Ангиография сетчатых микроваскулярных биомаркеров, измеряемая оптической согласованностью в нейродегенеративных и нейроаскальных расстройствах: систематический обзор и мета-анализ.

Introduction

Neurodegenerative and neurovascular disorders, including Alzheimer’s disease (AD), mild cognitive impairment (MCI), and cerebral small vessel disease (CSVD), represent a growing global health challenge and are associated with substantial personal, societal, and economic burdens. Early identification of these conditions remains difficult because pathological changes often begin years before the onset of clinically apparent symptoms. Detecting disease-related alterations during these early stages may facilitate timely intervention and potentially slow disease progression. However, currently available diagnostic approaches, such as positron emission tomography (PET) and cerebrospinal fluid (CSF) biomarkers, are costly, invasive, and not readily applicable for large-scale population screening. 1 , 2

The retina has emerged as a promising target for the investigation of neurological disease because of its unique developmental and anatomical relationship with the central nervous system (CNS). As an extension of the CNS, the retina shares structural, cellular, and microvascular characteristics with the brain, allowing retinal changes to potentially reflect pathological processes occurring within cerebral tissue. Consequently, the retina has frequently been described as a “window to the brain”, providing a unique opportunity to evaluate neurovascular alterations using non-invasive imaging techniques. 3

Recent advances in retinal imaging have expanded the ability to assess both neural and vascular structures in vivo. Optical coherence tomography (OCT) enables high-resolution visualization of retinal architecture and quantification of retinal nerve fiber layer (RNFL) and ganglion cell complex thickness. Building upon these capabilities, optical coherence tomography angiography (OCTA) allows non-invasive visualization of retinal and choroidal microvasculature without the need for dye injection. OCTA provides quantitative metrics, including vessel density and foveal avascular zone (FAZ) measurements, offering a rapid, reproducible, and clinically accessible assessment of retinal microcirculation. 4 , 5

Accumulating evidence suggests that retinal microvascular alterations may be associated with several neurological disorders. Studies involving patients with AD and MCI have reported reductions in vessel density within both the superficial capillary plexus (SCP) and deep capillary plexus (DCP), as well as alterations in FAZ parameters. Similar retinal vascular abnormalities have been described in CSVD and other conditions characterized by neurovascular dysfunction. 6 , 7 These findings support the hypothesis that retinal microvascular changes may serve as surrogate markers of broader neurovascular pathology.

Although disorders such as AD, CSVD, glaucoma-related optic neuropathy, and neuroinflammatory diseases differ in their primary etiologies, they share several overlapping mechanisms, including neurovascular dysfunction, endothelial injury, impaired autoregulation, chronic hypoperfusion, and microvascular compromise. These common pathways may result in measurable retinal vascular alterations that can be detected using OCTA. Consequently, evaluating retinal microvascular changes across these conditions may provide insights into shared mechanisms of neurovascular injury and their potential utility as imaging biomarkers.

Several systematic reviews and meta-analyses have previously examined OCTA findings in Alzheimer’s disease and related cognitive disorders. However, most have focused primarily on AD and MCI, while evidence from other neurovascular and neuroinflammatory conditions has remained comparatively underexplored. In addition, the growing number of recently published studies, together with ongoing advances in OCTA acquisition, segmentation algorithms, and image analysis techniques, warrants an updated synthesis of the available evidence. Furthermore, substantial variability in OCTA devices, scanning protocols, image processing methods, and study designs continues to contribute to heterogeneity across the literature, making interpretation of findings challenging. 8

Therefore, this systematic review and meta-analysis aimed to synthesize current evidence regarding OCTA-derived retinal microvascular alterations across neurodegenerative, neurovascular, and neuroinflammatory disorders and to evaluate their potential role as non-invasive biomarkers of neurovascular dysfunction.

Methods

Study Design

This systematic review and meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. 9 The study protocol was prospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO; registration number: CRD420261305354).

Literature Search Strategy

A comprehensive literature search was performed to identify studies evaluating retinal microvascular alterations using optical coherence tomography angiography (OCTA) in neurodegenerative, neurovascular, and neuroinflammatory disorders. The following electronic databases were searched: PubMed, Embase, Scopus, Web of Science, and the Cochrane Central Register of Controlled Trials (CENTRAL). The search strategy incorporated both Medical Subject Headings (MeSH) and free-text terms, including “optical coherence tomography angiography”, “OCTA”, “retinal vessel density”, “retinal microvasculature”, “foveal avascular zone”, “Alzheimer’s disease”, “mild cognitive impairment”, “multiple sclerosis”, “cerebral small vessel disease”, and “glaucoma”. Boolean operators (AND and OR) were used to combine search terms appropriately. Studies published between January 2015 and December 2025 were considered eligible. The final search was conducted on 15 December 2025. All retrieved records were exported to reference management software, and duplicate records were removed prior to screening.

Study Selection and Data Extraction

Study selection and data extraction were performed independently by two reviewers. Titles and abstracts were initially screened according to predefined eligibility criteria, followed by full-text assessment of potentially relevant studies. Disagreements between reviewers were resolved through discussion, and when necessary, a third reviewer was consulted to achieve consensus. Data were extracted using a standardized data collection form to ensure consistency across studies. Extracted variables included first author, publication year, study design, country of origin, sample size, participant characteristics, disease classification, diagnostic criteria, and study methodology. OCTA-related variables included device manufacturer, scan protocol, retinal layers evaluated, and segmentation methods. Quantitative outcome measures included superficial capillary plexus vessel density (SCP-VD), deep capillary plexus vessel density (DCP-VD), and foveal avascular zone (FAZ) parameters. Where reported, information regarding participant-level and eye-level analyses was recorded, and potential methodological limitations related to eye-level reporting and inter-eye correlation were considered during interpretation of the findings.

Quality Assessment

Methodological quality and risk of bias were assessed independently by two reviewers. Comparative observational studies were evaluated using the Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I) tool. 10 Although originally developed for non-randomized studies, ROBINS-I was selected because it enables systematic assessment of important sources of bias relevant to observational imaging studies, including confounding, participant selection, outcome measurement, missing data, and selective reporting. Cross-sectional non-comparative studies were additionally assessed using the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Analytical Cross-Sectional Studies. 11 This checklist evaluates study design, sampling methodology, validity of measurements, management of confounding variables, and appropriateness of statistical analyses. Risk-of-bias assessments were incorporated into the interpretation of pooled findings and overall certainty of evidence.

Statistical Analysis

Both qualitative narrative synthesis and quantitative meta-analysis were performed when studies were considered clinically and methodologically comparable. Continuous outcomes were pooled using standardized mean differences (SMDs) with corresponding 95% confidence intervals (CIs), as OCTA-derived parameters were reported using different measurement scales and imaging platforms across studies. Random-effects models with inverse-variance weighting were applied to account for anticipated between-study variability. Studies were grouped according to disease category, including Alzheimer’s disease (AD), mild cognitive impairment (MCI), cerebral small vessel disease (CSVD), glaucoma-related disorders, and neuroinflammatory conditions. Although these disorders differ in their primary etiologies, they share several pathophysiological mechanisms involving neurovascular dysfunction, endothelial impairment, impaired autoregulation, and microvascular compromise. Therefore, an overall pooled analysis was performed to evaluate whether OCTA-derived retinal microvascular alterations may represent a common neurovascular signature across disease entities, while disease-specific subgroup analyses were conducted to account for differences in underlying pathophysiology. Statistical heterogeneity was assessed using Cochran’s Q test and quantified using the I 2 statistic. Heterogeneity was interpreted according to commonly accepted thresholds, whereby I 2 values of less than 30% indicated low heterogeneity, 30–60% moderate heterogeneity, 50–90% substantial heterogeneity, and 75–100% considerable heterogeneity. Potential sources of heterogeneity were explored qualitatively based on differences in study populations, disease severity, OCTA device manufacturers, scan dimensions, segmentation algorithms, and imaging protocols. Owing to the limited number of studies within several disease-specific subgroups and inconsistent reporting of stratified data, additional quantitative exploration of these methodological factors was not considered sufficiently reliable. All statistical analyses were performed using Review Manager (RevMan) software, version 5.4 (The Cochrane Collaboration, Copenhagen, Denmark).

