Comparison of Longitudinal Retinal and Choroidal Imaging Findings in Patients With Mild Cognitive Impairment and Cognitively Normal Control Participants

Abstract

Purpose: To compare longitudinal retinal and choroidal imaging findings in patients with mild cognitive impairment and cognitively normal control participants. Methods: Eyes of patients with mild cognitive impairment and cognitively normal controls were imaged using Zeiss Cirrus HD-5000 optical coherence tomography (OCT) with AngioPlex OCT angiography (OCTA). Macular vessel density and perfusion density were assessed. The Mini-Mental State Examination was administered at study entry and at least 18 months later. The velocity of change in OCT and OCTA parameters was compared between controls, nonamnestic patients with mild cognitive impairment, and amnestic patients with mild cognitive impairment. Results: Eighty-five eyes of 43 patients with mild cognitive impairment and 85 eyes of 43 control participants were analyzed at baseline, and 45% of those were analyzed at an average follow-up of 28 months. The entire cohort of patients with mild cognitive impairment demonstrated faster rates of decline in perfusion density (/year) in 3-mm circle (P = .02) and ring (P = .01) as well as in vessel density (mm^-1/year) in 3-mm circle (P = .05) and ring (P = .04) vs control participants. Amnestic patients with mild cognitive impairment demonstrated faster decline in perfusion density in 3-mm circle compared with nonamnestic patients with mild cognitive impairment (P = .04) and control participants (P = .007). Amnestic patients had a greater rate of decline in vessel density in 3-mm circle compared with nonamnestic patients (P = .03) and control participants (P = .02). The entire group of patients with mild cognitive impairment and the subgroup of nonamnestic patients had a faster rate of decline in perfusion density compared with control participants (P = .01 and P < .001, respectively). Conclusions: The rate of decline in perfusion and vessel density was faster in the entire cohort of patients with mild cognitive impairment compared with control participants, with no significant change in Mini-Mental State Examination score. Amnestic patients with mild cognitive impairment demonstrated faster rates of decline in perfusion and vessel density compared with nonamnestic patients with mild cognitive impairment.

Keywords

OCT OCT angiography imaging retina diagnostic test

Introduction

Mild cognitive impairment, subdivided into amnestic and nonamnestic types, is a neurodegenerative disease defined as cognitive decline greater than anticipated for a person’s age and education level, with maintenance of functional independence.¹ Amnestic mild cognitive impairment, primarily characterized by memory loss, is associated with an annual conversion rate to Alzheimer disease of up to 28%.^2
–4 The nonamnestic type is characterized by cognitive impairment in either a single or multiple domains, such as executive function, visuospatial skills, language, attention, or orientation.⁴ This form of mild cognitive impairment is associated with conversion to other types of dementia, such as Lewy body dementia and frontotemporal dementia.^1,5,6

Despite its risk of progression to dementia, a diagnosis of mild cognitive impairment may be challenging and is dependent upon clinical interpretation. The core clinical criteria for a diagnosis of mild cognitive impairment, include evidence of concern about cognitive change, cognitive impairment in any cognitive domain, maintenance of daily functional independence, and absence of dementia.⁷ This approach is qualitative, however, and may be subject to interpretation. Quantifiable screening tests such as the Mini-Mental State Examination are used in clinical practice, but such tests may not be independently reliable for identifying or tracking progression of disease, and outcomes of testing may vary based on a patient’s education level.⁸ As a result, relying solely on these tests can lead to misdiagnosis, confounded by poor sensitivity to early signs of disease.⁹ In addition, these tests alone may be unable to accurately stage cognitive impairment or predict progression to dementia.⁸

Recent advances have resulted in improved biomarkers and include volumetric magnetic resonance imaging (MRI) measurements of the hippocampus,¹⁸ F-fluorodeoxyglucose positron emission tomography (FDG-PET), and cerebrospinal fluid (CSF) markers of amyloid-beta 42, total tau, and phosphorylated tau.¹⁰ Several blood-based biomarkers, such as cDNA clone MGC:31944 and polyadenylation-specific factor 3, have successfully distinguished between mild cognitive impairment and normal cognition.¹¹ However, these tests are expensive and invasive, thus limiting their clinical use for diagnosis and tracking disease progression.

The retina shares an embryonic origin with the brain, and as the only central nervous system (CNS) tissue that is directly optically accessible, is uniquely suited to high-resolution imaging with optical coherence tomography (OCT) and OCT angiography (OCTA). Structural retinal changes have been observed in individuals with mild cognitive impairment.^12–15 Compared with individuals with normal cognition, decreases in superficial capillary plexus parafoveal perfusion density in individuals with amnestic and nonamnestic mild cognitive impairment have been reported.^12,13

Data on longitudinal changes in the retina and choroid in individuals with mild cognitive impairment and whether they differ from normal aging are currently sparse. OCT and OCTA were used to longitudinally characterize and compare the velocity of change of retinal and choroidal measurements of a cohort of patients with nonamnestic and amnestic mild cognitive impairment and control participants with normal cognition.

Methods

Study Design

This longitudinal study (Clinicaltrials.gov identifier NCT03233646) was approved by the Duke Health Institutional Review Board (Pro00111831), complied with the Health Insurance Portability and Accountability Act of 1996, and adhered to all tenets of the Declaration of Helsinki. Written informed consent was obtained from all participants or their legally authorized representative before participation.

Participants

Participants were prospectively recruited and enrolled in the Eye Multimodal Imaging in Neurodegenerative Disease study between January 1, 2017, and January 1, 2023. Patients from Duke and UNC Memory Disorder Clinics who were diagnosed with mild cognitive impairment by a memory specialist (A.C.B., K.G.J., A.J.L.) were included. To address the diagnostic complexity of mild cognitive impairment, all participants underwent a second consensus review (A.J.L.) according to clinical diagnostic criteria (National Institute on Aging and Alzheimer’s Association clinical guidelines), cognition and functional assessments, and family collateral reports.⁷ Patients were further categorized as amnestic or nonamnestic based on Montreal Cognitive Assessment delayed memory subscore results and clinical history (K.G.J.).² Patients with dementia due to Alzheimer disease, Lewy body dementia, frontotemporal dementia, Parkinson disease, Huntington disease, stroke, or other neurodegenerative diseases were excluded. Control participants were cognitively normal volunteers aged 50 years and older, recruited from the surrounding community, Duke Neurological Disorders Clinic, or Duke Alzheimer’s Disease Prevention Registry. Control participants were age- and sex-matched to study patients (age-matched within 5 years). The minimum follow-up time was 18 months and was matched (within 3 months) for control participants and study patients.

