Biomarkers for Alzheimer's disease are upregulated in patients with diabetic retinopathy

Abstract

Background

Diabetes has been linked to increased prevalence of dementia, but the link between diabetic retinopathy (DR) and Alzheimer's disease (AD) remains unclear.

Objective

This study aimed to evaluate potential associations between DR and AD-related protein biomarkers in plasma and ocular fluid.

Methods

A prospective, cross-sectional study collected human blood, vitreous, aqueous, and tear samples and measured amyloid-β (Aβ₄₀, Aβ₄₂), total-tau (t-tau), phosphorylated-tau (ptau181), glial fibrillary acidic protein (GFAP), and neurofilament light chain (NfL) by digital immunoassays.

Results

The study included 79 eyes (79 patients) [41 females (59.4%); mean (SD) age 57.1 (12.2) years] of which DR was present in 44 (55.7%). All six biomarkers were significantly higher in plasma in participants with DR compared to those without DR [Aβ₄₀ p = 0.002, Aβ₄₂ p = 0.002, t-tau = 0.013, ptau181 p = 0.005, GFAP p = 0.010, and NfL p < 0.001]. Within vitreous, DR participants had significantly elevated t-tau (p = 0.002), ptau181 (p = 0.049), and NfL (p = 0.006); and within aqueous, higher NfL (p = < 0.001). Neuropsychological testing scores were lower in participants with DR than those without but did not reach statistical significance (Montreal-Cognitive-Assessment: p = 0.070; Mini-Mental-State-Exam: p = 0.057).

Conclusions

This study showed significant increases of AD associated protein biomarkers in plasma, vitreous, and aqueous in patients with DR. These results support a potential biological link between DR and AD pathology and suggest that DR, which tends to occur in younger individuals, may be a predictive factor for AD.

Keywords

Alzheimer’s disease amyloid-β diabetic retinopathy diabetes mellitus glial fibrillary acidic protein neurofilament light chain ptau181 t-tau

Introduction

Alzheimer's disease (AD) is the 6^th leading cause of death and comprises 60–80% of all cases of dementia.¹ More than 46 million people have dementia world-wide and by 2050 the number affected will likely increase by three-fold,² creating significant impact on personal caregivers, public health, and healthcare costs.¹ Diabetes mellitus (DM) is predicted to affect 643 million individuals by 2030³ and has been linked to dementia and AD in both cross sectional and longitudinal studies.^4–12 Those with type 2 diabetes have almost twice the risk of dementia, including both AD and vascular dementia, compared to those without DM.⁹ Pathologically, DM is linked to vascular dementia^13,14; however, the association of AD type dementia and diabetes is not yet fully understood. Studies have shown that common mechanisms occur in both DM and AD, such as ischemic injury,¹⁵ blood-brain barrier disruption,^16,17 and abnormal brain glucose utilization.^18–21

Recently, protein biomarkers have become increasingly important in understanding the pathogenesis of AD.²² Biomarkers such as amyloid-β (Aβ), tau, and others are currently used in cognitive aging research, therapeutic clinical trials, and are showing promise for early diagnosis and screening for AD. Since neuropathological changes in AD occur decades before the onset of symptoms, the early presence of biomarkers serves as a proxy for AD pathological changes in living persons.²³

Research has shown that patients with eye disease have a higher risk of AD development,²⁴ and recent studies indicate that the presence of diabetic retinopathy (DR) may be a risk factor for the development of AD dementia and cognitive impairment.^25,26 Our group studies the presence of AD-associated protein biomarkers in eye fluid in those with eye diseases.^27–32 In prior studies we have linked biomarkers in vitreous humor to cognitive status in living patients and to pathological findings in brain specimens, and we have correlated biomarkers in different ocular fluid chambers to plasma. In this current study, we investigated the association between diabetic retinopathy and AD by examining blood, vitreous humor, aqueous humor, and tear fluid for AD-associated biomarkers Aβ, t-tau, ptau181, glial fibrillary acidic protein (GFAP), and neurofilament light chain (NfL) and comparing findings between individuals with and without diabetic retinopathy.

Methods

Study design

This is a prospective, cross-sectional study approved by the Boston University Medical Center (BUMC) Institutional Review Board (IRB) (study reference number H-37370, principal investigator MLS) and carried out in accordance with the ethical standards of the Committee on Human Experimentation of BUMC, diversity, equity and inclusion of BUMC and 1964 the Declaration of Helsinki.

Inclusion criteria comprised adults who underwent pars plana vitrectomy for ophthalmic conditions such as rhegmatogenous retinal detachment, macular hole, epiretinal membrane, and complications of DR such as vitreous hemorrhage and tractional retinal detachment. DR severity was classified by experienced retina specialists (MLS, SN, XC, NHS) based on selection of the ICD-10 codes by the clinicians and confirmed by review of the clinical examination findings and review of multimodal ocular imaging. Written informed consent was obtained in English or Spanish language from all patients who enrolled in the study, and no individuals were excluded due to existing ocular or medical comorbidities or based on participants’ sex or race. Electronic medical records of the participants were reviewed to collect demographic, ocular and systemic information in a standardized format. Exclusion criteria included those who had any history of having their vitreous humor removed from prior vitrectomy, as well as those whose primary language was not English or Spanish, due to availability of IRB-approved consent forms only in these two languages.

Biospecimen collection

Vitreous and aqueous humor samples were collected during the surgical procedure. Approximately 0.05–0.15 mL of aqueous humor were collected at the beginning of each procedure through the limbus with a paracentesis using a supersharp blade to access the anterior chamber. This was immediately followed by aspiration of the undiluted aqueous fluid with a 27-gauge cannula attached to an empty syringe. The anterior chamber was immediately replaced with balanced salt solution to repressurize the anterior chamber. For collection of undiluted vitreous sample, 0.5–1.5 mL of fluid was aspirated at the start of each pars plana vitrectomy via the vitrectomy probe, and once the specimen was fully collected, immediate infusion of balanced salt solution commenced to replace the removed vitreous and repressurize the globe. Tear fluid and blood samples were collected on separate clinical visits from all study participants within 1–2 months of their surgery. Tears were collected from both eyes of the participants via Schirmer's tear strips (Eye Care and Cure Corp., AZ). In addition, 18 mL of whole blood was drawn from each patient into EDTA-treated purple top tubes. All biospecimens were collected, centrifuged and prepared for analysis at the Molecular Genetics Core Laboratory at BUMC.