Reporting Bias Assessment

Publication bias was considered during evidence synthesis. However, because fewer than ten studies contributed to most pooled analyses and subgroup-specific meta-analyses, formal statistical assessment of publication bias was not undertaken, as funnel plot asymmetry and related statistical tests may yield unreliable and potentially misleading results under these conditions.

Certainty Assessment

The certainty of evidence was evaluated qualitatively according to the principles of the Grading of Recommendations Assessment, Development and Evaluation (GRADE) framework. Given that all included studies were observational in design, the certainty of evidence was initially considered low. Additional downgrading was considered where substantial heterogeneity, moderate risk of bias, or methodological limitations were present. Consequently, the certainty of evidence was judged to be low for superficial capillary plexus vessel density (SCP-VD) and very low for deep capillary plexus vessel density (DCP-VD) and foveal avascular zone (FAZ) outcomes. These certainty ratings were primarily influenced by the observational nature of the included studies, moderate risk of bias, and considerable between-study heterogeneity.

Results

Study Selection

A detailed database search resulted in 612 records from 5 electronic databases: PubMed, Embase, Scopus, Web of Science, and Cochrane CENTRAL. Of 182 duplicates we identified 430 unique studies that were screened by title/abstract with 81 eligible articles included to be included in full-text review. Following full text assessment, 68 studies were excluded for inappropriate study design, lack of relevant outcome data, use of a different population without a comparison group, and/or overlap of cohorts. In total, 13 studies fulfilled the inclusion criteria and were selected for the qualitative and quantitative synthesis. 12–24 A study selection procedure is presented in Figure 1 . Figure 1 PRISMA flow diagram illustrating the study selection process for the systematic review and meta-analysis. Flowchart of study identification and screening process for systematic review. The flowchart outlines the identification and screening process for new studies via databases and registers. The process begins with records identified from databases totaling 612, including PubMed with 532, Embase with 31, Cochrane Library with 5, Scopus with 28 and Web of Science with 16. Records removed before screening due to duplicates amount to 182. The remaining records screened are 430. Reports sought for retrieval are 81, with none not retrieved. Reports assessed for eligibility are 81, with 68 excluded due to irrelevant study design (20), irrelevant outcome (40), irrelevant population (3), overlapping cohorts (2) and comparator issues (3). Finally, 13 new studies are included in the review, with reports of new included studies also totaling 13.