Medical exclusion criteria included diabetes, uncontrolled hypertension, glaucoma, and retinal or optic nerve pathology that might confound OCT or OCTA measurements. Ultra-widefield fundus photography (Optos) was performed at each visit to evaluate for retinal and choroidal pathology that would exclude participants from the study. Participants with corrected distance Snellen visual acuity (VA) <20/40 at image acquisition or a refractive error exceeding ± 6 diopters were excluded. At each visit, participants underwent cognitive evaluation using the Mini-Mental State Examination. A score of less than 27 was considered predictive of mild cognitive impairment.^16,17 Subjects’ age, sex, race, smoking history, family history of dementia, years of education, and personal history of diabetes, uncontrolled hypertension, stroke, or cardiac event were recorded.

OCT and OCTA Image Acquisition

At each visit, OCT and OCTA imaging were performed using the Zeiss Cirrus HD-5000 AngioPlex (software version 11.0.0.29946, Carl Zeiss Meditec). Image acquisition included a 512×128 macular cube scan, high-definition enhanced-depth imaging 21-line foveal scan, and 200×200 optic disc cube scan. Average retinal nerve fiber layer (RNFL), ganglion cell–inner plexiform layer (GC-IPL), and central subfield thickness (CST) were calculated using Zeiss AngioPlex software. A 3.46-mm diameter annulus around the optic disc was used to calculate RNFL thickness. GC-IPL thickness was calculated from the average of 6 segments of a 60-degree, 14.13-mm² ellipse centered around the fovea. CST was calculated as the measurement between the retinal pigment epithelium (RPE) and internal limiting membrane (ILM) on the 512×128 macular cube scan. Subfoveal choroidal thickness was manually measured as the vertical distance from the hyperreflective line of Bruch membrane to the innermost hyperreflective line of the choroid-scleral interface at the center of the fovea. Subfoveal choroidal thickness was manually measured by a group of trained graders (A.H., J.P.M., C.B.R., A.K., A.A., S.J.) masked to participants’ diagnosis and led by a third expert grader (D.S.G.).

The choroidal vascularity index was calculated using Comprehensive Ocular Imaging Network (www.ocularimaging.net) software.¹⁸ Using enhanced-depth imaging foveal scans, images were automatically binarized. The choroid, the area between the outer border of the hyperreflective RPE and the sclerochoroidal junction, was manually outlined by a trained grader (A.H.) and confirmed by an expert grader (D.S.G.). A 1500-μm wide horizontal region of interest centered on the fovea was used, and total choroidal area was calculated as the area of choroid outlined within the region of interest. All images were manually reviewed for accuracy and quality assessment by expert graders (V.G., R.A.). The choroidal vascularity index was then calculated with the Comprehensive Ocular Imaging Network software as the ratio of luminal area to total choroidal area.¹⁹

OCTA image acquisition for each eye included a 3-mm and 6-mm scan centered on the fovea and a 4.5-mm scan centered on the optic disc. All images were manually reviewed for image quality by trained study personnel. Images with signal strength less than 7 of 10, segmentation artifact, motion artifact, shadow artifact, or focal signal loss were excluded from analysis. Foveal avascular zone (FAZ) in the superficial capillary plexus was automatically segmented from the 3-mm OCTA scan using AngioPlex software. FAZ measurements were reviewed by trained personnel and adjusted if tracing or segmentation error was present. Superficial capillary plexus vessel and perfusion densities were automatically generated with AngioPlex software in the Early Treatment Diabetic Retinopathy Study (ETDRS) grid, centered on the fovea in the 3-mm and 6-mm circles and rings. Vessel density was calculated as the total length of blood vessels in the region of interest in inverse millimeters (mm^-1). Perfusion density, a unitless parameter, was calculated as the area of perfused vasculature per unit area in the region of interest.

In the 4.5-mm OCTA scan centered on the optic disc, the radial peripapillary capillary plexus was defined as the area between the ILM and the outer boundary of the RNFL. The average capillary perfusion density and average capillary flux index across the radial peripapillary capillary plexus were automatically calculated with AngioPlex software. Capillary perfusion density was defined as the percent area of perfused capillary within each region, and capillary flux index was defined as the ratio of the area of perfused capillary per unit area, weighted by the brightness of the flow signal.

Statistical Analysis

Categorical demographic variables were compared using Fisher exact test. Continuous variables were compared using Wilcoxon rank-sum test. Analysis of covariance was used, with the Mini-Mental State Examination as the dependent variable and years of education as the independent variable. OCT and OCTA variables were compared at study entry (baseline) and final follow-up, and velocity of change was calculated using each patient’s 2 study visits. OCT and OCTA velocity variables were calculated by subtracting values collected at study entry from those at follow-up, divided by the time interval in years. Percent change per year of OCTA variables was calculated by dividing the velocity by the average baseline value. OCT and OCTA retinal and choroidal imaging differences between groups (controls, nonamnestic mild cognitive impairment group, amnestic mild cognitive impairment group, entire mild cognitive impairment group) were analyzed using multivariate generalized estimating equations to account for the correlation between 2 eyes of the same subject. For all tests, P ≤ .05 was statistically significant. All data were analyzed using SAS/STAT software (version 9.4, SAS Institute Inc). Due to the exploratory nature of this study, we did not perform adjustments using Bonferroni correction for multiple comparisons.

Results

At baseline, 85 eyes of 43 patients with mild cognitive impairment and 85 eyes of 43 control participants were analyzed. Thirteen participants were diagnosed with nonamnestic mild cognitive impairment, and 29 were diagnosed with the amnestic type. One participant converted from amnestic mild cognitive impairment to Alzheimer disease over the course of the study and was excluded from both baseline and longitudinal analyses. Forty-two subjects in each group were analyzed. Each group comprised 52% male participants and 48% female participants. The mean (± SD) years of education were 17.6 ± 2.1 for control participants and 16.4 ± 2.3 for the entire group with mild cognitive impairment (P = .02). Demographic variables for the groups are shown in Table 1. The average Mini-Mental State Examination score at baseline was 29.5 ± 0.9 in control participants and 26.3 ± 3.8 in the entire group with mild cognitive impairment (P < .001). The average Mini-Mental State Examination score was 27.7 for nonamnestic patients with mild cognitive impairment and 26.1 for amnestic patients with mild cognitive impairment (P = .08). After adjusting for years of education, analysis of covariance found a significant difference in Mini-Mental State Examination score between the entire group with mild cognitive impairment and control participants (P < .001). There was no significant change in Mini-Mental State Examination score between the entire group with mild cognitive impairment and control participants (P = .63).

Table 1.

Demographic Information of the Entire Group With Mild Cognitive Impairment and Control Participants.