Immunoassay measurements

Ocular fluids and plasma samples were tested for Aβ₄₀, Aβ₄₂, phosphorylated-tau (ptau181), total tau (t-tau), GFAP, and NfL. Briefly, Aβ₄₂, Aβ₄₀, and t-tau relative concentrations were measured using Neurology 3-Plex A Assay (#101995, Quanterix, Billerica, MA) with a 4-fold dilution on HD-X analyzer (Quanterix). For ptau181, levels were measured using ptau181 V2 Advantage kit (#103714, Quanterix, MA) with a 4-fold dilution. Concentrations of GFAP and NfL were measured using the combined Neurology 2-Plex B assay (#103520, Quanterix, MA) with a 4-fold dilution for vitreous and plasma and an 8-fold dilution for aqueous samples. All samples were processed per manufacturer's instructions in the immunoassay kits.

Based on results from our prior studies, as well as our preference to target biomarkers more specific to AD, we prioritized the 6 biomarker assays in the following order: 1) ptau181, 2) t-tau, Aβ₄₀ and Aβ₄₂ (all 3 in the same kit); and 3) GFAP and NfL (both in same kit). The tear fluid volume was not sufficient to conduct the immunoassays for GFAP and NfL.

Neuropsychological testing

All enrolled participants underwent neuropsychological testing within 180 days from collection of fluid specimens. This timeframe was recommended by collaborators from the Boston University Alzheimer's Disease and Research Center, who additionally trained research staff to conduct the examinations. Testing included Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA).

Statistical analysis

Statistical analyses were performed by using statistical software (SAS, version 9.4; SAS Institute Inc). Some biomarkers underwent log transformation after visual inspection of residuals to reduce the skewness of the data. Analysis of covariance was used to assess the correlation between the biomarkers and DR distribution as well as between neuropsychological scoring and DR distribution. Multivariate regression models were adjusted for age, gender, Apolipoprotein E ε4 (APOE4) allele genotype, and eye disease [cataract, glaucoma, and age-related macular degeneration (AMD)], and the Tukey method was applied to control for multiple comparisons. p-values ≤ 0.05 were considered significant and ≤ 0.1 as trend.

Two models of statistical analyses were run. The first ‘Model 1’ included all patients with DR compared to those without DR based on clinical exam (DR versus No-DR) The No-DR comparison group included patients without diabetes (No-DM) as well as those with DM and without clinical evidence of DR (DM with No-DR). The second ‘Model 2’ was a sensitivity analysis that compared those with DR to No-DM, excluding those with DM and No-DR, in order to reduce any potential confounding effects that participants with diabetes without clinical retinopathy may have had on the analysis. All statistical analyses were performed on the available data, with missing values treated as absent without imputation or correction.

Results

In total, 79 eyes of 79 patients (1 eye per patient) were enrolled in the study. Participants’ demographic and clinical information can be found in Table 1. Of the 79 patients who consented and underwent sample collection, we ultimately collected 77 vitreous humor samples, 67 blood samples, 56 tear fluid samples, and 51 aqueous humor samples. Missing samples were due to challenges with specimen collection, patients refusing blood draws, and inadequate sample volume (for tear fluid and aqueous fluid). The racial and ethnic breakdown of our study population is a close representation of the patient population typically seen at the eye clinic at BUMC, as the largest safety-net hospital in New England, with 31.9% of patients being Black or African American, 20.3% White, and 46.2% Hispanic.

Table 1.

Participants’ demographic and ocular characteristics.

Parameter	DR n (%) or Mean ± SD	DM with No-DR n (%) or Mean ± SD	No-DM n (%) or Mean ± SD
Participants/Eyes	44 (55.7)	9 (11.4)	26 (32.9)
Age (y)	55.3 ± 10.3	63.0 ± 11.2	58.2 ± 14.9
Gender (female)	26 (63.4)	5 (55.6)	10 (52.6)
Ethnicity (Hispanic)	21(48.8)	3 (33.3)	12 (46.2)
Race
Asian	2 (4.5)	0	1 (3.8)
Black or African American	13 (29.5)	4 (44.4)	5 (19.2)
Native Hawaiian	1 (2.3)	0	0
White	8 (18.2)	2 (22.2)	4 (21.5)
Unknown/Other	3 (6.8)	0	7 (26.9)
Diabetes Type (Type 2)	34 (82.9)	9 (100)	n/a
Diabetes Duration (y)	6.8 ± 7.0	6.4 ± 4.0	n/a
HbA1c (%)	8.4 ± 8.0	7.1 ± 6.9	n/a
Nonproliferative DR (NPDR)	5 (12.2)	n/a	n/a
Mild NPDR	1 (2.4)	n/a	n/a
Moderate NPDR	2 (4.9)	n/a	n/a
Severe NPDR	2 (4.9)	n/a	n/a
Proliferative DR	36 (87.8)	n/a	n/a
History of Cataract	11 (25.0)	2 (22.2)	6 (23.1)
History of Age-related Macular Degeneration	41 (93.2)	9 (100)	19 (73.1)
History of Glaucoma	27 (61.4)	7 (77.8)	16 (61.5)
Indication for Surgery
Epiretinal Membrane	1 (2.3)	1 (11.1)	7 (26.9)
Macular Hole	1 (2.3)	2 (22.2)	7 (26.9)
Rhegmatogenous Retinal Detachment	6 (13.6)	3 (33.3)	8 (30.8)
Tractional Retinal Detachment*	8 (18.2)	0	1 (3.9)
Vitreous Hemorrhage*	26 (59.1)	1 (11.1)	1 (3.9)
Other	2 (4.6)	2 (22.2)	2 (7.7)

DM: diabetes mellitus; DR: diabetic retinopathy; SD: standard deviation; n refers to the sample size; % refers to the percent of the sample size; * Tractional retinal detachment and vitreous hemorrhage are common complications of advanced diabetic retinopathy.