Study Characteristics

Table 1 presents the characteristics of the included studies, 12–24 which were from diverse geographical regions and consisted mainly of observational designs, namely cross-sectional and case-control studies. The study population included patients with different neurological and neurovascular conditions, such as Alzheimer’s disease (AD), mild cognitive impairment (MCI), cerebral small vessel disease (CSVD), and glaucoma-related optic neuropathy. Participants were matched to healthy controls for different ages, and diagnostic criteria differed among studies but generally conformed to established standards, including the National Institute on Aging Alzheimer’s Association (NIA-AA) definition of Alzheimer’s disease and Petersen criteria of mild cognitive impairment. Sample sizes varied from small exploratory cohorts to large observational ones, reflective of the increasing research interest in OCTA derived retinal biomarkers for neurological disorders. Table 1 Baseline Characteristics of the Included Studies First Author, Year Study Design, Country Sample Size Age (Mean, SD) Gender Distribution (M/F) Neurodegenerative Disorder Type Diagnostic Criteria Used Disease Stage or Subtype Control Type El Nahas et al, 2025 12 Observational case-control study in Egypt 60 CSVD: 40 Healthy controls (HC): 20 Median (IQR) reported: CSVD group 59 (52.2–63), HC 59.5 (55.2–62.7) 55% male in both groups (22 males in the CSVD group, 11 in the HC) Cerebral small vessel disease (CSVD) MRI findings (TOAST classification, Fazekas scale, SVD score) 70% had SVD score = 1 (low severity); Fazekas grades 0–3 distributed Healthy age- and sex-matched controls Di Pippo et al, 2024 13 Observational cross-sectional single-site study in Italy 36 Early AD/MCI: 18 HC: 18 Analysis based on 36 eyes per group. AD/MCI: 75.22 ± 6.1 years HC: 75.64 ± 7.38 years AD/MCI: 10/8 HC: 10/8 Alzheimer’s Disease (AD) and Mild Cognitive Impairment (MCI) due to AD National Institute on Aging-Alzheimer’s Association (NIA-AA) criteria; confirmed by Amyloid PET scan Early AD / MCI due to AD Healthy age- and sex-matched subjects Xie et al, 2024 14 Cross-sectional study in China 158 HC: 62 AD: 55 MCI: 41 HC: 69.5 ± 6.7 years MCI: 71.8 ± 7.0 years AD: 74.2 ± 8.1 years HC: 28/34 MCI: 20/21 AD: 26/29 AD and MCI AD: National Institute on Aging and Alzheimer’s (NIA-AA) criteria for probable AD. MCI: Petersen criteria. Diagnosis made by an experienced neurologist. Defined by MMSE scores: AD: MMSE 13–20 MCI: MMSE 21–24 HC: MMSE > 24 Healthy participants were matched for educational background, age, and sex. Excluded for neurological disorders, hypotension, uncontrolled hypertension, or substance abuse. Cheng et al, 2021 15 Prospective observational study in the UK 57 eyes of 31 patients with NTG (14 eyes excluded due to poor OCTA quality) 66.2 ± 7.9 years 50% female (16 of 32 patients) Normal Tension Glaucoma (NTG) – classified as an optic neuropathy Characteristic optic nerve head/RNFL changes + glaucomatous visual field defect on SAP (24–2 SITA Fast). No recorded history of IOP >21 mmHg. Severity staged by Mean Deviation (MD): Worse eye MD: −6.5 ± 5.0 dB; Better eye MD: −1.4 ± 1.8 dB. No HC (a single cohort study) Rogaczewska et al, 2021 16 Observational study in Poland 40 MS patients, 13 NMOSD patients, 20 HCs MS: 35.15 ± 7.47; NMOSD: 42.08 ± 10.23; Controls: 37.90 ± 11.47 MS: 8/32 NMOSD: 2/11 Controls: 3/17 Multiple Sclerosis (MS), Neuromyelitis Optica Spectrum Disorder (NMOSD) MS: Revised 2017 McDonald criteria; NMOSD: Revised 2015 NMOSD diagnostic criteria; AQP4-IgG seropositive Relapsing-remitting MS; AQP4-IgG seropositive NMOSD Age- and sex-matched healthy volunteers Wang et al, 2021 17 Cross-sectional study in China 62 AD, 47 MCI, 49 HCs. AD: 71.81 ± 7.98; MCI: 72.73 ± 7.75; HC: 69.50 ± 5.94 AD: 27/35 MCI: 18/29 HC: 17/32 AD MCI 2011 NIA-AA (National Institute on Aging–Alzheimer’s Association) guidelines Early AD; MCI (prodromal stage) Age-matched healthy subjects Chua et al, 2020 18 Cross-sectional case-control in Singapore 24 AD, 37 MCI, 29 HCs. AD: 74.9 ± 6.0 MCI: 77.9 ± 6.4 HCs: 76.7 ± 5.3 AD: 7/17 MCI: 21/16; HCs: 16/13 AD MCI AD: DSM-IV; MCI: impairment in ≥1 neuropsychological domain without affecting daily living NS Cognitively normal individuals from the same clinics Lee et al, 2020 19 Prospective cross-sectional in South Korea 69 participants enrolled: 29 ADCI, 25 SVCI, 15 CN (controls). After exclusions: 60 participants (117 eyes) included for analysis: 28 ADCI (55 eyes), 18 SVCI (34 eyes), 14 CN (28 eyes) ADCI: 67.5 ± 9.5 years SVCI: 77.0 ± 6.3 years CN: 67.2 ± 6.1 years ADCI: 11/17 SVCI: 6/12 CN: 4/10 Alzheimer’s Disease Cognitive Impairment (ADCI); Subcortical Vascular Cognitive Impairment (SVCI) ADCI: NIA-AA criteria (probable AD dementia or amnestic MCI due to AD), amyloid PET positive; SVCI: clinical criteria + severe ischemia on MRI; CN: normal cognition, amyloid PET negative ADCI includes both AD dementia and amnestic MCI due to AD Cognitively normal, amyloid-negative, no severe white matter hyperintensities van de Kreeke et al, 2020 20 Cross-sectional cohort study (part of EMIF-AD PreclinAD cohort) in the Netherlands 124 68.6 ± 6.3 years 58/66 AD Amyloid PET positivity ([18F]flutemetamol), visual rating according to manufacturer criteria (GE Healthcare); cognitive health defined by TICS-m >22, GDS <11, CERAD word list recall > −1.5 SD, CDR = 0 Preclinical AD (cognitively healthy with amyloid positivity) Cognitively healthy, amyloid-negative monozygotic twins Wu et al, 2020 21 Prospective study in China 60 (21 HC, 21 MCI, 18 AD) HC: 68.67 ± 5.85 MCI: 67.81 ± 5.96 AD: 69.94 ± 6.39 HC: 11/10 MCI: 12/9 AD: 10/8 AD MCI AD: NINCDS-ADRDA criteria; MCI: Petersen criteria; Cognitive screening: MMSE ≥ 24 MCI (preclinical) and AD (dementia stage) Healthy, age- and sex-matched controls Zabel et al, 2019 22 Cross-sectional study in Poland 81 (27 AD, 27 POAG, 27 HC) HC: 74.26 ± 7.66 POAG: 71.85 ± 7.06 AD: 74.11 ± 5.87 HC: 8/19 POAG: 10/17 AD: 6/21 AD Primary Open-Angle Glaucoma (POAG) AD: DSM-IV & NIA-AA criteria, amyloid PET (Florbetapir SUVr >1.5); POAG: optic nerve neuropathy + visual field loss; HC: normal exam AD: mild to moderate dementia (MMSE 10–23); POAG: perimetric, moderate severity Healthy, age- and sex-matched controls Lahme et al, 2018 23 Prospective study in Germany 74 eyes (36 AD, 38 HC) AD: 67.97 ± 9.30; HC: 66.08 ± 10.11 AD: 15/21; HC: 14/23 Alzheimer’s disease National Institute on Aging–Alzheimer’s Association (NIA-AA) criteria Early-stage Alzheimer’s disease (MMSE ≥10) Healthy, age- and sex-matched controls Kwon et al, 2018 24 Retrospective cross-sectional study in Korea 166 eyes (126 OAG, 40 HCs) HCs: 50 ± 14 years OAG (total): 54 ± 14 years CVF+: 57 ± 14 years; CVF-: 51 ± 13 years HCs: 15/25 OAG total: 62/64 CVF+: 35/32 CVF-: 27/32 OAG Open iridocorneal angle, glaucomatous ONH damage, glaucomatous VF defect regardless of IOP. VF defect is defined as ≥3 adjacent points with P<0.05 on the pattern deviation map and one point with P<0.01, or PSD with P<0.05 on two consecutive tests. Stratified by presence (CVF+) or absence (CVF-) of central visual field defects Normal healthy eyes (no glaucoma, IOP <21 mmHg, normal VF, and optic disc) Abbreviations : ADCI, Alzheimer’s Disease Cognitive Impairment; AD, Alzheimer’s Disease; AQP4-IgG, aquaporin-4 immunoglobulin G; CDR, Clinical Dementia Rating; CERAD, Consortium to Establish a Registry for Alzheimer’s Disease; CN, cognitively normal; CSVD, cerebral small vessel disease; CVF, central visual field; DSM-IV, Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition; EMIF-AD, European Medical Information Framework for Alzheimer’s Disease; GDS, Geriatric Depression Scale; HC, healthy controls; IOP, intraocular pressure; IQR, interquartile range; MD, Mean Deviation; MCI, Mild Cognitive Impairment; MMSE, Mini-Mental State Examination; MRI, magnetic resonance imaging; MS, Multiple Sclerosis; NIA-AA, National Institute on Aging–Alzheimer’s Association; NINCDS-ADRDA, National Institute of Neurological and Communicative Disorders and Stroke–Alzheimer’s Disease and Related Disorders Association; NMOSD, Neuromyelitis Optica Spectrum Disorder; NR, not reported; NTG, Normal Tension Glaucoma; OAG, Open-Angle Glaucoma; OCTA, optical coherence tomography angiography; ONH, optic nerve head; PET, positron emission tomography; POAG, Primary Open-Angle Glaucoma; PSD, pattern standard deviation; RNFL, retinal nerve fiber layer; SAP, standard automated perimetry; SITA, Swedish Interactive Thresholding Algorithm; SUVr, standard uptake value ratio; SVCI, Subcortical Vascular Cognitive Impairment; SVD, small vessel disease; TICS-m, Telephone Interview for Cognitive Status modified; TOAST, Trial of Org 10172 in Acute Stroke Treatment; VF, visual field.