Variable	Statistic	Entire Group With Mild Cognitive Impairment	Control Participants	P Value^a
Age at baseline (y)	N	42	42
	Mean (SD)	71.8 (6.7)	71.8 (5.7)
	Min, median, max	58.2, 72.8, 82.2	58.4, 71.9, 83.8	.67
Years of education	Mean (SD)	16.4 (2.3)	17.6 (2.1)
	Min, median, max	12.0, 16.0, 22.0	14.0, 18,0, 23.0	.02
Mini-Mental State Examination score	Mean (SD)	26.3 (3.8)	29.5 (0.9)
	Min, median, max	15.0, 27.0, 30.0	27.0, 30.0, 30.0	< .001
Male sex	N (%)	22 (52.4)	22 (52.4)	.99
Race	N (%)			.57
White		36 (85.7)	36 (85.7)
Black		0	2 (4.8)
Hispanic		5 (11.9)	4 (9.5)
Other		1 (2.4)	0
Hypertension	N (%)	21 (48.8)	18 (41.9)	.67
Stroke	N (%)	1 (2.3)	0	.99
Coronary disease	N (%)	8 (18.6)	6 (13.9)	.771
Smoking history	N (%)	23 (54.8)	19 (45.2)	.822
Family history of dementia	N (%)	22 (52.4)	20 (47.6)	.999

P value age and years of education based on Wilcoxon rank-sum test. P value for categorical variables based on Fisher exact test.

There were no significant differences in baseline OCT and OCTA parameters between the entire group with mild cognitive impairment and control participants (Table 2). These data are shown in Supplementary Table 1. At baseline, significantly lower perfusion density in the 3-mm circle and ring was found in amnestic patients with mild cognitive impairment compared with nonamnestic patients (P = .03 and P = .04, respectively). Amnestic patients with mild cognitive impairment also had significantly lower vessel density in the 3-mm circle and ring compared with nonamnestic patients (P = .03 and P = .04, respectively). A significantly smaller FAZ area was found in amnestic patients with mild cognitive impairment compared with control participants (P = .03).

Table 2.

Average OCT and OCTA Measures in the Entire Group With Mild Cognitive Impairment and Control Participants at Baseline.

Parameters	Mean(± SD)
	Entire Group With Mild Cognitive Impairment	Control Participants	P Value^c
	N = 83 eyes(42 subjects)	N = 83 eyes(42 subjects)	P Value^c
OCT parameters
Central subfield thickness (µm)	263 (22.6)	257 (19.0)	.07
Average ganglion-cell inner plexiform layer thickness (µm)	76.1 (7.25)	76.4 (6.63)	.69
Average retinal nerve fiber layer thickness (µm)	88.9 (9.26)	88.4 (9.89)	.96
Subfoveal choroidal thickness (µm)	259 (73.4)	273 (80.0)	.39
Choroidal vascularity index (%)	65.8 (1.66)	66.5 (2.35)	.11
OCTA parameters
Foveal avascular zone area (mm²)	0.220 (0.092)	0.244 (0.081)	.17
3-mm circle perfusion density^a	0.359 (0.038)	0.355 (0.025)	.56
3-mm ring perfusion density^a	0.378 (0.038)	0.376 (0.026)	.80
3-mm circle vessel density (mm^-1)	19.8 (2.23)	19.6 (1.54)	.70
3-mm ring vessel density (mm^-1)	20.8 (2.22)	20.7 (1.54)	.98
6-mm circle perfusion density^a	0.430 (0.042)	0.424 (0.037)	.38
6-mm inner ring perfusion density^a	0.422 (0.048)	0.415 (0.043)	.38
6-mm outer ring perfusion density^a	0.440 (0.043)	0.436 (0.035)	.61
6-mm circle vessel density (mm^-1)	17.54 (1.61)	17.4 (1.37)	.56
6-mm inner ring vessel density (mm^-1)	17.63 (1.82)	17.3 (1.66)	.35
6-mm outer ring vessel density (mm^-1)	17.76 (1.62)	17.7 (1.34)	.87
Average capillary perfusion density (%)	0.444 (0.024)	0.438 (0.018)	.36
Average capillary flux index^b	0.427 (0.039)	0.430 (0.036)	.62

Abbreviations: OCT, optical coherence tomography; OCTA, optical coherence tomography angiography.

Perfusion density is a unitless measure.

Flux is a unitless measure.

P value based on test of difference between means using generalized estimating equations to adjust for the correlation between eyes of the same subject.

The average patient follow-up was 28 months (range, 19-42). Follow-up imaging was obtained in 38 eyes of 19 patients with mild cognitive impairment and 38 eyes of 19 control participants (45% return rate, partly due to COVID-19 and caregiver stressors). Follow-up OCT and OCTA data are shown in Table 3. At follow-up, choroidal vascular index and FAZ area were significantly lower in the entire group with mild cognitive impairment compared with control participants (P = .03 and P = .006, respectively). Follow-up OCT and OCTA data between controls, nonamnestic patients with mild cognitive impairment, and amnestic patients with mild cognitive impairment are shown in Supplementary Table 2. Subfoveal choroidal thickness increased over time in the entire group with mild cognitive impairment (7.224 ± 19.961 µm/year) while it decreased in the control participants (-1.811 ± 16.397 µm/year) (P = .05).

Table 3.

Average OCT and OCTA Measures in the Entire Group With Mild Cognitive Impairment and Control Participants at Follow-Up (Average of 28 Months).

Parameters	Mean(± SD)
	Entire Group With Mild Cognitive Impairment	Control Participants	P Value^c
	N = 38 eyes (19 subjects)	N = 38 eyes (19 subjects)	P Value^c
OCT parameters
Central subfield thickness (µm)	265 (29.9)	254 (18.7)	.13
Average ganglion-cell inner plexiform layer thickness (µm)	75.5 (9.18)	76.5 (6.74)	.68
Average retinal nerve fiber layer thickness (µm)	85.3 (12.95)	86.8 (9.95)	.59
Subfoveal choroidal thickness (µm)	276 (78.5)	265 (93.5)	.66
Choroidal vascularity index (%)	66.1 (1.34)	67.1 (1.87)	.03
OCTA parameters
Foveal avascular zone area (mm²)	0.195 (0.112)	0.268 (0.071)	.006
3-mm circle perfusion density^a	0.336 (0.043)	0.346 (0.039)	.25
3-mm ring perfusion density^a	0.353 (0.045)	0.368 (0.039)	.15
3-mm circle vessel density (mm^-1)	18.5 (2.54)	18.9 (2.27)	.37
3-mm ring vessel density (mm^-1)	19.5 (2.65)	20.2 (2.34)	.22
6-mm circle perfusion density^a	0.408 (0.049)	0.412 (0.047)	.86
6-mm inner ring perfusion density^a	0.404 (0.043)	0.409 (0.053)	.73
6-mm outer ring perfusion density^a	0.417 (0.052)	0.422 (0.047)	.81
6-mm circle vessel density (mm^-1)	16.8 (1.76)	16.9 (1.77)	.96
6-mm inner ring vessel density (mm^-1)	17.0 (1.51)	17.1 (2.02)	.83
6-mm outer ring vessel density (mm^-1)	16.9 (1.90)	17.1 (1.77)	.83
Average capillary perfusion density (%)	0.434 (0.023)	0.438 (0.019)	.58
Average capillary flux index^b	0.425 (0.037)	0.439 (0.025)	.14

Abbreviations: OCT, optical coherence tomography; OCTA, optical coherence tomography angiography.

Perfusion density is a unitless measure.