A total of 53 (67.1%) participants had diabetes while 26 did not have diabetes (No-DM). Clinically-evident retinopathy was present in 44 (55.7%) of the 53 patients with diabetes (Table 1), and there were 9 participants with DM with No-DR. In the model 1 analysis, which compared 44 patients with DR and 35 patients without DR, biomarkers were detected and measured as shown in Table 2. All six biomarkers had significantly higher concentrations in plasma in those with DR when compared with those without DR (Table 3), after adjusting for the covariates of age, gender, APOE4 allele number, eye disease, and additionally correcting for multiple comparisons. Moreover, eyes with DR had significantly higher levels of ptau181, t-tau, and NfL in vitreous humor, and NfL was found to be significantly higher in aqueous humor (Table 3). Of the four biomarkers tested in tear fluid (Aβ₄₀, Aβ₄₂, ptau181, and t-tau), none were significantly different with DR. Complete MMSE testing was performed in 57 (82.3%) and MoCA exam in 63 (79.7%) participants. In multivariate regression analysis, we found that both MMSE and MoCA scores were lower in participants with DR compared to patients without DR, which trended toward but did not achieve statistical significance (Table 4).

Table 2.

Overview of the biomarkers descriptive data for the total cohort.

	Aβ₄₀	Aβ₄₂	t-tau	ptau181	GFAP	NfL
Plasma (pg/mL)
Mean (SD)	373 (324)	14.7 (12.3)	5.97 (4.15)	1.02 (1.19)	120 (74.3)	52.5 (74.7)
Median [Min, Max]	267 [125, 1850]	10.8 [5.02, 69.6]	4.55 [2.01, 26.8]	0.609 [0.172, 7.17]	99.4 [20.2, 409]	21.5 [3.85, 346]
IQR [Q1, Q3]	125 [217, 342]	5.45 [8.44, 13.9]	3.24 [3.47, 6.71]	0.488 [0.439, 0.927]	77.4 [71.7, 149]	34.1 [14.3, 48.4]
Missing n (%)	12 (15.2%)	12 (15.2%)	12 (15.2%)	12 (15.2%)	12 (15.2%)	12 (15.2%)
Vitreous (pg/mL)
Mean (SD)	606 (279)	31.6 (16.3)	40.3 (77.7)	12.6 (10.5)	17500 (21300)	419 (872)
Median [Min, Max]	603 [0, 1220]	29.6 [0, 81.2]	20.9 [0, 639]	9.84 [0.797, 47.5]	8550 [0, 116000]	120 [0, 5870]
IQR [Q1, Q3]	345 [440, 785]	20.2 [21.8, 42.0]	29.0 [11.6, 40.6]	11.5 [5.23, 16.7]	20300 [3830, 24100]	319 [28.4, 347]
Missing n (%)	2 (2.5%)	2 (2.5%)	2 (2.5%)	2 (2.5%)	1 (1.3%)	1 (1.3%)
Aqueous (pg/mL)
Mean (SD)	134 (134)	5.54 (4.72)	20.0 (33.7)	7.92 (11.1)	2350 (4320)	66.3 (177)
Median [Min, Max]	86.1 [10.8, 539]	4.07 [0.450, 20.5]	8.35 [1.54, 208]	4.66 [0.580, 62.1]	868 [18.9, 20800]	16.5 [0.481, 1070]
IQR [Q1, Q3]	157 [35.2, 192]	5.72 [2.03, 7.75]	14.8 [5.50, 20.3]	5.47 [2.61, 8.08]	1470 [200, 1670]	47.9 [6.06, 53.9]
Missing n (%)	34 (43.0%)	34 (43.0%)	34 (43.0%)	28 (35.4%)	38 (48.1%)	38 (48.1%)
Tears (pg/mL)
Mean (SD)	2.90 (3.40)	0.459 (0.603)	128 (180)	32.5 (51.3)	n/a	n/a
Median [Min, Max]	2.22 [0, 11.8]	0.0979 [0, 2.23]	112 [0, 819]	12.4 [0, 295]	n/a	n/a
IQR [Q1, Q3]	4.58 [0, 4.58]	0.718 [0, 0.718]	180 [0.589, 180]	43.3 [0.479, 43.8]	n/a	n/a
Missing n (%)	53 (67.1%)	53 (67.1%)	53 (67.1%)	25 (31.6%)	79 (100%)	79 (100%)

Aβ: amyloid-beta; t-tau: total tau; ptau81: phosphorylated tau 181; GFAP: glial fibrillary acidic protein; NfL: neurofilament light chain; SD: standard deviation; IQR: interquartile range; Q1: first quartile; Q3: third quartile.

Table 3.

Regression analysis of biomarkers in each biospecimen in Model 1, adjusted for the covariates of age, gender, APOE4 allele number, eye disease, and multiple comparisons.

Model 1: DR versus No-DR
	Aβ40		Aβ₄₂		t-tau		ptau181		GFAP		NfL
	Estimate (95% CI)	p	Estimate (95% CI)	p	Estimate (95% CI)	p	Estimate (95% CI)	p	Estimate (95% CI)	p	Estimate (95% CI)	p
Plasma	0.5 (0.2, 0.7)	0.002*	0.5 (0.2, 0.7)	0.002*	0.3 (0.1, 0.6)	0.013*	0.7 (0.4,1.1)	0.005*	0.3 (0.001, 0.6)	0.01*	1.4 (1.0, 1.7)	<0.0001*
Vitreous	−80.1 (−207.3, 47.0)	0.18	−3.9 (-, 3.6)	0.19	0.9 (0.5, 1.3)	0.002*	0.4, (0.03, 0.8)	0.049*	0.7(0.03, 1.3)	0.09	1.3 (0.5, 2.0)	0.006*
Aqueous	0.1(−0.5, 0.8)	0.75	0.1 (0.4, 0.6)	0.91	0.1 (−0.5, 0.7)	0.57	0.1 (0.4, 0.7)	0.78	0.4 (−0.7, 1.4)	0.49	1.3 (0.3, 2.3)	0.0002*
Tears	1.5 (−1.3, 4.3)	0.77	0.3 (0.2, 0.8)	0.46	67.8 (−82.2, 217.8)	0.98	16.2 (11.8, 44.1)	0.49	-	-	-	-

Aβ: amyloid-β; t-tau: total tau; ptau81: phosphorylated tau 181; GFAP: glial fibrillary acidic protein; NfL: neurofilament light chain; CI: confidence intervals; “*”: statistical significance.

Eyes with diabetic retinopathy (DR) against eyes without DR. Those without DR included those without diabetes (No-DM) and those with diabetes but without DR (DM with No-DR).

Table 4.

Neuropsychological testing regression analysis for Model 1 and Model 2.