OCTA Methodology

Descriptive description regarding the employed OCTA imaging protocols within the reported studies in Table 2 . 12–24 was provided. Most of the studies used commercially available OCTA systems (Optovue RTVue XR Avanti or Zeiss Cirrus HD-5000 AngioPlex as well, and other commercial OCTAs), but the Topcon DRI OCT Triton model was used in some studies. Imaging protocols differed between studies, with most studies performing 3×3 mm, 4.5×4.5 mm, and 6×6 mm scans of the macular or peripapillary region. The vast majority of research reported vascular parameters from the superficial capillary plexus (SCP) and deep capillary plexus (DCP), whereas numerous studies also investigated radial peripapillary capillaries and the flow of choriocapillaris. Automated retinal layer segmentation was typically done through device software, with manual correction or deep learning–based retinal layer segmentation techniques reported in some studies. Different criteria for image quality, from study to study, were also applied: most studies required a minimum signal strength index or predefined quality threshold prior to image analysis. Table 2 OCTA Methodology Across the Included Studies First Author, Year OCTA Device/Manufacturer Scan Size (3×3 mm, 6×6 mm, etc). Retinal Layers Analyzed Segmentation Method (Automatic/Manual) Image quality Threshold/Signal Strength El Nahas et al, 2025 12 RTVueXR Avanti (Optovue Inc., Fremont, California, USA) Macula map: 5×5 mm; Macular OCTA: 6×6 mm; ONH OCTA: 4.5×4.5 mm Superficial capillary plexus, deep capillary plexus, macular thickness (whole image, foveal, parafoveal, perifoveal), ganglion cell complex (GCC), optic nerve head (ONH) parameters (RNFL thickness, rim area, cup/disc ratio), radial peripapillary capillaries (RPC) Automated segmentation via incorporated software (software version 2017.1) Scans with good quality (>6/10) were included; 7 patients were excluded due to poor scan quality Di Pippo et al, 2024 13 Rtvue XR Avanti (Optovue, Inc, Fremont, CA) HD Angio Retina 6×6 mm scan centered on the fovea Superficial capillary plexus, Deep capillary plexus, Choriocapillaris, Foveal Avascular Zone (FAZ) Automatic segmentation via built-in software (AngioAnalytics). Slabs are defined as: • Superficial: ILM to IPL-10µm • Deep: IPL-10µm to OPL+10µm • Choriocapillaris: Bruch membrane −10µm to +30 Scans with a quality rating below 7/10 were rejected and reacquired. Xie et al, 2024 14 Avanti RTVue XR (Optovue, Fremont, California, USA) 3x3 mm 2 and 6×6 mm 2 macular scans Superficial Vascular Complex (SVC) Deep Vascular Complex (DVC) The Inner Vascular Complex (IVC) is a combined slab of SVC and DVC. Fully automated deep learning-based framework. • Microvasculature: OCTA-Net model trained on the ROSE-1 dataset. • FAZ: FAZ-Net model. Segmentation was performed on device-generated en face angiograms of SVC, DVC, and IVC. Images with a scan quality below 7 or with artefacts were excluded. The average signal strength index (SSI) was similar between groups (AD: 8.26 vs Control: 8.07). Cheng et al, 2021 15 RTVue XR Avanti (Optovue, Inc., Fremont, CA, USA) 3×3 mm macular scan Superficial retinal capillary plexus (from ILM to posterior margin of IPL) Automated segmentation via the device’s built-in software for pfVD; VAMPIRE software for fractal dimension and FAZ analysis. Signal strength index (SSI) ≥ 40. OCTA scans with quality scores ≤ 4, motion artefact, decentration, or poor clarity were excluded. Rogaczewska et al, 2021 16 RTVue XR Avanti with AngioVue, Optovue Inc., Fremont, CA, USA; software version 2017.1.0.151 4.5 × 4.5 mm rectangle scan centered on optic nerve head; peripapillary annulus: 1.0 mm wide. Superficial retinal layers (internal limiting membrane to outer boundary of nerve fiber layer); radial peripapillary capillary (RPC) plexus. Automatic (AngioVue software) Excluded if signal strength index < 50 or significant motion artifacts Wang et al, 2021 17 Optovue Angiovue System (Optovue Inc., Fremont, CA, USA), software ReVue v2017.1.0.155 Macula: 3×3 mm 2 ; Optic nerve head (ONH): 4.5×4.5 mm 2 Superficial vascular plexus (SVP), Deep vascular plexus (DVP), Radial peripapillary capillaries (RPC) Automatic (software-based) Image quality > 6 (on device scale) Chua et al, 2020 18 Zeiss Cirrus HD-5000 with AngioPlex (Carl Zeiss Meditec) 3×3 mm 2 (macular region) Superficial capillary plexus (SCP), Deep capillary plexus (DCP) Automatic (instrument software) with manual FAZ delineation Signal strength index ≥7; poor-quality scans excluded Lee et al, 2020 19 Topcon DRI OCT Triton plus (swept-source OCT) 4.5 × 4.5 mm (peripapillary) Radial peripapillary capillary (RPC) network; Retinal nerve fiber layer (RNFL) thickness Automatic (IMAGEnet 6 software) Image quality score < 40 excluded; scans with motion artifacts, poor clarity, or off-centered disc excluded van de Kreeke et al, 2020 20 Cirrus 5000 Angioplex (Zeiss Meditec, Inc., Germany) 6 × 6 mm NR Automated (software-derived ETDRS grid) NS Wu et al, 2020 21 RTVue XR Avanti (Optovue, Inc., USA) 6 × 6 mm Superficial Retinal Capillary Plexus (SRCP) and Deep Retinal Capillary Plexus (DRCP) Automated (AngioVue software) Signal Strength Index (SSI) > 40 Zabel et al, 2019 22 Avanti RTVue XR (Optovue, Inc., USA) Macula: 6×6 mm; ONH: 4.5×4.5 mm SVP DVP RPC Automated (AngioVue software with 3D PAR) Signal Quality (SQ) ≥ 6 (on a 10-point scale) Lahme et al, 2018 23 Avanti RTVue XR (Optovue, Inc., USA) Macula: 6×6 mm; ONH: 4.5×4.5 mm SVP DVP RPC Automated (AngioVue software with 3D PAR) Signal Quality (SQ) ≥ 6 (on a 10-point scale) Kwon et al, 2018 24 Angiovue (Optovue, Inc., Fremont, CA, USA) Optic nerve head: 4.5×4.5 mm 2 Macula: 3×3 mm 2 ONH scan: Radial peripapillary capillary (RPC) layer (ILM to NFL). Macular scan: Superficial vascular complex (ILM to posterior boundary of IPL). Automatic (manufacturer’s default software) for VD; Manual for FAZ measurements using ImageJ. Signal Strength Index (SSI) <48 excluded. Minimum quality threshold for SD-OCT images: signal strength ≥7. Abbreviations : 3D PAR, 3D Projection Artifact Removal; DCP, Deep Capillary Plexus; DRI, Deep Range Imaging; DVC, Deep Vascular Complex; ETDRS, Early Treatment Diabetic Retinopathy Study; FAZ, Foveal Avascular Zone; GCC, Ganglion Cell Complex; HD, High Definition; ILM, Internal Limiting Membrane; IPL, Inner Plexiform Layer; IVC, Inner Vascular Complex; mm, millimeter; NR, Not Reported; NS, Not Specified; OCT, Optical Coherence Tomography; OCT-A, Optical Coherence Tomography Angiography; ONH, Optic Nerve Head; OPL, Outer Plexiform Layer; RPC, Radial Peripapillary Capillaries; RNFL, Retinal Nerve Fiber Layer; SCP, Superficial Capillary Plexus; SD-OCT, Spectral-Domain Optical Coherence Tomography; SQ, Signal Quality; SSI, Signal Strength Index; SS-OCT, Swept-Source Optical Coherence Tomography; SVC, Superficial Vascular Complex; VD, Vessel Density; XR, extended range.