Flux is a unitless measure.

P value based on test of difference between means using generalized estimating equations to adjust for the correlation between eyes of the same subject.

When analyzing velocity of change of OCTA parameters from baseline, the entire group with mild cognitive impairment had a significantly faster decline per year in perfusion and vessel density in the 3-mm circle and ring compared with control participants. In addition, amnestic patients with mild cognitive impairment had a significantly faster decline in perfusion density in the 3-mm circle (−0.013 ± 0.017/year) compared with nonamnestic patients (−0.004 ± 0.008/year; P = .04) and control participants (−0.001 ± 0.018/year; P = .007). Amnestic patients with mild cognitive impairment had a significantly faster velocity of decline in perfusion density in the 3-mm ring (-0.013 ± 0.018/year) compared with control participants (−0.001 ± 0.017/year; P = .008). Amnestic patients with mild cognitive impairment also had a faster velocity of decline in vessel density in the 3-mm circle (−0.761 ± 0.971 mm^-1/year) compared with nonamnestic patients (-0.205 ± 0.416 mm^-1/year; P = .03) and control participants (−0.123 ± 1.030 mm^-1/year; P = .02). Similarly, amnestic patients with mild cognitive impairment had a faster velocity of decline of vessel density in the 3-mm ring (−0.735 ± 1.015 mm^-1/year) compared with control participants (−0.093 ± 1.041 mm^-1/year; P = .02).

The velocity of decline from baseline per year for perfusion and vessel density in the entire group with mild cognitive impairment was greater in all regions of the 6-mm ETDRS circle by up to 2.6%. The entire group with mild cognitive impairment had a faster decline in average capillary perfusion density compared with control participants (P = .01). Nonamnestic patients with mild cognitive impairment had a faster velocity of decline in average capillary perfusion density compared with control participants (P < .001). Nonamnestic patients with mild cognitive impairment demonstrated a decrease in FAZ area of 0.002 mm²/year, while control participants had an increase of 0.006 mm²/year (P = .01). The average OCT and OCTA velocity of change measurements through final follow-up are shown in Tables 4 and 5.

Table 4.

Velocity of Change of OCT and OCTA Measures in the Entire Group With Mild Cognitive Impairment and Control Participants.

Parameters	Mean(± SD)
	Entire Group With Mild Cognitive Impairment	Control Participants	P Value^c
	N = 38 eyes (19 subjects)	N = 38 eyes (19 subjects)	P Value^c
OCT parameters
Central subfield thickness (µm/y)	0.854 (6.194)	−0.059 (2.781)	.48
Average ganglion-cell inner plexiform layer thickness (µm/y)	−0.422 (1.264)	0.146 (1.281)	.07
Average retinal nerve fiber layer thickness (µm/y)	−0.598 (1.608)	0.035 (1.467)	.15
Subfoveal choroidal thickness (µm/y)	7.224 (19.961)	−1.811 (16.397)	.05
Choroidal vascularity index (%/y)	0.026 (0.895)	0.003 (1.037)	.89
OCTA parameters
Foveal avascular zone area (mm²/y)	−0.001 (0.019)	0.006 (0.014)	.10
3-mm circle perfusion density (/y)^a	−0.011 (0.016)	−0.001 (0.018)	.02
3-mm ring perfusion density (/y)^a	−0.011 (0.016)	−0.001 (0.017)	.01
3-mm circle vessel density (mm^-1/y)	−0.610 (0.892)	−0.123 (1.030)	.05
3-mm ring vessel density (mm^-1/y)	−0.608 (0.910)	−0.093 (1.041)	.04
6-mm circle perfusion density (/y)^a	−0.011 (0.021)	−0.004 (0.015)	.09
6-mm inner ring perfusion density (/y)^a	−0.008 (0.024)	−0.001 (0.022)	.13
6-mm outer ring perfusion density (/y)^a	−0.011 (0.021)	−0.005 (0.013)	.11
6-mm circle vessel density (mm^-1/y)	−0.388 (0.779)	−0.140 (0.528)	.07
6-mm inner ring vessel density (mm^-1/y)	−0.315 (0.913)	−0.048 (0.797)	.11
6-mm outer ring vessel density (mm^-1/y)	−0.419 (0.804)	−0.168 (0.472)	.08
Average capillary perfusion density (%/y)	−0.004 (0.006)	0.000 (0.005)	.01
Average capillary flux index (/y)^b	−0.003 (0.011)	0.002 (0.012)	.20

Abbreviations: OCT, optical coherence tomography; OCTA, optical coherence tomography angiography.

Perfusion density is a unitless measure.

Flux is a unitless measure.

P value based on test of difference between means using generalized estimating equations to adjust for the correlation between eyes of the same subject.

Table 5.

Velocity of Change of OCT and OCTA Measures in Control Participants and Nonamnestic and Amnestic Patients With Mild Cognitive Impairment.

Parameters	Mean(± SD)
	Control Participants	Nonamnestic Patients With Mild Cognitive Impairment	Amnestic Patients With Mild Cognitive Impairment	P Value^c	P Value^c	P Value^c
	N = 38 eyes (19 subjects)	N = 28 eyes (14 subjects)	N = 10 eyes (5 subjects)	Control vsNonamnestic	Control vsAmnestic	Nonamnestic vsAmnestic
OCT parameters
Central subfield thickness (µm/y)	−0.059 (2.781)	−3.814 (5.382)	2.064 (5.884)	.04	.12	.008
Average ganglion-cell inner plexiform layer thickness (µm/y)	0.146 (1.281)	−0.680 (1.533)	−0.350 (1.205)	.17	.124	.590
Average retinal nerve fiber layer thickness (µm/y)	0.035 (1.467)	−0.106 (1.428)	−0.755 (1.660)	.88	.087	.286
Subfoveal choroidal thickness (µm/y)	−1.811 (16.397)	−1.050 (18.019)	10.089 (20.120)	.90	.020	.108
Choroidal vascularity index (%/y)	0.003 (1.037)	0.004 (0.445)	0.033 (1.011)	.99	.879	.877
OCTA parameters
Foveal avascular zone area (mm²/y)	0.006 (0.014)	−0.002 (0.015)	0.000 (0.021)	.01	.21	.66
3-mm circle perfusion density (/y)^a	−0.001 (0.018)	−0.004 (0.008)	−0.013 (0.017)	.46	.007	.04
3-mm ring perfusion density (/y)^a	−0.001 (0.017)	−0.006 (0.008)	−0.013 (0.018)	.26	.008	.13
3-mm circle vessel density (mm^-1/y)	−0.123 (1.030)	−0.205 (0.466)	−0.761 (0.971)	.73	.02	.03
3-mm ring vessel density (mm^-1/y)	−0.093 (1.041)	−0.270 (0.416)	−0.735 (1.015)	.46	.02	.06
6-mm circle perfusion density (/y)^a	−0.004 (0.015)	−0.010 (0.020)	−0.011 (0.022)	.21	.13	.76
6-mm inner ring perfusion density (/y)^a	−0.001 (0.022)	−0.010 (0.027)	−0.008 (0.024)	.06	.21	.70
6-mm outer ring perfusion density (/y)^a	−0.005 (0.013)	−0.009 (0.018)	−0.012 (0.022)	.32	.14	.63
6-mm circle vessel density (mm^-1/y)	−0.140 (0.528)	−0.288 (0.706)	−0.425 (0.817)	.37	.07	.45
6-mm inner ring vessel density (mm^-1/y)	−0.048 (0.797)	−0.343 (1.025)	−0.304 (0.895)	.11	.18	.83
6-mm outer ring vessel density (mm^-1/y)	−0.168 (0.472)	−0.263 (0.599)	−0.475 (0.872)	.58	.07	.28
Average capillary perfusion density (%/y)	0.000 (0.005)	−0.005 (0.005)	−0.004 (0.007)	< .001	.11	.48
Average capillary flux index (/y)^b	0.002 (0.012)	0.003 (0.010)	−0.005 (0.011)	.89	.10	.19