	Model 1: DR versus No-DR		Model 2: DR versus No-DM
	Estimate (95% CI)	p	Estimate (95% CI)	p
MoCA	2.5 (−0.02, 4.9)	0.07	1.7 (−1.1, 4.5)	0.38
MMSE	1.8 (−0.2, 3.8)	0.057	1.6 (−0.5, 3.8)	0.059

DM: diabetes mellitus; DR: diabetic retinopathy; MoCA: Montreal cognitive assessment; MMSE: mini mental state examination; CI: confidence intervals; P values were adjusted for the covariates of age, gender, APOE4 allele number, eye disease, and multiple comparisons.

In the model 2 sensitivity analysis, we compared 44 patients with DR to 26 patients without DM (No-DM); this model excluded 9 participants with DM with No-DR (Table 5). In regression analyses, 5 of the 6 biomarkers, including Aβ₄₀, Aβ₄₂, t-tau, ptau181, and NfL were significantly higher in plasma in participants with DR when compared with those with No-DM (Table 5). Moreover, in vitreous samples, eyes with DR had significantly higher levels of t-tau, ptau181, and NfL than eyes with No DM (Table 5). In regression analyses for neuropsychological testing, MMSE scores showed a similar trend with reduced scores in those with DR compared to those with No-DM (Table 4).

Table 5.

Regression analysis of biomarkers in each biospecimen in the Model 2 sensitivity analysis, eyes with DR against eyes without diabetes (No-DM). Those with diabetes without DR were removed from this analysis.

Model 2: DR versus No-DM
	Aβ₄₀		Aβ₄₂		t-tau		ptau181		GFAP		NfL
	Estimate (95% CI)	p	Estimate (95% CI)	p	Estimate (95% CI)	p	Estimate (95% CI)	p	Estimate (95% CI)	p	Estimate (95% CI)	p
Plasma	0.5 (0.2, 0.8)	0.009*	0.5 (0.2, 0.7)	0.009*	0.3 (0.06, 0.6)	0.06	0.8 (0.4, 1.1)	0.01*	0.2 (0.1, 0.5)	0.14	1.4 (0.9, 1.8)	<0.0001*
Vitreous	−49.9 (−185.5, 85.7)	0.33	−2.3 (−10.1, 5.5)	0.33	0.9 (0.5, 1.3)	0.007*	0.5 (0.07, 1.0)	0.049*	0.6 (−0.05, 1.2)	0.31	1.1 (0.3, 1.9)	0.008*
Aqueous	0.1(−0.6, 0.8)	0.56	0.09 (−0.5, 0.6)	0.73	0.3 (−0.4, 0.9)	0.87	0.3 (−0.3, 0.8)	0.45	0.2 (−0.9, 1.3)	0.99	1.2 (0.1, 2.5)	0.19
Tears	1.2 (−1.8, 4.1)	0.83	0.2 (−0.3, 0.8)	0.53	53.6 (−105.8, 212.9)	0.93	15.9 (−16.6, 48.3)	0.60	-	-	-	-

Aβ: amyloid-β; t-tau: total tau; ptau81: phosphorylated tau 181; GFAP: glial fibrillary acidic protein; NfL: neurofilament light chain; CI: confidence intervals; “*”: statistical significance; P values were adjusted for the covariates of age, gender, APOE4 allele number, eye disease, and multiple comparisons.

Discussion

One of the goals of AD research is the development of a low cost, minimally-invasive test that can diagnose AD in early, pre-symptomatic stages when current therapies have more meaningful effect in delaying progression. In 2016, the A/T/N classification scheme was proposed by Jack et. al to categorize major AD associated biomarkers into a straightforward system: “A” for amyloid-β biomarker (amyloid on PET scans or Aβ₄₂ in cerebrospinal fluid (CSF)); “T” for tau biomarker (CSF ptau, or tau PET); and “N” for neurodegeneration or neuronal injury (MRI atrophy, CSF t-tau, or NfL).³³ Research is showing that the presence of these biomarkers in serum and plasma, while not yet ready for clinical implementation, show promise as a potential diagnostic blood test and may soon serve as a diagnostic standard for early AD.^23,34,35 Our study showed significant increases of all 6 of the AD pathological proteins we tested in plasma, 3 of the 6 proteins in vitreous, and 1 protein in aqueous, in those with DR compared to those without DR. In addition, we found that participants with DR trended toward lower cognitive scores compared to those without DR, providing some clinical validation of the potential for DR patients to progress to mild cognitive impairment at an earlier age. To our knowledge, this is the first study to show that AD protein biomarkers are significantly upregulated in plasma and eye fluid in those with DR compared to those without DR.

While the link between diabetes and risk of dementia has been well studied in cross-sectional and longitudinal studies (3–11), the association between diabetic retinopathy and AD-type dementia is less well understood. A longitudinal cohort study by Lee et al. in 2019 found that DR was associated with a 44% increased risk of AD even after controlling for age, sex, education, smoking and APOE genotype.²⁴ A follow up study by the same group found that the presence of DR for more than 5 years was strongly associated with increased risk of dementia and AD even after adjusting for factors that indicate an increased severity of DM (such as microalbuminuria, duration of glucose impairment, and eGFR).²⁵ A Danish cohort study published in 2022 found that patients with DR were more likely to develop AD than patients without DM (HR 1.24) and patients with DM but without DR (HR 1.34).³⁶ A recent mendelian randomization study by Ouyang et al. showed those with an increased genetic susceptibility to AD are at increased risk for development of DR.³⁷ Taken together, these prior studies indicate that DR may be an independent risk factor not only for cognitive impairment and dementia, but specifically for AD-type dementia. Here we show increased AD biomarkers associated with DR, supporting the potential association of DR with AD neuropathology.