Measurement Across Studies

The main quantitative OCTA outcomes reported in these included studies are presented in Table 3 . 12–24 Several neurological and neurovascular disorders exhibited a similar retinal microvascular remodeling pattern. Many experiments showed decreased capillary densities in the superficial (SCP) and deep (DCP) retinal capillaries when comparing the disease groups with healthy controls. 12–24 Such degradations were especially pronounced in AD and mild cognitive impairment. Enlargement of the foveal avascular zone (FAZ) has been described in a number of studies and possibly reflects compromised retinal microcirculation. Other microvascular abnormalities were reported across conditions, including cerebral small vessel disease and glaucoma-related optic neuropathy. The aggregate analysis for superficial capillary plexus vessel density (SCP-VD) showed a significant reduction in patients with both neurological and ophthalmic diseases as compared with healthy controls, with a pooled standardized mean difference (SMD) of −0.60. Similar trends were also seen across disease modes, with Alzheimer’s disease having an SMD of −0.46 and mild cognitive impairment having an SMD of −0.62. Substantial decreases were greater in cerebral small vessel disease (SMD −1.09) and glaucoma-related disorders (SMD −0.78). There was moderate between-study heterogeneity for SCP-VD (I 2 = 43%). The pooled analysis is provided in Figure 2 . Table 3 Primary and Secondary Outcomes of the Included Studies First Author, Year Vessel Density (Superficial Capillary Plexus) Vessel Density (Deep Capillary Plexus) Foveal Avascular Zone (FAZ) Perfusion density Conclusion El Nahas et al, 2025 12 Whole image CSVD: 47.6 (42.2–50.7) HC: 53.4 (52.2–54.7) Foveal CSVD: 14.85 (8.1–17.9) HC: 19.4 (15.2–22.1) Parafoveal CSVD: 50.15 (45.6–53.0) HC: 52.95 (46.5–56.1) Perifoveal CSVD: 49.2 (43.6–51.7) HC: 53.3 (49.8–54.2) Whole image CSVD: 45.95 (42.5–50.9) HC: 55.05 (49.3–60.1) Foveal CSVD: 32.45 (26.1–37.4) HC: 39.05 (34.0–41.4) Parafoveal CSVD: 53.3 (50.4–56.1) HC: 58.55 (56.1–60.4) Perifoveal CSVD: 47.15 (41.4–54.6) HC: 55.75 (48.7–60.9) Whole image thickness CSVD: 274.5 (259.5–280.7) HC: 281.0 (273.5–287.5) Foveal thickness CSVD: 240.5 (226.5–258.0) HC: 254.0 (235.0–263.0) Parafoveal thickness CSVD: 312.5 (300.5–319.7) HC: 320.5 (314.2–336.0) Perifoveal thickness CSVD: 271.0 (254.5–279.5) HC: 278.5 (267.7–291.2) FAZ area (mm 2 ) CSVD: 0.36 (0.26–0.42) HC: 0.29 (0.20–0.30) FAZ perimeter (mm) CSVD: 2.29 (1.96–2.50) HC: 2.14 (1.80–2.20) FAZ Acircularity Index (AI) CSVD: 1.10 (1.00–1.10) HC: 1.11 (1.00–1.10) NR OCTA is a noninvasive tool that can detect early retinal microvascular changes in CSVD, making it useful for identifying at-risk individuals. Di Pippo et al, 2024 13 AD/MCI: 42.77 ± 3.26 HC: 44.39 ± 3.59 AD/MCI: 43.56 ± 6.30 HC: 44.64 ± 3.89 Foveal Avascular Zone (FAZ) Area (mm 2 ) AD/MCI: 0.267 ± 0.08 HC: 0.267 ± 0.10 Flow Area - Choriocapillaris (mm 2 ) AD/MCI: 1.94 ± 0.18 HC: 2.07 ± 0.12 Flow Area - Outer Retina (mm 2 ) AD/MCI: 1.05 ± 0.50 HC: 1.00 ± 0.37 NR OCTA has revealed microvascular alterations in the retina and choriocapillaris in AD, but further studies and technical advances are needed to confirm their value as early biomarkers. Xie et al, 2024 14 Vascular Area Density (VAD): HC: 16.65 ± 2.32 MCI: 15.35 ± 2.59 AD: 15.51 ± 2.25 Vascular Length Density (VLD) HC: 6.17 ± 0.92 MCI: 5.61 ± 1.00 AD: 5.74 ± 0.89 Vascular Fractal Dimension (VFD): HC: 1.50 ± 0.03 MCI: 1.48 ± 0.04 AD: 1.49 ± 0.04 Vascular Bifurcations (VB): HC: 157 ± 35 MCI: 133 ± 38 AD: 141 ± 36 Vascular Area Density (VAD): HC: 25.06 ± 5.05 MCI: 22.06 ± 4.67 AD: 23.82 ± 4.98 Vascular Length Density (VLD): HC: 9.81 ± 1.80 MCI: 8.70 ± 1.71 AD: 9.32 ± 1.83 FAZ Area (FA): HC: 0.69 ± 0.20 MCI: 0.65 ± 0.26 AD: 0.75 ± 0.21 FAZ Roundness (FR): HC: 34.42 ± 7.00 MCI: 40.30 ± 13.27 AD: 39.58 ± 10.75 FAZ Circularity (FC): HC: 0.74 ± 0.11 MCI: 0.71 ± 0.17 AD: 0.70 ± 0.11 NR The study supports OCTA as a practical tool for diagnosing AD and MCI, with retinal biomarkers that can assist clinical decision-making. Cheng et al, 2021 15 Worse eye: 38.4 ± 4.9 Better eye: 41.3 ± 2.9 NR FAZ perimeter (mm) Worse eye: 2.26 ± 0.61 Better eye: 2.27 ± 0.52 NR Perifoveal vascular density is lower in worse glaucoma eyes and links to field loss (but is age/image-quality dependent); MSFD, reduced in NTG patients with hypertension, may be a broader vascular biomarker. Rogaczewska et al, 2021 16 HC: 52.97 ± 2.51 MS-ON: 49.87 ± 3.10 MS+ON: 46.76 ± 4.96 NMOSD-ON: 50.60 ± 1.71 NMOSD+ON: 40.80 ± 8.99 NR NA NR Optic neuritis produces a larger RPC vessel-density drop in NMOSD than in MS; preferential involvement of the superior and inferior sectors may help differentiate the two. Wang et al, 2021 17 AD: whole enface 44.66 ± 3.36% MCI: 44.00 ± 3.07% HC: 46.82 ± 2.08% AD: whole enface 49.42 ± 3.40% MCI: 49.57 ± 2.89% HC: 50.89 ± 2.86% AD: 0.34 ± 0.11 mm 2 ; MCI: 0.36 ± 0.12 mm 2 ; HC: 0.33 ± 0.12 mm 2 (no significant difference) NR Superficial retinal vascular density was lower in MCI and AD than in healthy controls, indicating retinal microvascular dysfunction. Chua et al, 2020 18 AD: 14.78 ± 1.14% MCI: 14.94 ± 1.02% HCs: 15.66 ± 0.96% AD: 20.42 ± 1.60% MCI: 20.81 ± 1.65% HCs: 21.54 ± 1.55% AD: 0.34 ± 0.14 MCI: 0.35 0.12 HCs: 0.31 ± 0.12 AD: 7.02 ± 1.29 MCI: 6.39 1.27 HCs: 6.92 ± 1.16 AD eyes show reduced macular vascular density in both plexuses, while MCI shows loss limited to the superficial plexus, highlighting retinal microvascular changes as a potential window into brain pathology. Lee et al, 2020 19 Capillary Density in RPC network (%): - Superior: ADCI: 64.15 ± 6.39 SVCI: 60.14 ± 6.42 CN: 63.16 ± 6.18 - Inferior: ADCI: 67.19 ± 7.34 SVCI: 64.06 ± 6.07 CN: 63.43 ± 7.80 - Temporal: ADCI: 45.76 ± 7.13; SVCI: 42.34 ± 6.29; CN: 48.45 ± 7.08 - Nasal: ADCI: 49.69 ± 5.52; SVCI: 50.25 ± 6.29; CN: 50.51 ± 5.59 NR NR NR First study linking peripapillary capillary density/RNFL to brain imaging in ADCI/SVCI: peripapillary microvascular changes (not RNFL) correlate with CSVD, OCTA may be a quick, noninvasive screen. van de Kreeke et al, 2020 20 NR NR 0.22 ± 0.09 mm 2 (mean ± SD) NR People with preclinical Alzheimer’s had higher retinal vessel density, suggesting early vascular changes. Wu et al, 2020 21 HC: Parafovea: 50.47 ± 2.72 Perifovea: 50.21 ± 2.1 MCI: Parafovea: 50.37 ± 2.33 Perifovea: 50.57 ± 2.57 AD: Parafovea: 49.56 ± 2.81 Perifovea: 49.37 ± 2.36 HC: Parafovea: 52.28 ± 2.89 Perifovea: 51.46 ± 3.7 MCI: Parafovea: 48.09 ± 3.88 Perifovea: 49.21 ± 4.25 AD: Parafovea: 43.1 ± 2.75 Perifovea: 43.82 ± 3.42 Area in mm 2 : HC: 0.26 ± 0.07; MCI: 0.37 ± 0.06; AD: 0.44 ± 0.08 NR MCI and AD show macular microvascular loss, most pronounced in the deep capillary plexus; OCTA can detect these early changes, with deep plexus density as a potential early AD biomarker. Zabel et al, 2019 22 HC: 48.15 ± 3.03 POAG: 39.72 ± 4.97 AD: 47.42 ± 3.04 HC: 49.46 ± 4.27 POAG: 47.44 ± 6.07 AD: 43.95 ± 5.15 Area in mm 2 : HC: 0.21 ± 0.07; POAG: 0.26 ± 0.08; AD: 0.32 ± 0.09 NR Both AD and POAG cause neurodegeneration and retinal microvascular damage. POAG affects superficial vessels, while AD targets deeper retinal layers. Lahme et al, 2018 23 Whole SVP density: HC: 48.15 ± 3.03 POAG: 39.72 ± 4.97 AD: 47.42 ± 3.04 Whole DVP density: HC: 49.46 ± 4.27 POAG: 47.44 ± 6.07 AD: 43.95 ± 5.15 Area in mm 2 : HC: 0.21 ± 0.07 POAG: 0.26 ± 0.08 AD: 0.32 ± 0.09 NR AD patients show lower peripapillary and superficial retinal flow on OCTA, reflecting cerebral vascular damage rather than AD-specific pathology. Kwon et al, 2018 24 Reported as %: - wiVD disc: Normal: 54.1 ±3.4 OAG total: 48.0 ± 6.3 - cpVD: Normal: 63.6 ± 3.5 OAG total: 57.0±7.2 - wiVD macula: Normal: 51.0 ± 4.4 OAG total: 47.2 ± 5.0 - pVD (parafoveal): Normal: 53.8 ± 5.2 OAG total: 51.2 ± 4.8 - fVD (foveal): Normal: 30.1 ± 5.0 OAG total: 28.4 ± 5.6 NR - Area (mm 2 ): Normal: 0.383 ± 0.11 OAG total: 0.431 ± 0.12 CVF+: 0.467 ± 0.12 CVF-: 0.393 ± 0.11 - Perimeter (mm): Normal: 2.702 ± 0.47 OAG total: 3.042 ± 0.63 CVF+: 3.241 ± 0.66 CVF-: 2.825 ± 0.42 NR In OAG, CTA-based FAZ size correlates with vessel density across corresponding retinal regions. Figure 2 Forest plot showing pooled retinal vessel density differences between patients with neurodegenerative or neurovascular diseases and healthy controls. Forest plot of foveal avascular zone area differences across disease subgroups versus controls, mostly positive. Forest plot analysis of foveal avascular zone area comparing disease and control groups. Columns include study details, mean differences and confidence intervals. The effect axis shows standardized mean differences with intervals from -2 to 2. Left caption indicates smaller zone in disease group; right caption indicates larger zone. Subgroup 2.1.1 AD: Studies show varied differences, eg, Chua et al, 2020 (0.23), Wu et al, 2020 (2.36). Subtotal: 0.76. Subgroup 2.1.2 MCI: Differences include Chua et al, 2020 (0.33), Wu et al, 2020 (1.66). Subtotal: 0.46. Subgroup 2.1.3 CSVD: El Nahas et al, 2025 shows subtotal of 0.65. Subgroup 2.1.4 Glaucoma: Kwon et al, 2018 shows subtotal of 0.41. Overall total effect size is 0.62 with a confidence interval of 0.29 to 0.94.