Abbreviations: OCT, optical coherence tomography; OCTA, optical coherence tomography angiography.

Perfusion density is a unitless measure.

Flux is a unitless measure.

P value based on test of difference between means using generalized estimating equations to adjust for the correlation between eyes of the same subject.

Conclusions

Longitudinal studies are important when studying conditions that are affected by aging, such as mild cognitive impairment. Our study demonstrated that the velocity of decline in perfusion and vessel density in the 3-mm circles and rings as well as capillary perfusion density was significantly greater in the entire group with mild cognitive impairment compared with control participants. Subgroup analysis showed that amnestic patients with mild cognitive impairment had a faster velocity of decline in perfusion and vessel density in the 3-mm circles and rings than nonamnestic patients and control patients, and nonamnestic patients with mild cognitive impairment had a faster velocity of decline in capillary perfusion density compared with control patients.

Over an average follow-up of 28 months, there was a significantly greater velocity of decline in perfusion and vessel density in the 3-mm circle and ring in the entire group with mild cognitive impairment and the amnestic subgroup compared with control participants. Several cross-sectional studies using the RTVue-XR SD-OCT (Optovue, Inc.) in cognitively normal individuals have demonstrated modest age-related decreases in macular vessel density.²⁰ Rao et al²¹ reported decreases in macular vessel density of 0.2%/year in the 3-mm overlay. Jo et al²² found that parafoveal vessel density decreased by 0.051%/year in the 6-mm overlay. This distinction is important for differentiating age-related pathologic changes in retinal perfusion and vessel density. Although the absolute decrease in perfusion and vessel density in the entire group with mild cognitive impairment over our follow-up period is relatively small, the percent change, up to 2.6% reduction per year in all regions of the 6-mm circle, is greater than in cognitively normal adults, both in our study and previous research.^20–22 Given that literature on longitudinal OCTA changes in mild cognitive impairment is limited, this study may serve as a reference for distinguishing changes in retinal perfusion and vessel density attributed to normal aging from neurocognitive disease.

The entire cohort with mild cognitive impairment had a significantly greater decline in perfusion and vessel density compared with control participants. This faster decline may mirror microvascular loss in the brain, as CNS inflammation has been associated with impaired cognition.^23–31 The neuroangiogenesis hypothesis proposed by Ambrose postulates that reduced capillary density seen with aging may result from decreases in growth factors and cytokines.³⁰ As an embryologic derivative of the brain, the retina has been shown to manifest vascular changes associated with cognitive decline. Using ultra-widefield retinal imaging, increased rates of arteriolar and venular thinning with reduced vessel branching have been measured in patients with mild cognitive impairment.³²

Amnestic patients with mild cognitive impairment also had a significantly greater decline in perfusion and vessel density compared with nonamnestic patients. This is consistent with a study by Criscuolo et al^33,34 that used the Optovue Angiovue System (Optovue, Inc.) and reported reduced superficial capillary plexus vessel density in 19 amnestic patients with mild cognitive impairment after a 2-year follow-up. The more rapid decrease in perfusion and vessel density in amnestic patients with mild cognitive impairment is consistent with impaired memory. Velocity of change of perfusion and vessel density may be a strong indicator of memory decline and may serve as an adjunctive diagnostic tool to differentiate amnestic from nonamnestic mild cognitive impairment and to monitor disease progression.

Peripapillary capillary perfusion density decreased faster in the entire group with mild cognitive impairment compared with control participants, with subanalysis revealing significantly decreased velocity of change in average capillary perfusion density in nonamnestic patients vs control patients. Previously, Joseph et al³⁵ reported modest decreases in mean peripapillary capillary perfusion density with normal aging. However, Ma et al³⁶ reported no significant difference in rate of change of capillary perfusion density between APOE ε4 carriers and noncarriers over a 2-year period. There is no previous literature examining longitudinal capillary perfusion density changes in patients with mild cognitive impairment. It is unclear why only nonamnestic patients with mild cognitive impairment exhibited a significantly faster decline in capillary perfusion density compared with control participants. Peripapillary vascular metrics, including capillary perfusion density, may vary during the presumed inflammatory transition between early and late mild cognitive impairment, as described by Fan et al.’s³⁷ observation of higher microglial activation during early stages of mild cognitive impairment. Our observations of a significant decline in capillary perfusion density among nonamnestic patients with mild cognitive impairment and among the entire cohort with mild cognitive impairment over 28 months represents a novel finding and underscores the need for better understanding of the pathophysiology.

In this study, we found that cross-sectional results for perfusion and vessel density did not differ significantly between control participants, the nonamnestic and amnestic subgroups, and the entire group with mild cognitive impairment at baseline or follow-up, which is consistent with previous studies with similar cohort sizes.^15,38 However, FAZ area was smaller at baseline in amnestic patients with mild cognitive impairment compared with control participants and smaller in the entire group with mild cognitive impairment compared with control patients over our follow-up period. Moreover, nonamnestic patients with mild cognitive impairment demonstrated a decrease in FAZ area over time compared with an increase in control patients, which may be partly attributed to perifoveal vascular remodeling. Previous studies found that FAZ area is significantly larger in patients with mild cognitive impairment compared with control patients, which may be expected due to capillary dropout related to inflammatory microvascular injury and perivascular amyloid accumulation in those individuals at risk for Alzheimer disease.^38–40 Other studies demonstrated no significant FAZ area enlargement in patients with mild cognitive impairment or patients with preclinical Alzheimer disease.^13,41 Given our conflicting findings, longitudinal work with larger cohorts may shed further insight.