Research has supported that there are 2 pathological mechanisms that lead to vision loss in DR.³⁸ The first is the vascular etiology, leading to development of clinically evident hemorrhages, microaneurysms, exudates, and neovascularization that is visualized on examination and multimodal imaging using color fundus photography, fluorescein angiography, and optical coherence tomography (OCT). The second pathological mechanism is retinal neurodegeneration, which is more insidious and chronic in nature, causing retinal nerve thinning from cumulative loss of neurons and glial cells, and may be visualized clinically on OCT. Retinal neurodegeneration occurs early in the course of diabetes before the development of retinopathy.^39,40 Animal studies have confirmed features of apoptotic cell death and neurodegeneration of the retina that began within weeks of the onset of hyperglycemia.^41–44 Early studies using scanning laser polarimetry have shown that retinal nerve fiber and ganglion cell layer thicknesses are reduced in those with DM with no clinically evident retinopathy.^45–47 Subsequent studies using OCT have found that retinal thickness is reduced in patients with DM without DR compared to healthy controls,^48–54 and more recent studies using higher resolution spectral domain OCT have confirmed the thinning is primarily localized to the nerve fiber and ganglion cell layers.^55–60 Additional evidence of retinal neurodegeneration includes subtle alterations on electroretinogram in patients with DM before the development of DR^61–64 with reduction in b-wave amplitudes and delayed oscillatory potentials.³⁸ Notably, a series of studies have shown that multifocal electroretinogram may be able to predict the eventual development of DR and macular edema in those with Type 1 diabetes.^65–68

The retinal vasculature provides nutritional support to the neural retinal tissue, while the neural cells send signals to the vascular smooth muscle cells. This interplay, referred to as the retinal neurovascular unit, creates a tight blood retinal barrier, which is broken down early in diabetic retinopathy by neurodegeneration, leading to breakdown of neurovascular coupling and disruption of the blood retinal barrier, loss of autoregulation, and alterations in blood flow.^69–71 Similarly, and perhaps simultaneously, in diabetic cerebrovascular disease, neurovascular coupling breakdown leads to the disruption of the blood-brain barrier,^72–74 leading to clinical changes in the brain that further contribute to cognitive dysfunction.

Both DR and AD share a compromised neurovascular unit marked by persistent neuroinflammation, glial activation, and vascular degeneration, disrupting normal neuronal function in both retina and brain. In diabetes, chronic hyperglycemia and insulin resistance trigger oxidative stress and the buildup of advanced glycation end-products, pathological processes that similarly exacerbate inflammation and neuronal injury in AD.^39,75 Other key pathways, such as impaired insulin signaling and low-grade inflammation, are common to both DR and AD, and studies indicate that many molecular triggers in AD-affected brain tissue are also active in diabetic retinas. Collectively, these shared mechanisms underscore a retina–brain axis,” wherein retinal changes in diabetes may mirror AD-related neurodegenerative changes, highlighting potential common targets for therapeutic intervention.⁷⁶ Thus, early retinal neurodegenerative changes in DR may have contributed to the significant upregulation of the AD-related protein biomarkers observed in the vitreous and aqueous humor in our study, and the well-established overlap in pathogenic mechanisms between DM and AD^{15–21,39,75,76} may account for the biomarker elevations in the plasma. With a mean cohort age of 57.1 years, well before the typical onset of AD, our findings suggest that DR may serve as an early clinical biomarker of AD and identify a population that could benefit from targeted early dementia screening.

In our cohort, DR frequently coexisted with AMD and glaucoma, two other prevalent ocular conditions in aging populations that share mechanistic overlaps with AD. Like DR, both AMD and AD are associated with common systemic risk factors such as hypertension, smoking, and dyslipidemia.⁷⁷ These conditions are characterized by chronic inflammation, oxidative stress, amyloid accumulation, complement activation, cellular senescence, and mitochondrial dysfunction.⁷⁸ Similarly, glaucoma involves progressive loss of retinal ganglion cells and optic nerve degeneration, paralleling neurodegenerative processes observed in AD.⁷⁹ Notably, both diseases may share impairment of the glymphatic system, a paravascular clearance pathway in the brain and eye responsible for removing metabolic waste, including Aβ, further supporting the potential link between ocular pathology and early AD-related neurodegeneration.⁸⁰ Despite the coexistence of AMD and glaucoma with DR in our cohort, we adjusted for these comorbidities in our analysis and the presence of DR remained independently associated with elevated AD-related biomarkers in the ocular fluids.

This study has several notable strengths. Its prospective design allowed for better control of variables and helped minimize both recall and selection bias. Conducted at an urban academic health center, the study included a racially and ethnically diverse patient cohort, enhancing the generalizability of our findings.

However, the study also has limitations. Although prospectively designed, it remains cross-sectional in nature, limiting our ability to assess longitudinal changes in biomarker levels over time. The relatively small sample size restricted our ability to stratify participants by ocular disease severity or surgical indication (Table 1), such as different stages of DR. Additionally, because DR-related complications that occur in the advanced stages of retinopathy were the primary indication for surgery in many cases, our findings may be skewed by reflecting biomarker changes associated with advanced retinopathy. Furthermore, including participants undergoing pars plana vitrectomy for a range of indications introduced heterogeneity, as conditions such as rhegmatogenous retinal detachment, macular hole, and epiretinal membrane may have independent neuroinflammatory or vascular effects. Many participants with DR also had AMD or glaucoma. Although we adjusted for ocular comorbidities in our analyses, their coexistence may have compounded underlying neurovascular compromise. On the other hand, patients with eye disease represent an at-risk population for dementia, often developing it at a younger age. We believe this group warrants further in-depth study. While overlapping eye conditions could have contributed to the observed upregulation of biomarkers in vitreous and aqueous humor, they are unlikely to fully explain the systemic elevations seen in plasma, given the eye's status as an immunologically privileged site.

Another limitation is the absence of a gold-standard comparison for AD biomarkers, such as cerebrospinal fluid or MRI-based markers. However, since our cohort included younger individuals without a known diagnosis of cognitive dysfunction, there was no clinical or research justification for lumbar puncture or neuroimaging. Despite this, our study offers some insights into potential differences in AD biomarker levels between individuals with and without DR.

Lastly, while we measured phosphorylated tau at threonine 181 (ptau181), we did not assess ptau217, which recent studies suggest may have superior diagnostic accuracy in AD.⁸¹ Future research should explore the detectability of ptau217 in ocular fluids and compare its performance to ptau181 in individuals at risk of developing AD.

Conclusion

In summary, this study showed significant upregulation of AD associated biomarkers in the plasma, vitreous, and aqueous in those with DR compared to those without DR. These results demonstrate biological evidence supporting the link between DR and AD neuropathology at early disease stages and that those with DR represent an at-risk target population that may benefit from additional screening for cognitive disorders.

Footnotes

Acknowledgements

The authors have no acknowledgments to report.

ORCID iD

Thor D Stein

Ethical considerations

This study was approved by the Ethics Committee of Boston University Medical Center (BUMC) Institutional Review Board (study reference number H-37370, principal investigator MLS) on September 30, 2021. All participants provided written informed consent prior to enrolment in the study. This research was conducted ethically in accordance with the World Medical Association Declaration of Helsinki.