Deep Capillary Plexus Vessel Density: Meta-Analysis

The pooled analysis of deep capillary plexus vessel density (DCP-VD) in eight studies 12–24 showed a significant reduction in DCP-VD between disease versus control conditions, as indicated by an overall SMD of −0.84. Reductions were also remarkable for Alzheimer’s disease (SMD −0.92), mild cognitive impairment (SMD −0.61), and cerebral small vessel disease (SMD −1.31). Nonetheless, heterogeneity between studies was relatively large (I 2 = 78%), which may be attributed to differences in cohort, imaging, and OCTA methods associated with the selected studies. The DCP vascular density pooled forest plot is shown in Figure 3 . Figure 3 Pooled analysis of deep capillary plexus vessel density (DCP-VD) in neurodegenerative and cerebrovascular diseases. Forest plot of deep capillary plexus vessel density comparing disease and control across subgroups. Analysis of deep capillary plexus vessel density (DCP-VD) in neurodegenerative and cerebrovascular diseases. The forest plot shows effect sizes with the x-axis labeled Std. Mean Difference, ranging from -4 to 4. “Lower DCP VD in disease group” is on the left, “Higher DCP VD in disease group” on the right. Studies are grouped into Alzheimer′s disease, mild cognitive impairment and cerebral small vessel disease. Alzheimer′s disease effects: -1.15, -0.25, -3.18, -0.46, -1.16, -0.20, -0.70. Subtotal: -0.92 (95% CI: -1.43 to -0.41). Mild cognitive impairment effects: -0.61, -1.20, -0.46, -0.45. Subtotal: -0.61 (95% CI: -0.88 to -0.34). Cerebral small vessel disease effect: -1.31 (95% CI: -1.90 to -0.72). Overall total: -0.84 (95% CI: -1.15 to -0.53). Heterogeneity: I superscript 2 = 78%.

Meta-Analysis of Foveal Avascular Zone Area

At the stroke scale for FAZ area, its size and structure, meta-analysis of nine studies involved pooled FAZ area. 12–24 It showed significant FAZ enlargement in disease groups (compared with healthy control subjects), with a pooled SMD of 0.61. Subgroup analyses identified similar trends for other disease entities (eg, Alzheimer’s disease, cerebral small vessel disease, and glaucoma-associated diseases). Statistical heterogeneity was significant (I 2 = 83%), which may have resulted from differences in OCTA devices, scanning sizes, segmentation methods, and anatomical differences among the study participants. The pooled FAZ analysis is demonstrated in Figure 4 . Figure 4 Pooled analysis of foveal avascular zone (FAZ) area comparing disease groups with healthy controls. Forest plot of foveal avascular zone area differences between disease subgroups and controls. The study′s forest plot shows standardized mean differences (SMD) for subgroups: AD, MCI, CSVD, Glaucoma and Inflammatory. The x-axis ′Std. Mean Difference IV, Random, 95% CI′ ranges from -2 to +2, indicating ′Lower VD in disease group′ on the left and ′Higher VD in disease group′ on the right. Each study offers a point estimate with a 95% confidence interval and weight percentage. AD studies: Zabel et al (2019) SMD -0.24, Xie et al (2024) -0.50, Wu et al (2020) -0.32, Wang et al (2021) -0.75, Lahme et al (2018) -0.24, Di Pippo et al (2024) -0.05, Chua et al (2020) -0.83. AD subtotal SMD: -0.46. MCI studies: Xie et al (2024) -0.53, Wu et al (2020) -0.04, Wang et al (2021) -1.07, Chua et al (2020) -0.72. MCI subtotal: -0.62. CSVD: El Nahas et al (2025) -1.09. Glaucoma: Kwon et al (2018) -0.78. Inflammatory: Rogaczewska et al (2021) -1.08. Overall total SMD: -0.60, 95% CI: -0.78 to -0.43, with 572 disease and 485 control participants.

Risk of Bias Assessment

The methodological quality of the included studies was assessed with a Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I) tool. This model was used in evaluating 12 comparative observational studies, and the overall risk of bias was considered at a moderate level. The patterns of bias within various areas are shown in Figure 5 . A study with a single-cohort cross-sectional design and no comparator group was evaluated individually with the Joanna Briggs Institute (JBI) Critical Appraisal Checklist for Analytical Cross-Sectional Studies, reporting findings in Table 4 . 15 Observed heterogeneity across studies may be attributed to variations in OCTA devices, scan protocols, and disease characteristics. Sensitivity analyses were not performed due to limited subgroup sizes. Table 4 JBI Checklist Quality Assessment Findings Study 1 2 3 4 5 6 7 8 Quality Cheng et al, 2021 15 No Yes Yes Yes Yes Yes Yes Yes Low Figure 5 Risk of bias assessment of included studies using the ROBINS-I tool. Table showing risk of bias domains for various studies with judgments of moderate and low bias. The table summarizes studies with columns: Study, D1-D7, Overall. Each row lists a study, year and bias judgments for domains D1-D7, with an overall assessment. All studies (El Nahas et al, 2025; Di Pippo et al, 2024; Xie et al, 2024; Wang et al, 2021; Rogaczewska et al, 2021; Wu et al, 2020; Chua et al, 2020; Lee et al, 2020; van de Kreeke et al, 2020; Zabel et al, 2019; Lahme et al, 2018; Kwon et al, 2018) show: D1 minus, D2 plus, D3 plus, D4 plus, D5 minus, D6 plus, D7 plus, with an overall yellow circle indicating moderate bias. Domain biases are: D1 confounding, D2 participant selection, D3 intervention classification, D4 deviations from intended interventions, D5 missing data, D6 outcome measurement, D7 reported result selection. Judgement: Yellow circle indicates moderate bias, green circle indicates low bias.