Both the entire group with mild cognitive impairment and the amnestic subgroup had a significantly lower choroidal vascular index at follow-up compared with control participants. This difference shows the dependence of cross-sectional studies on subject cohort, especially for age-dependent diseases.⁴² Compared with choroidal vascular index, subfoveal choroidal thickness may vary with multiple systemic factors, reducing its reliability.⁴³ There were no significant differences in subfoveal choroidal thickness at baseline or follow-up between the entire group with mild cognitive impairment, nor with the subgroups and control participants. However, the subfoveal choroidal thickness increased from baseline to follow-up in the entire group with mild cognitive impairment and the amnestic subgroup, while over time it decreased in control participants. Our observation of decreased choroidal vascular index may be attributed to choroidal remodeling in response to possible choroidal vascular insult associated with mild cognitive impairment. We speculate that choroidal vascularity is influenced by inflammatory changes in amnestic mild cognitive impairment, in which higher levels of C-reactive protein and IL-12p70 have been reported.^44,45 This may cause ischemic changes, resulting in vessel narrowing, closure of small vessels, and a decrease in the luminal area of the choroid. Ischemia may also lead to compensatory stromal thickening in the choroid, increasing structural support for the remaining vasculature and enhancing diffusion of nutrients. This stromal thickening may partly explain the observed increase in subfoveal choroidal thickness.

Our study has several limitations. First, the overall return rate of 45% was directly affected by the COVID-19 pandemic, during which study recruitment was paused for 6 months, after which several participants and caregivers maintained concerns about COVID-19 and exhibited caregiver fatigue. Follow-up compliance is an ongoing challenge with observational noninterventional trials even without these factors, particularly for individuals with cognitive impairment. Our study cohort was also comprised of more amnestic than nonamnestic patients at baseline. However, at follow-up, only 17% of amnestic patients returned, compared with 100% of nonamnestic patients and 44% of control participants. Inadequate follow-up among amnestic patients may reflect greater cognitive decline, limiting study compliance. This attrition may reduce the precision of subgroup analyses and emphasizes the importance of future studies with larger cohorts, although this may be an inherent limitation in longitudinal studies of individuals with progressive neurocognitive degeneration. Notably, amnestic mild cognitive impairment is more likely to progress to Alzheimer disease than the nonamnestic type.^46–48 Thus, it may be expected that a larger cohort of nonamnestic patients with mild cognitive impairment at follow-up would decrease the likelihood of significant findings. Yet, in the entire cohort of patients with mild cognitive impairment, the rate of decline in perfusion and vessel density was faster than control participants. In addition, of the 66 participants prospectively enrolled in our study, 23 were excluded due to baseline OCT images failing specified quality thresholds. Although consensus diagnoses of amnestic and nonamnestic mild cognitive impairment were not determined for the excluded participants, poor image quality may reflect greater cognitive or functional impairment, including difficulty following instructions, fixation instability, or head tremor. Even though this exclusion may introduce selection bias in our sample toward patients with relatively milder disease, it becomes even more compelling that the rate of decline over time was faster in the cohort of patients with mild cognitive impairment compared with control patients.

Although this is one of the largest longitudinal studies evaluating retinal (macular and peripapillary) and choroidal microvascular changes in patients with mild cognitive impairment, our sample size was modest, with a follow-up duration of just over 2 years, which is relatively short compared with the decades-long continuum of neurodegeneration. Over this shorter follow-up duration, however, we observed reductions of up to 2.6% per year in several parameters, suggesting these are sensitive metrics. Interpretability and comparison of OCTA measurements across different devices remains an unmet challenge and limits comparison across different cohorts. We also acknowledge that quantification of choroidal vascular index is challenging given the need for software to binarize and segment choroidal images as well as experienced raters to accurately grade image quality and region of interest. Mini-Mental State Examination scores may lack the sensitivity needed to clinically diagnose and evaluate progression of neurocognitive disease. Even so, we found significant reductions in retinal perfusion and vessel density over our follow-up period in patients with mild cognitive impairment despite no significant decline in Mini-Mental State Examination scores, suggesting that biomarkers for microvascular loss may help detect cognitive changes before measurable changes in Mini-Mental State Examination score. Other neurocognitive vehicles may offer greater precision in detecting cognitive changes; however, none may be as sensitive as retinal microvascular evaluation.

Future studies to stratify the severity of mild cognitive impairment and better characterize retinal changes across its spectrum would require significant funding to incorporate biomarkers including CSF biomarker concentrations and amyloid PET imaging, as well as APOE genotyping. Previous studies demonstrated that amnestic mild cognitive impairment is associated with higher rates of Aβ and tau pathology.^49
–51 Additionally, a previous longitudinal mouse model of tauopathy demonstrated significant decreases in vessel density in tau mice compared with wild-type controls.⁵² Future research could explore the relationship between retinal vascular changes and the presence of Aβ and tau by incorporating CSF and PET imaging in conjunction with OCT and OCTA. These tests were not performed in our study, which may be a confounding factor given that previous studies found increased rates of progression to dementia in patients with mild cognitive impairment who are biomarker positive.^49,51,53,54 Similarly, APOE genetic testing was available for only a small subset of our mild cognitive impairment cohort. Given that previous studies have linked APOE4 carrier status to increased rates of Aβ protein accumulation, future longitudinal OCT and OCTA studies should incorporate genetic testing to stratify risk for Alzheimer disease.^53,55

Longitudinal differences in OCT and OCTA parameters of individuals with mild cognitive impairment show that cross-sectional studies alone may not elucidate the changes observed over time in such an age-dependent disease. Compared with control participants, the velocity of decline in perfusion and vessel density over time was faster in the entire cohort of individuals with mild cognitive impairment (and greater in individuals with amnestic mild cognitive impairment). Our study also explores the relationship of the FAZ area and choroidal vascular index with mild cognitive impairment and warrants further investigation. Longitudinal changes in retinal and choroidal microvascular metrics, including perfusion density and vessel density, may serve as sensitive biomarkers of progression of mild cognitive impairment, potentially preceding measurable decline in Mini-Mental State Examination scores.

Supplemental Material

sj-docx-1-vrd-10.1177_24741264261429294 – Supplemental material for Comparison of Longitudinal Retinal and Choroidal Imaging Findings in Patients With Mild Cognitive Impairment and Cognitively Normal Control Participants

Supplemental material, sj-docx-1-vrd-10.1177_24741264261429294 for Comparison of Longitudinal Retinal and Choroidal Imaging Findings in Patients With Mild Cognitive Impairment and Cognitively Normal Control Participants by Angela Hemesath, Wufan Zhao, Cason B. Robbins, Anita Kundu, Ariana Allen, Suzanna Joseph, Justin P. Ma, Alice Haystead, Andrea C. Bozoki, Kim G. Johnson, Andy J. Liu, Sandra S. Stinnett, Xin Wei, Rupesh Agrawal, Dilraj S. Grewal and Sharon Fekrat in Journal of VitreoRetinal Diseases

Footnotes

Ethical Approval

This study (NCT03233646) was reviewed and approved by the Duke University Health System Institutional Review Board (Pro00111831). This study adhered to the tenets of the Declaration of Helsinki with regard to enrollment of human subjects.