The ethics review committee of the BUMC Institutional Review Board (study reference number H-37370, principal investigator MLS) on September 30, 2021. Written informed consent for inclusion in this research was obtained from the patients prior to surgery.

The experimental protocols were approved by the Institutional Review Board (IRB) of BUMC (study reference number H-37370, principal investigator MLS) on September 30, 2021. All research activities complied with ethical regulations and were performed in accordance with regulations of each hospital. Informed consent to use histopathological samples and pathological diagnostic reports for research purposes was obtained from all patients prior to surgery. They were given the option to refuse to participate by opting out.

Consent to participate

Written informed consent was obtained in English or Spanish language from all participants who enrolled in the study.

Consent for publication

Not applicable.

Author contribution(s)

Manju L Subramanian: Conceptualization; Funding acquisition; Investigation; Methodology; Project administration; Resources; Software; Supervision; Validation; Visualization; Writing – original draft; Writing – review & editing.

Konstantina Sampani: Data curation; Formal analysis; Writing – original draft; Writing – review & editing.

Steven Ness: Investigation; Methodology; Writing – review & editing.

Fatima Tuz-Zahra: Data curation; Formal analysis; Writing – review & editing.

Sreevardhan Alluri: Methodology; Project administration; Resources; Software; Writing – review & editing.

Xuejing Chen: Investigation; Methodology; Writing – review & editing.

Nicole H Siegel: Investigation; Methodology; Writing – review & editing.

Nurgul Aytan: Methodology; Project administration; Resources; Software; Writing – review & editing.

Weiming Xia: Methodology; Project administration; Resources; Software; Writing – review & editing.

Yorghos Tripodis: Formal Analysis; Writing – review & editing.

Michael L Alosco: Conceptualization; Writing – review & editing.

Thor D Stein: Conceptualization; Funding acquisition; Investigation; Methodology; Project administration; Resources; Software; Supervision; Validation; Visualization; Writing – review & editing.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by NIH/NIA/ERP: 1R03AG063255-01, PI Manju Subramanian; P30AG072978, the United States Department of Veterans Affairs, Veterans Health Administration, BLRD Merit Award (I01BX005933), PI Thor Stein.

U.S. Department of Veterans Affairs, National Institute on Aging, (grant number P30AG072978, 1R03AG063255-01).

Declaration of conflicting interests

The authors declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. MLA received a single time honorarium from the Michael J Fox Foundation for services unrelated to this study. MLA also receives royalties from Oxford University Press Inc.

Data availability statement

The datasets generated and analyzed in this study are available upon request to any qualified researchers. As corresponding authors, we have full access to all data in the study and can provide relevant information as needed. The data adheres to FAIR principles—findable, accessible, interoperable, and reusable—ensuring transparency and reproducibility

References

Alzheimer's Association . 2023 Alzheimer's disease facts and figures. Alzheimers Dement 2023; 19: 1598–1695.

Prince

Wimo

Guerchet

, et al. World Alzheimer Report 2015. The Global Impact of Dementia. An Analysis of Prevalence, Incidence, Cost and Trends. London: Alzheimer's Disease International, 2015.

Magliano

Boyko

. IDF DIABETES ATLAS. 10th ed. Brussels: International Diabetes Federation, 2021.

Ciudin

Espinosa

Simó-Servat

, et al. Type 2 diabetes is an independent risk factor for dementia conversion in patients with mild cognitive impairment. J Diabetes Complications 2017; 31: 1272–1274.

de Almeida Faria

ACR

Dall'Agnol

Gouveia

, et al. Risk factors for cognitive decline in type 2 diabetes mellitus patients in Brazil: a prospective observational study. Diabetol Metab Syndr 2022; 14: 105.

Jash

Gondaliya

Kirave

, et al. Cognitive dysfunction: a growing link between diabetes and Alzheimer's disease. Drug Dev Res 2020; 81: 144–164.

Lacy

Gilsanz

Karter

, et al. Long-term glycemic control and dementia risk in type 1 diabetes. Diabetes Care 2018; 41: 2339–2345.

Xia

Huang

, et al. The factors contributing to cognitive dysfunction in type 2 diabetic patients. Ann Transl Med 2020; 8: 104.

Cheng

Huang

Deng

, et al. Diabetes as a risk factor for dementia and mild cognitive impairment: a meta-analysis of longitudinal studies. Intern Med J 2012; 42: 484–491.

10.

Gudala

Bansal

Schifano

, et al. Diabetes mellitus and risk of dementia: a meta-analysis of prospective observational studies. J Diabetes Investig 2013; 4: 640–650.

11.

Nguyen

QTH

Nguyen

TKO

, et al. Type 3 diabetes and its role implications in Alzheimer's disease. Int J Mol Sci 2020; 21: 3165.

12.

Nguyen

QTH

Nguyen

TTD

, et al. Role of insulin resistance in the Alzheimer's disease progression. Neurochem Res 2020; 45: 1481–1491.

13.

Abner

Nelson

Kryscio

, et al. Diabetes is associated with cerebrovascular but not Alzheimer's disease neuropathology. Alzheimers Dement 2016; 12: 882–889.

14.

Ahtiluoto

Polvikoski

Peltonen

, et al. Diabetes, Alzheimer disease, and vascular dementia: a population-based neuropathologic study. Neurology 2010; 75: 1195–1202.

15.

Etiene

Kraft

Ganju

, et al. Cerebrovascular pathology contributes to the heterogeneity of Alzheimer's disease. J Alzheimers Dis 1998; 1: 119–134.

16.

Mann

. The neuropathology of Alzheimer's disease: a review with pathogenetic, aetiological and therapeutic considerations. Mech Ageing Dev 1985; 31: 213–255.

17.

Ott

Jones

Daiello

, et al. Blood-cerebrospinal fluid barrier gradients in mild cognitive impairment and Alzheimer's disease: relationship to inflammatory cytokines and chemokines. Front Aging Neurosci 2018; 10: 245.

18.

de Leon

George

Ferris

, et al. Regional correlation of PET and CT in senile dementia of the Alzheimer type. AJNR Am J Neuroradiol 1983; 4: 553–556.

19.

Faulstich

. Brain imaging in dementia of the Alzheimer type. Int J Neurosci 1991; 57: 39–49.

20.

Waldron

Wintmolders

Bottelbergs

, et al. In vivo molecular neuroimaging of glucose utilization and its association with fibrillar amyloid-β load in aged APPPS1-21 mice. Alzheimers Res Ther 2015; 7: 76.