Reporting Biases

A formal assessment of publication bias was not performed due to the limited number of included studies.

Certainty of Evidence

The overall certainty of evidence was interpreted as limited, considering the observational design of the included studies and the presence of heterogeneity.

Discussion

The present systematic review and meta-analysis demonstrated consistent alterations in OCTA-derived retinal microvascular parameters among patients with neurodegenerative, neurovascular, and neuroinflammatory disorders compared with healthy controls. Reduced vessel density was observed in both the superficial capillary plexus (SCP) and deep capillary plexus (DCP), while the foveal avascular zone (FAZ) was generally enlarged across disease groups. These findings support the concept that retinal microvascular alterations may reflect broader neurovascular dysfunction within the central nervous system. Given the close embryological, anatomical, and physiological relationship between the retina and the brain, retinal vascular imaging offers a unique opportunity to evaluate microvascular changes associated with neurological disease using a rapid and non-invasive approach. 25–29

The observed vascular abnormalities are biologically plausible and consistent with current understanding of neurovascular dysfunction in neurological disease. Previous studies have demonstrated impaired retinal vascular responses and altered neurovascular coupling in patients with Alzheimer’s disease and mild cognitive impairment, while similar retinal microvascular alterations have also been reported in dementia with Lewy bodies and cerebral small vessel disease. 25–27 The greater reduction observed in DCP vessel density compared with SCP vessel density may reflect the increased susceptibility of the deep retinal circulation to hypoperfusion and microvascular injury because of its location within metabolically active retinal layers and its more complex vascular architecture. Although the included disorders differ in their primary etiologies, they share several mechanisms involving endothelial dysfunction, impaired autoregulation, chronic hypoperfusion, and microvascular compromise, which may contribute to the retinal vascular changes detected by OCTA. 25–28

The findings of this review are broadly consistent with previous systematic reviews and meta-analyses evaluating OCTA biomarkers in Alzheimer’s disease and related cognitive disorders. 29 Unlike previous reviews that focused predominantly on Alzheimer’s disease and mild cognitive impairment, the present review additionally incorporated studies of cerebral small vessel disease, glaucoma-related optic neuropathy, and neuroinflammatory disorders, thereby providing a broader assessment of retinal microvascular alterations across conditions characterized by neurovascular dysfunction. By evaluating retinal microvascular alterations across multiple disease entities, this study expands the existing evidence base and explores whether OCTA-derived vascular changes may represent a common manifestation of neurovascular dysfunction rather than disease-specific phenomena alone. Nevertheless, the available evidence does not currently support the use of OCTA as a standalone diagnostic tool. Instead, OCTA-derived vascular parameters may serve as complementary non-invasive biomarkers alongside clinical assessment, structural retinal imaging, and neuroimaging modalities. 28 , 29

Several limitations should be considered when interpreting these findings. Moderate to considerable heterogeneity was observed across pooled analyses, particularly for DCP vessel density and FAZ outcomes, likely reflecting differences in study populations, disease severity, diagnostic criteria, OCTA devices, scanning protocols, segmentation algorithms, and image quality criteria. In addition, most included studies were observational in design and demonstrated a moderate overall risk of bias, particularly with respect to confounding and participant selection. Residual confounding related to age, vascular risk factors, systemic disease, and medication use may also have influenced reported associations. Furthermore, variability in whether analyses were performed at the participant level or eye level may represent an additional methodological limitation. Despite these limitations, the consistency of findings across multiple studies and disease categories supports a relationship between retinal microvascular alterations and neurovascular dysfunction. Future studies should prioritize standardized OCTA acquisition protocols, harmonized reporting methodologies, and prospective longitudinal study designs to determine whether retinal microvascular alterations precede clinical disease progression and to better define their role in disease detection, monitoring, and prognostication.

Conclusion

This systematic review and meta-analysis demonstrated consistent associations between OCTA-derived retinal microvascular alterations and neurodegenerative, neurovascular, and neuroinflammatory disorders. Reduced retinal vessel density, particularly within the superficial and deep capillary plexuses, together with enlargement of the foveal avascular zone, support the concept of a shared retina–brain vascular axis and highlight the potential of retinal imaging as a non-invasive marker of neurovascular dysfunction. However, the current evidence remains limited by methodological heterogeneity, moderate risk of bias, and the predominantly observational design of the included studies. Although OCTA offers a rapid and clinically accessible approach for evaluating retinal microcirculation, further prospective longitudinal studies using standardized acquisition and analysis protocols are required to establish the diagnostic, prognostic, and monitoring value of retinal microvascular parameters across neurological disorders.

Abbreviations

AD, Alzheimer’s disease; ADCI, Alzheimer’s disease cognitive impairment; AQP4-IgG, aquaporin-4 immunoglobulin G; CDR, Clinical Dementia Rating; CERAD, Consortium to Establish a Registry for Alzheimer’s Disease; CI, confidence interval; CN, cognitively normal; CSF, cerebrospinal fluid; CSVD, cerebral small vessel disease; DCP, deep capillary plexus; DVP, deep vascular plexus; ETDRS, Early Treatment Diabetic Retinopathy Study; FAZ, foveal avascular zone; GCC, ganglion cell complex; HC, healthy control; ILM, internal limiting membrane; IPL, inner plexiform layer; IOP, intraocular pressure; JBI, Joanna Briggs Institute; MCI, mild cognitive impairment; MMSE, Mini-Mental State Examination; MRI, magnetic resonance imaging; MS, multiple sclerosis; NMOSD, neuromyelitis optica spectrum disorder; OAG, open-angle glaucoma; OCTA, optical coherence tomography angiography; ONH, optic nerve head; PET, positron emission tomography; POAG, primary open-angle glaucoma; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses; RNFL, retinal nerve fiber layer; RPC, radial peripapillary capillaries; SAP, standard automated perimetry; SCP, superficial capillary plexus; SD, standard deviation; SMD, standardized mean difference; SSI, signal strength index; VD, vessel density.

Data Sharing Statement

The data supporting the findings of this study were obtained from previously published articles included in this systematic review and meta-analysis. All relevant data are presented within the manuscript and its tables. No additional unpublished datasets were generated during the current study.

Ethics Approval and Informed Consent

Ethical approval was not required for this study because it is a systematic review and meta-analysis based solely on previously published data. No human participants were directly involved in this study.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors declare that they have no financial or non-financial competing interests related to this work.