Statement of Informed Consent

Written informed consent was obtained from all participants and/or their designated legally authorized representative prior to study enrollment.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported in part by the Alzheimer’s Drug Discovery Foundation.

ORCID iDs

Wufan Zhao

Ariana Allen

Andrea C. Bozoki

Kim G. Johnson

Dilraj S. Grewal

Sharon Fekrat

Data Availability Statement

The datasets generated and/or analyzed during the current study are not publicly available due to patient confidentiality but are available from the corresponding author upon reasonable request.

Supplemental Material

Supplemental material is available online with this article.

References

Csukly

Siraly

Fodor

, et al. The differentiation of amnestic type MCI from the non-amnestic types by structural MRI. Front Aging Neurosci. 2016;8:52. doi:10.3389/fnagi.2016.00052

Grundman

Petersen

Ferris

, et al. Mild cognitive impairment can be distinguished from Alzheimer disease and normal aging for clinical trials. Arch Neurol. 2004;61(1):59-66. doi:10.1001/archneur.61.1.59

Schmidtke

Hermeneit

High rate of conversion to Alzheimer’s disease in a cohort of amnestic MCI patients. Int Psychogeriatr. 2008;20(1):96-108. doi:10.1017/S1041610207005509

UCI MIND. Mild Cognitive Impairment. 2013. Accessed May 5, 2023. https://mind.uci.edu/dementia/mild-cognitive-impairment/

Ferman

Smith

Kantarci

, et al. Nonamnestic mild cognitive impairment progresses to dementia with Lewy bodies. Neurology. 2013;81(23):2032-2038. doi:10.1212/01.wnl.0000436942.55281.47

Ward

Arrighi

Michels

Cedarbaum

JM.

Mild cognitive impairment: disparity of incidence and prevalence estimates. Alzheimers Dement. 2012;8(1):14-21. doi:10.1016/j.jalz.2011.01.002

Albert

DeKosky

Dickson

, et al. The diagnosis of mild cognitive impairment due to Alzheimer’s disease: recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement. 2011;7(3):270-279. doi:10.1016/j.jalz.2011.03.008

Arevalo-Rodriguez

Smailagic

Roque-Figuls

, et al. Mini-Mental State Examination (MMSE) for the early detection of dementia in people with mild cognitive impairment (MCI). Cochrane Database Syst Rev. 2021;7(7):CD010783. doi:10.1002/14651858.CD010783.pub3

Bondi

Smith

GE.

Mild cognitive impairment: a concept and diagnostic entity in need of input from neuropsychology. J Int Neuropsychol Soc. 2014;20(2):129-134. doi:10.1017/S1355617714000010

10.

Dukart

Mueller

Villringer

, et al. Relationship between imaging biomarkers, age, progression and symptom severity in Alzheimer’s disease. Neuroimage Clin. 2013;3:84-94. doi:10.1016/j.nicl.2013.07.005

11.

DeMarshall

Nagele

Sarkar

, et al. Detection of Alzheimer’s disease at mild cognitive impairment and disease progression using autoantibodies as blood-based biomarkers. Alzheimers Dement (Amst). 2016;3:51-62. doi:10.1016/j.dadm.2016.03.002

12.

Wang

Zhao

Tao

, 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. 2020;12:572484. doi:10.3389/fnagi.2020.572484

13.

Robbins

Akrobetu

, et al. Assessment of retinal microvascular alterations in individuals with amnestic and nonamnestic mild cognitive impairment using optical coherence tomography angiography. Retina. 2022;42(7):1338-1346. doi:10.1097/IAE.0000000000003458

14.

Robbins

Grewal

Thompson

, et al. Choroidal structural analysis in Alzheimer disease, mild cognitive impairment, and cognitively healthy controls. Am J Ophthalmol. 2021;223:359-367. doi:10.1016/j.ajo.2020.09.049

15.

Yoon

Grewal

Thompson

, et al. Retinal microvascular and neurodegenerative changes in Alzheimer’s disease and mild cognitive impairment compared with control participants. Ophthalmol Retina. 2019;3(6):489-499. doi:10.1016/j.oret.2019.02.002

16.

O’Bryant

Humphreys

Smith

, et al. Detecting dementia with the mini-mental state examination in highly educated individuals. Arch Neurol. 2008;65(7):963-967. doi:10.1001/archneur.65.7.963

17.

Zhang

Qiu

Qian

, et al. Determining appropriate screening tools and cutoffs for cognitive impairment in the chinese elderly. Front Psychiatry. 2021;12:773281. doi:10.3389/fpsyt.2021.773281

18.

Betzler

Ding

Wei

, et al. Choroidal vascularity index: a step towards software as a medical device. Br J Ophthalmol. 2022;106(2):149-155. doi:10.1136/bjophthalmol-2021-318782

19.

Agrawal

Gupta

Tan

Cheung

Wong

Cheng

CY.

Choroidal vascularity index as a measure of vascular status of the choroid: measurements in healthy eyes from a population-based study. Sci Rep. 2016;6:21090. doi:10.1038/srep21090

20.

Jiang

Wang

, et al. Macular perfusion in healthy Chinese: an optical coherence tomography angiogram study. Invest Ophthalmol Vis Sci. 2015;56(5):3212-3217. doi:10.1167/iovs.14-16270

21.

Rao

Pradhan

Weinreb

, et al. Determinants of peripapillary and macular vessel densities measured by optical coherence tomography angiography in normal eyes. J Glaucoma. 2017;26(5):491-497. doi:10.1097/IJG.0000000000000655

22.

Sung

Shin

JW.

Effects of age on peripapillary and macular vessel density determined using optical coherence tomography angiography in healthy eyes. Invest Ophthalmol Vis Sci. 2019;60(10):3492-3498. doi:10.1167/iovs.19-26848

23.

Hunter

Kwan

Malek-Ahmadi

, et al. Morphological and pathological evolution of the brain microcirculation in aging and Alzheimer’s disease. PLoS One. 2012;7(5):e36893. doi:10.1371/journal.pone.0036893

24.

Brown

Thore

CR.

Review: cerebral microvascular pathology in ageing and neurodegeneration. Neuropathol Appl Neurobiol. 2011;37(1):56-74. doi:10.1111/j.1365-2990.2010.01139.x

25.

Buee

Hof

Delacourte

Brain microvascular changes in Alzheimer’s disease and other dementias. Ann N Y Acad Sci. 1997;826:7-24. doi:10.1111/j.1749-6632.1997.tb48457.x

26.

Cruz

Hernandez JC

Bracko

Kersbergen

, et al. Neutrophil adhesion in brain capillaries reduces cortical blood flow and impairs memory function in Alzheimer’s disease mouse models. Nat Neurosci. 2019;22(3):413-420. doi:10.1038/s41593-018-0329-4

27.

Nortley

Korte

Izquierdo

, et al. Amyloid β oligomers constrict human capillaries in Alzheimer’s disease via signaling to pericytes. Science. 2019;365(6450):eaav9518. doi:10.1126/science.aav9518

28.