21.

Wurtman

. Biomarkers in the diagnosis and management of Alzheimer's disease. Metabolism 2015; 64: S47–S50.

22.

Jack

Jr Bennett

Blennow

, et al. NIA-AA Research framework: toward a biological definition of Alzheimer's disease. Alzheimers Dement 2018; 14: 535–562.

23.

Teunissen

Verberk

IMW

Thijssen

, et al. Blood-based biomarkers for Alzheimer's disease: towards clinical implementation. Lancet Neurol 2022; 21: 66–77.

24.

Lee

Larson

Gibbons

, et al. Associations between recent and established ophthalmic conditions and risk of Alzheimer's disease. Alzheimers Dement 2019; 15: 34–41.

25.

Lee

Krakauer

, et al. Diabetic retinopathy and dementia association, beyond diabetes severity. Am J Ophthalmol 2023; 249: 90–98.

26.

Chan

RNF

Tang

Chan

VTT

, et al. The cross-sectional and longitudinal relationship of diabetic retinopathy to cognitive impairment: a systematic review and meta-analysis. Eye (Lond) 2023; 37: 220–227.

27.

Wright

Stein

Jun

, et al. Association of cognitive function with amyloid-β and tau proteins in the vitreous humor. J Alzheimers Dis 2019; 68: 1429–1438.

28.

Subramanian

Vig

Chung

, et al. Neurofilament light chain in the vitreous humor of the eye. Alzheimers Res Ther 2020; 12: 111.

29.

Vig

Garg

Tuz-Zahra

, et al. Vitreous humor biomarkers reflect pathological changes in the brain for Alzheimer's disease and chronic traumatic encephalopathy. J Alzheimers Dis 2023; 93: 1181–1193.

30.

Fereshetian

Agranat

Siegel

, et al. Protein and imaging biomarkers in the eye for early detection of Alzheimer's disease. J Alzheimers Dis Rep 2021; 5: 375–387.

31.

Hussain

Sheikh

Subramanian

. The eye as a diagnostic tool for Alzheimer's disease. Life (Basel) 2023; 13: 726.

32.

Sampani

Ness

Tuz-Zahra

, et al. Neurodegenerative biomarkers in different chambers of the eye relative to plasma: an agreement validation study. Alzheimers Res Ther 2024; 16: 192.

33.

Jack

Jr Bennett

Blennow

, et al. A/T/N: an unbiased descriptive classification scheme for Alzheimer disease biomarkers. Neurology 2016; 87: 539–547.

34.

Hansson

Edelmayer

Boxer

, et al. The Alzheimer's association appropriate use recommendations for blood biomarkers in Alzheimer's disease. Alzheimers Dement 2022; 18: 2669–2686.

35.

Montoliu-Gaya

Alosco

Yhang

, et al. Optimal blood tau species for the detection of Alzheimer's disease neuropathology: an immunoprecipitation mass spectrometry and autopsy study. Acta Neuropathol 2023; 147: 5.

36.

Pedersen

Stokholm

Pouwer

, et al. Diabetic retinopathy predicts risk of Alzheimer's disease: a Danish registry-based nationwide cohort study. J Alzheimers Dis 2022; 86: 451–460.

37.

Ouyang

Yuan

, et al. Alzheimer's disease as a causal risk factor for diabetic retinopathy: a Mendelian randomization study. Front Endocrinol (Lausanne) 2024; 15: 1340608.

38.

Barber

. Diabetic retinopathy: recent advances towards understanding neurodegeneration and vision loss. Sci China Life Sci 2015; 58: 541–549.

39.

Simó

Stitt

Gardner

. Neurodegeneration in diabetic retinopathy: does it really matter? Diabetologia 2018; 61: 1902–1912.

40.

Fragiotta

Pinazo-Durán

Scuderi

. Understanding neurodegeneration from a clinical and therapeutic perspective in early diabetic retinopathy. Nutrients 2022; 14: 792.

41.

Mizutani

Kern

Lorenzi

. Accelerated death of retinal microvascular cells in human and experimental diabetic retinopathy. J Clin Invest 1996; 97: 2883–2890.

42.

Barber

Lieth

Khin

, et al. Neural apoptosis in the retina during experimental and human diabetes. Early onset and effect of insulin. J Clin Invest 1998; 102: 783–791.

43.

Ning

Baoyu

Yuzhen

, et al. Neuro-optic cell apoptosis and microangiopathy in KKAY mouse retina. Int J Mol Med 2004; 13: 87–92.

44.

Barber

Antonetti

Kern

, et al. The Ins2Akita mouse as a model of early retinal complications in diabetes. Invest Ophthalmol Vis Sci 2005; 46: 2210–2218.

45.

Lopes de Faria

Russ

Costa

. Retinal nerve fibre layer loss in patients with type 1 diabetes mellitus without retinopathy. Br J Ophthalmol 2002; 86: 725–728.

46.

Ozdek

Lonneville

Onol

, et al. Assessment of nerve fiber layer in diabetic patients with scanning laser polarimetry. Eye (Lond) 2002; 16: 761–765.

47.

Takahashi

Goto

Shoji

, et al. Diabetes-associated retinal nerve fiber damage evaluated with scanning laser polarimetry. Am J Ophthalmol 2006; 142: 88–94.

48.

Asefzadeh

Fisch

Parenteau

, et al. Macular thickness and systemic markers for diabetes in individuals with no or mild diabetic retinopathy. Clin Exp Ophthalmol 2008; 36: 455–463.

49.

Biallosterski

van Velthoven

Michels

, et al. Decreased optical coherence tomography-measured pericentral retinal thickness in patients with diabetes mellitus type 1 with minimal diabetic retinopathy. Br J Ophthalmol 2007; 91: 1135–1138.

50.

Bronson-Castain

Bearse

Jr Neuville

, et al. Adolescents with type 2 diabetes: early indications of focal retinal neuropathy, retinal thinning, and venular dilation. Retina 2009; 29: 618–626.

51.

Browning

Fraser

Clark

. The relationship of macular thickness to clinically graded diabetic retinopathy severity in eyes without clinically detected diabetic macular edema. Ophthalmology 2008; 115: 533–539.e532.

52.

Nilsson

von Wendt

Wanger

, et al. Early detection of macular changes in patients with diabetes using Rarebit Fovea Test and optical coherence tomography. Br J Ophthalmol 2007; 91: 1596–1598.