References

1. Ashraf G , McGuinness M , Khan MA , Obtinalla C , Hadoux X , van Wijngaarden P . Retinal imaging biomarkers of Alzheimer’s disease: a systematic review and meta‐analysis of studies using brain amyloid beta status for case definition . Alzheimers Dement . 2023 ; 15 ( 2 ): e12421 . 10.1002/dad2.12421 PMC10210353 37250908

2. Liao H , Zhu Z , Peng Y . Potential utility of retinal imaging for Alzheimer’s disease: a review . Front Aging Neurosci . 2018 ; 10 : 188 . doi: 10.3389/fnagi.2018.00188 29988470 PMC6024140

3. London A , Benhar I , Schwartz M . The retina as a window to the brain—from eye research to CNS disorders . Nat Rev Neurol . 2013 ; 9 ( 1 ): 44 – 19 . doi: 10.1038/nrneurol.2012.227 23165340

4. Pellegrini M , Vagge A , Ferro Desideri L , et al. Optical coherence tomography angiography in neurodegenerative disorders . J Clin Med . 2020 ; 9 ( 6 ): 1706 . doi: 10.3390/jcm9061706 32498362 PMC7356677

5. Rifai OM , McGrory S , Robbins CB , et al. The application of optical coherence tomography angiography in Alzheimer’s disease: a systematic review . Alzheimers Dement . 2021 ; 13 ( 1 ): e12149 . 10.1002/dad2.12149 PMC7927164 33718582

6. García-Bermúdez MY , Vohra R , Freude K , et al. Potential retinal biomarkers in Alzheimer’s disease . Int J Mol Sci . 2023 ; 24 ( 21 ): 15834 . doi: 10.3390/ijms242115834 37958816 PMC10649108

7. Shen Z , Yu W , Gao S , Zhang S , Hong C . Unveiling potential retinal and choroidal markers with SS-OCTA in detecting cerebral small vessel disease . Sci Rep . 2025 ; 15 ( 1 ): 39221 . doi: 10.1038/s41598-025-21628-9 41214033 PMC12603291

8. Ibrahim Y , Xie J , Macerollo A , et al. A systematic review on retinal biomarkers to diagnose dementia from OCT/OCTA images . J Alzheimer’s Dis Rep . 2023 ; 7 ( 1 ): 1201 – 1235 . doi: 10.3233/ADR-230042 38025800 PMC10657718

9. Page MJ , McKenzie JE , Bossuyt PM , et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews . BMJ . 2021 ; 372 : n71 . doi: 10.1136/bmj.n71 33782057 PMC8005924

10. Sterne JAC , Hernán MA , Reeves BC , et al. ROBINS-I: a tool for assessing risk of bias in non-randomised studies of interventions . BMJ . 2016 ; 355 : i4919 . doi: 10.1136/bmj.i4919 27733354 PMC5062054

11. Moola S , Munn Z , Tufanaru C , et al. Chapter 7: systematic reviews of etiology and risk . In: Aromataris E , Munn Z , editors. Joanna Briggs Institute Reviewer’s Manual . The Joanna Briggs Institute ; 2017 .

12. El Nahas N , Roushdy T , Wahid El Din M , et al. Retinal vessel density using optical coherence tomography angiography as a potential biomarker in patients with cerebral small vessel disease . Egypt J Neurol Psychiatry Neurosurg . 2025 ; 61 ( 1 ): 45 . doi: 10.1186/s41983-025-00973-6

13. Di Pippo M , Cipollini V , Giubilei F , Scuderi G , Abdolrahimzadeh S . Retinal and choriocapillaris vascular changes in early Alzheimer disease patients using optical coherence tomography angiography . J Neuroophthalmol . 2024 ; 44 ( 2 ): 184 – 189 . doi: 10.1097/WNO.0000000000001907 37351953

14. Xie J , Yi Q , Wu Y , et al. Deep segmentation of OCTA for evaluation and association of changes of retinal microvasculature with Alzheimer’s disease and mild cognitive impairment . Br J Ophthalmol . 2024 ; 108 ( 3 ): 432 – 439 . doi: 10.1136/bjo-2022-321399 36596660 PMC10894818

15. Cheng KK , Tan BL , Brown L , et al. Macular vessel density, branching complexity and foveal avascular zone size in normal tension glaucoma . Sci Rep . 2021 ; 11 ( 1 ): 1056 . doi: 10.1038/s41598-020-80080-z 33441769 PMC7807020

16. Rogaczewska M , Michalak S , Stopa M . Optical coherence tomography angiography of peripapillary vessel density in multiple sclerosis and neuromyelitis optica spectrum disorder: a comparative study . J Clin Med . 2021 ; 10 ( 4 ): 609 . doi: 10.3390/jcm10040609 33562808 PMC7915450

17. Wang X , Zhao Q , Tao R , et al. Decreased retinal vascular density in Alzheimer’s disease (AD) and mild cognitive impairment (MCI): an optical coherence tomography angiography (OCTA) study . Front Aging Neurosci . 2021 ; 12 : 572484 . doi: 10.3389/fnagi.2020.572484 33519415 PMC7843508

18. Chua J , Hu Q , Ke M , et al. Retinal microvasculature dysfunction is associated with Alzheimer’s disease and mild cognitive impairment . Alzheimer’s Res Ther . 2020 ; 12 ( 1 ): 161 . doi: 10.1186/s13195-020-00724-0 33276820 PMC7718666

19. Lee JY , Kim JP , Jang H , et al. Optical coherence tomography angiography as a potential screening tool for cerebral small vessel diseases . Alzheimer’s Res Ther . 2020 ; 12 ( 1 ): 73 . doi: 10.1186/s13195-020-00638-x 32527301 PMC7291486

20. Van De Kreeke JA , Nguyen HT , Konijnenberg E , et al. Optical coherence tomography angiography in preclinical Alzheimer’s disease . Br J Ophthalmol . 2020 ; 104 ( 2 ): 157 – 161 . doi: 10.1136/bjophthalmol-2019-314127 31118186 PMC7025728

21. Wu J , Zhang X , Azhati G , Li T , Xu G , Liu F . Retinal microvascular attenuation in mental cognitive impairment and Alzheimer’s disease by optical coherence tomography angiography . Acta ophthalmologica . 2020 ; 98 ( 6 ): e781 – 7 . doi: 10.1111/aos.14381 32153141

22. Zabel P , Kaluzny JJ , Wilkosc-Debczynska M , et al. Comparison of retinal microvasculature in patients with Alzheimer’s disease and primary open-angle glaucoma by optical coherence tomography angiography . Invest Ophthalmol Visual Sci . 2019 ; 60 ( 10 ): 3447 – 3455 . doi: 10.1167/iovs.19-27028 31408108

23. Lahme L , Esser EL , Mihailovic N , et al. Evaluation of ocular perfusion in Alzheimer’s disease using optical coherence tomography angiography . J Alzheimers Dis . 2018 ; 66 ( 4 ): 1745 – 1752 . doi: 10.3233/JAD-180738 30507577

24. Kwon J , Choi J , Shin JW , Lee J , Kook MS . An optical coherence tomography angiography study of the relationship between foveal avascular zone size and retinal vessel density . Invest Ophthalmol Visual Sci . 2018 ; 59 ( 10 ): 4143 – 4153 . doi: 10.1167/iovs.18-24168 30105369

25. Katsimpris A , Karamaounas A , Sideri AM , Katsimpris J , Georgalas I , Petrou P . Optical coherence tomography angiography in Alzheimer’s disease: a systematic review and meta-analysis . Eye . 2022 ; 36 ( 7 ): 1419 – 1426 . doi: 10.1038/s41433-021-01648-1 34193983 PMC9232568

26. Gao Y , Kwapong WR , Zhang Y , et al. Retinal microvascular changes in white matter hyperintensities investigated by swept source optical coherence tomography angiography . BMC Ophthalmol . 2022 ; 22 ( 1 ): 77 . doi: 10.1186/s12886-021-02143-7 35168582 PMC8845341

27. Chua J , Singh S , Gong S , et al. Combining OCT and OCT angiography improves differentiation of mild cognitive impairment and Alzheimer’s disease . Alzheimer’s Res Ther . 2026 ; 18 . doi: 10.1186/s13195-026-01955-3 PMC12918493 41593765

28. Querques G , Borrelli E , Sacconi R , et al. Functional and morphological changes of the retinal vessels in Alzheimer’s disease and mild cognitive impairment . Sci Rep . 2019 ; 9 ( 1 ): 63 . doi: 10.1038/s41598-018-37271-6 30635610 PMC6329813

29. Albanese GM , Gharbiya M , Visioli G , et al. Neuroretinal and microvascular retinal features in dementia with Lewy body assessed by optical coherence tomography angiography . Neurol Sci . 2025 ; 46 ( 1 ): 185 – 194 . doi: 10.1007/s10072-024-07683-6 39152330 PMC11698750