulation may occur earlier than in the brain 1). Aβ deposits inside and outside degenerating melanopsin RGCs, paralleling cell loss in the RGC layer, inner nuclear layer, and outer nuclear layer. Cell lines and animal models have confirmed the toxicity of Aβ to retinal neurons 1).
In transgenic mice, tau aggregates have been detected in the retina. Abnormal neurotrophic factor signaling, increased excitotoxicity susceptibility, early axonal damage, and RGC dysfunction have been reported 1).
However, the presence of AD pathology in the retina is inconsistent across studies. Some studies have not detected Aβ or fibrillar tau. Differences in tissue processing methods and immunohistochemistry protocols are considered a cause of this inconsistency 1).
There are two hypotheses regarding the mechanism of retinal thinning1).
Retrograde degeneration hypothesis: Retrograde degeneration of the visual pathway due to brain lesions secondarily thins the optic nerve and retinal layers.
Common pathology hypothesis: The idea that AD pathology (Aβ plaques, tau, neuroinflammation) occurs simultaneously in the brain and retina.
Some studies have reported RNFL thickening in early AD, suggesting that inner retinal reactive gliosis (inflammatory response) may precede thinning1). It has also been suggested that RNFL thinning may precede GC-IPL thinning as the temporal sequence of the neurodegenerative process1).
Cerebral small vessel disease contributes to AD pathology. A decrease in retinal vascular fractal dimension is thought to reflect a deviation from optimal cerebral microcirculation 1). Narrowing of the central retinal vein diameter (CRVE) may be due to venous wall thickening from collagen deposition 1). Increased retinal oxygen saturation is thought to reflect hypometabolism associated with AD 1).
QDoes amyloid beta also accumulate in the retina?
A
Histological studies have reported cases of Aβ deposition in the retina (especially in the GC-IPL) 1). However, consistency across studies is low, possibly due to differences in tissue processing methods. Retinal Aβ accumulation has been suggested to occur earlier than brain pathology, but this is not an established finding.
7. Latest Research and Future Prospects (Research Stage Reports)
Efforts are underway to use deep learning (DL) to recognize Alzheimer’s disease-specific “retinal fingerprints.” Research goals include applying routine retinal images from eye clinics to large-scale AD screening and building a physician support system through two-step risk stratification1).
Since retinal imaging alone has limited disease specificity, combining it with blood biomarkers (a “multiple marker approach”) is being investigated to improve diagnostic accuracy1).
The clinical trial of lens Aβ detection using aftobetin hydrochloride (NCT02928211) evaluates a method for detecting Aβ deposition in the lens in vivo. Lens amyloid pathology may precede clinical symptoms by up to 10 years, drawing attention as a potential ultra-early biomarker. A trial of AD diagnosis using eye movement tracking (NCT01434940) is also ongoing.
The ENVIS-ion study, which evaluates the effect of low-dose aspirin in suppressing progression of white matter lesions and silent cerebral infarction, is examining retinal vascular changes as a treatment outcome 1). This is an attempt to explore whether retinal vascular parameters can be used as therapeutic evaluation indicators for cerebrovascular disease.
Cheung CY, Mok V, Foster PJ, Trucco E, Chen C, Wong TY. Retinal imaging in Alzheimer’s disease. J Neurol Neurosurg Psychiatry. 2021;92(9):983–994.
Javaid FZ, Brenton J, Guo L, Cordeiro MF. Visual and Ocular Manifestations of Alzheimer’s Disease and Their Use as Biomarkers for Diagnosis and Progression. Front Neurol. 2016;7:55. PMID: 27148157.
Heaton GR, Davis BM, Turner LA, Cordeiro MF. Ocular biomarkers of Alzheimer’s disease. Cent Nerv Syst Agents Med Chem. 2015;15(2):117-25. PMID: 25788142.
Copy the article text and paste it into your preferred AI assistant.
Article copied to clipboard
Open an AI assistant below and paste the copied text into the chat box.