Wang

Tan

Role of pro-inflammatory cytokines released from microglia in Alzheimer’s disease. Ann Transl Med. 2015;3(10):136. doi:10.3978/j.issn.2305-5839.2015.03.49

29.

Ostergaard

Jespersen

Engedahl

, et al. Capillary dysfunction: its detection and causative role in dementias and stroke. Curr Neurol Neurosci Rep. 2015;15(6):37. doi:10.1007/s11910-015-0557-x

30.

Ambrose

CT.

Pro-angiogenesis therapy and aging: a mini-review. Gerontology. 2017;63(5):393-400. doi:10.1159/000477402

31.

Steinman

Sun

Feng

ZP.

Microvascular alterations in Alzheimer’s disease. Front Cell Neurosci. 2020;14:618986. doi:10.3389/fncel.2020.618986

32.

Pead

Thompson

Grewal

, et al. Retinal vascular changes in Alzheimer’s dementia and mild cognitive impairment: a pilot study using ultra-widefield imaging. Transl Vis Sci Technol. 2023;12(1):13. doi:10.1167/tvst.12.1.13

33.

Chiara

Gilda

Daniela

, et al. A two-year longitudinal study of retinal vascular impairment in patients with amnestic mild cognitive impairment. Front Aging Neurosci. 2022;14:993621. doi:10.3389/fnagi.2022.993621

34.

Criscuolo

Cennamo

Montorio

, et al. Corrigendum: a two-year longitudinal study of retinal vascular impairment in patients with amnestic mild cognitive impairment. Front Aging Neurosci. 2023;15:1197979. doi:10.3389/fnagi.2023.1197979

35.

Joseph

Woo

Robbins

, et al. Longitudinal assessment of peripapillary microvasculature using optical coherence tomography angiography in cognitively normal adults. J Vitreoretin Dis. 2024;8(5):571-576. doi:10.1177/24741264241263167

36.

Robbins

Lee

, et al. Longitudinal analysis of the retina and choroid in cognitively normal individuals at higher genetic risk of Alzheimer disease. Ophthalmol Retina. 2022;6(7):607-619. doi:10.1016/j.oret.2022.03.001

37.

Fan

Brooks

Okello

Edison

An early and late peak in microglial activation in Alzheimer’s disease trajectory. Brain. 2017;140(3):792-803. doi:10.1093/brain/aww349

38.

Criscuolo

Cennamo

Montorio

, et al. Assessment of retinal vascular network in amnestic mild cognitive impairment by optical coherence tomography angiography. PLoS One. 2020;15(6):e0233975. doi:10.1371/journal.pone.0233975

39.

Zhang

Azhati

Liu

Retinal microvascular attenuation in mental cognitive impairment and Alzheimer’s disease by optical coherence tomography angiography. Acta Ophthalmol. 2020;98(6):e781-e787. doi:10.1111/aos.14381

40.

Koronyo-Hamaoui

Koronyo

Ljubimov

, et al. Identification of amyloid plaques in retinas from Alzheimer’s patients and noninvasive in vivo optical imaging of retinal plaques in a mouse model. Neuroimage. 2011;(54 Suppl 1):S204-S217. doi:10.1016/j.neuroimage.2010.06.020

41.

Curro

van Nispen

RMA

den Braber

, et al. Longitudinal assessment of retinal microvasculature in preclinical Alzheimer’s disease. Invest Ophthalmol Vis Sci. 2024;65(12):2. doi:10.1167/iovs.65.12.2

42.

Teti

DM.

Handbook of Research Methods in Developmental Science. John Wiley & Sons; 2008.

43.

Aksoy

Asena

Tekindal

Ayvazoglu Soy

Yilmaz

Haberal

Choroidal vascularity index and choroidal thickness changes following renal transplantation. Turk J Ophthalmol. 2023;53(2):97-104. doi:10.4274/tjo.galenos.2022.02489

44.

Karim

Hopkins

Purandare

, et al. Peripheral inflammatory markers in amnestic mild cognitive impairment. Int J Geriatr Psychiatry. 2014;29(3):221-226. doi:10.1002/gps.3988

45.

Trollor

Smith

Baune

, et al. Systemic inflammation is associated with MCI and its subtypes: the Sydney Memory and Aging Study. Dement Geriatr Cogn Disord. 2010;30(6):569-578. doi:10.1159/000322092

46.

Jungwirth

Zehetmayer

Hinterberger

Tragl

Fischer

The validity of amnestic MCI and non-amnestic MCI at age 75 in the prediction of Alzheimer’s dementia and vascular dementia. Int Psychogeriatr. 2012;24(6):959-966. doi:10.1017/S1041610211002870

47.

Yaffe

Petersen

Lindquist

Kramer

Miller

Subtype of mild cognitive impairment and progression to dementia and death. Dement Geriatr Cogn Disord. 2006;22(4):312-319. doi:10.1159/000095427

48.

Vos

van Rossum

Verhey

, et al. Prediction of Alzheimer disease in subjects with amnestic and nonamnestic MCI. Neurology. 2013;80(12):1124-1132. doi:10.1212/WNL.0b013e318288690c

49.

Okello

Koivunen

Edison

, et al. Conversion of amyloid positive and negative MCI to AD over 3 years: an 11C-PIB PET study. Neurology. 2009;73(10):754-760. doi:10.1212/WNL.0b013e3181b23564

50.

Fleisher

Sowell

Taylor

, et al. Clinical predictors of progression to Alzheimer disease in amnestic mild cognitive impairment. Neurology. 2007;68(19):1588-1595. doi:10.1212/01.wnl.0000258542.58725.4c

51.

Hampel

Teipel

Fuchsberger

, et al. Value of CSF β-amyloid1-42 and tau as predictors of Alzheimer’s disease in patients with mild cognitive impairment. Mol Psychiatry. 2004;9(7):705-710. doi:10.1038/sj.mp.4001473

52.

Buscho

Palacios

Xia

, et al. Longitudinal characterization of retinal vasculature alterations with optical coherence tomography angiography in a mouse model of tauopathy. Exp Eye Res. 2022;224:109240. doi:10.1016/j.exer.2022.109240

53.

Mares

Kumar

Sharma

, et al. APOE ε4–associated heterogeneity of neuroimaging biomarkers across the Alzheimer’s disease continuum. Alzheimers Dement. 2024;21(1):e13977. doi:10.1002/alz.14392

54.

Bucci

Chiotis

Nordberg

; Alzheimer’s Disease Neuroimaging Initiative. Alzheimer’s disease profiled by fluid and imaging markers: tau PET best predicts cognitive decline. Mol Psychiatry. 2021;26(10):5888-5898. doi:10.1038/s41380-021-01263-2

55.

DiBattista

Heinsinger

Rebeck

GW.

Alzheimer’s disease genetic risk factor APOE-ε4 also affects normal brain function. Curr Alzheimer Res. 2016;13(11):1200-1207. doi:10.2174/1567205013666160401115127