53.

Oshitari

Hanawa

Adachi-Usami

. Changes of macular and RNFL thicknesses measured by Stratus OCT in patients with early stage diabetes. Eye (Lond) 2009; 23: 884–889.

54.

van Dijk

Kok

Garvin

, et al. Selective loss of inner retinal layer thickness in type 1 diabetic patients with minimal diabetic retinopathy. Invest Ophthalmol Vis Sci 2009; 50: 3404–3409.

55.

El-Fayoumi

Badr Eldine

Esmael

, et al. Retinal nerve fiber layer and ganglion cell complex thicknesses are reduced in children with type 1 diabetes with no evidence of vascular retinopathy. Invest Ophthalmol Vis Sci 2016; 57: 5355–5360.

56.

Xiao

Lin

, et al. Two-year longitudinal study on changes in thickness of the retinal nerve fiber layer and ganglion cell layer in children with type 1 diabetes mellitus without visual impairment or diabetic retinopathy. Curr Eye Res 2022; 47: 1218–1225.

57.

Lee

Ryu

, et al. Peripapillary retinal nerve fiber layer and microvasculature in prolonged type 2 diabetes patients without clinical diabetic retinopathy. Invest Ophthalmol Vis Sci 2021; 62: 9.

58.

Gundogan

Akay

Uzun

, et al. Early neurodegeneration of the inner retinal layers in type 1 diabetes mellitus. Ophthalmologica 2016; 235: 125–132.

59.

Karti

Nalbantoglu

Abali

, et al. Retinal ganglion cell loss in children with type 1 diabetes mellitus without diabetic retinopathy. Ophthalmic Surg Lasers Imaging Retina 2017; 48: 473–477.

60.

Mohd-Ilham

Tai

ELM

Suhaimi

, et al. Evaluation of macular and retinal nerve fiber layer thickness in children with type 1 diabetes mellitus without retinopathy. Korean J Ophthalmol 2021; 35: 287–294.

61.

Huang

Chen

, et al. Multifocal electroretinogram can detect the abnormal retinal change in early stage of type2 DM patients without apparent diabetic retinopathy. J Diabetes Res 2021; 2021: 6644691.

62.

Jia

Lin

Xiao

, et al.

Are we overlooking the neurodegeneration in the early stage of type 1 diabetes mellitus without visual impairment or diabetic retinopathy: is it probably occurred before retinal vasculature dysfunction?

Curr Eye Res 2023; 48: 433–440.

63.

Fortune

Schneck

Adams

. Multifocal electroretinogram delays reveal local retinal dysfunction in early diabetic retinopathy. Invest Ophthalmol Vis Sci 1999; 40: 2638–2651.

64.

Zeng

Cao

, et al. Early retinal neurovascular impairment in patients with diabetes without clinically detectable retinopathy. Br J Ophthalmol 2019; 103: 1747–1752.

65.

Bearse

Jr Adams

Han

, et al. A multifocal electroretinogram model predicting the development of diabetic retinopathy. Prog Retin Eye Res 2006; 25: 425–448.

66.

Han

Bearse

Jr Schneck

, et al. Multifocal electroretinogram delays predict sites of subsequent diabetic retinopathy. Invest Ophthalmol Vis Sci 2004; 45: 948–954.

67.

Harrison

Bearse

Jr Schneck

, et al. Prediction, by retinal location, of the onset of diabetic edema in patients with nonproliferative diabetic retinopathy. Invest Ophthalmol Vis Sci 2011; 52: 6825–6831.

68.

Bearse

Jr Schneck

, et al. Local diabetic retinopathy prediction by multifocal ERG delays over 3 years. Invest Ophthalmol Vis Sci 2008; 49: 1622–1628.

69.

Antonetti

Silva

Stitt

. Current understanding of the molecular and cellular pathology of diabetic retinopathy. Nat Rev Endocrinol 2021; 17: 195–206.

70.

Garhöfer

Chua

Tan

, et al. Retinal neurovascular coupling in diabetes. J Clin Med 2020; 9: 2829.

71.

Starr

Wardlaw

Ferguson

, et al. Increased blood-brain barrier permeability in type II diabetes demonstrated by gadolinium magnetic resonance imaging. J Neurol Neurosurg Psychiatry 2003; 74: 70–76.

72.

van Sloten

Sedaghat

Carnethon

, et al. Cerebral microvascular complications of type 2 diabetes: stroke, cognitive dysfunction, and depression. Lancet Diabetes Endocrinol 2020; 8: 325–336.

73.

Woerdeman

van Duinkerken

Wattjes

, et al. Proliferative retinopathy in type 1 diabetes is associated with cerebral microbleeds, which is part of generalized microangiopathy. Diabetes Care 2014; 37: 1165–1168.

74.

Sanahuja

Alonso

Diez

, et al. Increased burden of cerebral small vessel disease in patients with type 2 diabetes and retinopathy. Diabetes Care 2016; 39: 1614–1620.

75.

Little

Llorián-Salvador

Scullion

, et al. Common pathways in dementia and diabetic retinopathy: understanding the mechanisms of diabetes-related cognitive decline. Trends Endocrinol Metab 2022; 33: 50–71.

76.

Ebrahimi

Thompson

Lauer

, et al. The retina-brain axis and diabetic retinopathy. Eur J Ophthalmol 2023; 33: 2079–2095.

77.

Kaštelan

Nikuševa-Martić

Pašalić

, et al. Genetic and epigenetic biomarkers linking Alzheimer's disease and age-related macular degeneration. Int J Mol Sci 2024; 25: 7271.

78.

Butovsky

Rosenzweig

. Alzheimer's disease and age-related macular degeneration: shared and distinct immune mechanisms. Immunity 2025; 58: 1120–1139.

79.

Sen

Saxena

Tripathi

, et al. Neurodegeneration in Alzheimer's disease and glaucoma: overlaps and missing links. Eye (Lond) 2020; 34: 1546–1553.

80.

Bodi

Sharma

. Normal-tension glaucoma and potential clinical links to Alzheimer's disease. J Clin Med 2024; 13: 1948.

81.

Thijssen

La Joie

Strom

, et al. Plasma phosphorylated tau 217 and phosphorylated tau 181 as biomarkers in Alzheimer's disease and frontotemporal lobar degeneration: a retrospective diagnostic performance study. Lancet Neurol 2021; 20: 739–752.