Abstract
Objective: Previous studies have shown a relationship between retinopathy and cognition including population with and without chronic kidney disease (CKD) but data regarding peritoneal dialysis (PD) are limited. This study aims to investigate the relationship between retinopathy and cognitive impairment in patients undergoing peritoneal dialysis (PD). Methods: In this observational study, we recruited a total of 107 participants undergoing PD, consisting of 48 men and 59 women, ages ranging from 21 to 78 years. The study followed a cross-sectional design. Retinal microvascular characteristics, such as geometric changes in retinal vascular including tortuosity, fractal dimension (FD), and calibers, were assessed. Retinopathy (such as retinal hemorrhage or microaneurysms) was evaluated using digitized photographs. The Modified Mini-Mental State Examination (3MS) was performed to assess global cognitive function. Results: The prevalence rates of retinal hemorrhage, microaneurysms, and retinopathy were 25%, 30%, and 43%, respectively. The mean arteriolar and venular calibers were 63.2 and 78.5 µm, respectively, and the corresponding mean tortuosity was 37.7 ± 3.6 and 37.2 ± 3.0 mm−1. The mean FD was 1.49. After adjusting for age, sex, education, mean arterial pressure, and Charlson index, a negative association was revealed between retinopathy and 3MS scores (regression coefficient: −3.71, 95% confidence interval: −7.09 to −0.33, p = 0.03). Conclusions: Retinopathy, a condition common in patients undergoing PD, was associated with global cognitive impairment. These findings highlight retinopathy, can serve as a valuable primary screening tool for assessing the risk of cognitive decline.
Keywords: Retinopathy, cognitive impairment, peritoneal dialysis, retinal microvascular abnormalities
Introduction
Peritoneal dialysis (PD), an at-home therapy, involves self-management and self-monitoring, thus partly relying on patients having normal cognitive function [1]. Impaired cognition not only reduces a patient’s ability to follow complex medical or dietary requirements [2] but also may lead to PD-related peritonitis [3], which is an independent mortality predictor in this patient population [4]. Previous studies found a high prevalence (27% to 67%) of cognitive impairment in patients with end-stage kidney disease (ESKD) [4–5]. Hence, it is important to determine the risk factors of cognitive dysfunction in clinical practice.
We have previously shown that the risk of executive dysfunction and impairment of immediate memory, overall cognition, and visuospatial skills in PD patients with diabetes and retinopathy was twice as high as that in non-diabetes patients undergoing PD [6]. After multivariable adjustment, the overall and special cognition functions of PD patients with diabetes were almost equal to those in non-diabetes PD patients, suggesting that the mechanisms correlating with cognitive impairment in type 2 diabetes mellitus (T2DM) may be involved in PD patients with ESKD.
Not only retinopathy but also cognitive dysfunction is an important comorbidity of diabetes [7]. Retinal microvasculature disease and cognitive impairment have been reported as potential evidence for a microvascular component [8]. Diabetic retinopathy (DR) is associated with cognitive impairment, and screening for it would be helpful in the early identification of individuals with cognitive impairment [9]. Several studies have investigated the association between retinopathy and cognitive function in the non-kidney disease population [9–11]. Peng et al.’s study supported that deep vascular plexus vessel density could be associated with early cognitive impairment in patients with CKD [12]. However, the association between retinopathy and cognition among PD patients has not been separately examined. Therefore, we designed this cross-sectional study to investigate the relationship between retinopathy and cognitive function among patients undergoing PD by evaluating global cognitive function.
Materials and methods
Patient selection and study design
We recruited 153 patients who underwent cognitive assessments at Peking University Shenzhen Hospital. Among them, 107 patients underwent retinal examination between August 2018 and December 2020. All participants provided signed informed consent. The inclusion criteria were as follows: age ≥ 18 years; capable of completing all necessary tests and questionnaires; receiving continuous peritoneal dialysis for at least 3 months and clinically stable. Excluded patients included those with a history of systemic infection, cancer, hepatitis, or acute cardiovascular events; those who underwent surgery or suffered trauma within the past month; those with conditions limiting study participation (language incompatibility, severe visual impairment, confusion, illiteracy, preexisting dementia, various mental disorders, or upper limb disability); or those with unobtainable retinal examination results or no ophthalmology referral. All participants received routine glucose-buffered lactate-based peritoneal dialysis solution (Baxter Healthcare, Guangzhou, China). The study complied with the Helsinki Declaration (revised in 2013) and obtained approval from the hospital’s Institutional Review Board (IRB of Peking University Shenzhen Hospital [2019]048th). During and after data collection, all authors had no access to information identifying individual participant identities.
Clinical characteristics
We recorded comorbid conditions and demographics, including education level, age, body mass index (BMI), sex, primary kidney disease, systolic and diastolic blood pressure, PD vintage, and Charlson index [13]. Educational level was recorded as the highest diploma received by patients: high school or higher, middle school, and elementary school or lower.
Laboratory methods
Venous blood count and biochemical measurements were performed after overnight fasting during continuous peritoneal dialysis. Venous blood cell counts and biochemical measurements were performed after overnight fasting during continuous peritoneal dialysis. Biochemical data (including triglycerides, serum albumin, high-sensitivity C-reactive protein [hs-CRP], total cholesterol, hemoglobin A1c [HbA1c], and hemoglobin levels) were collected using an automated Hitachi chemistry analyzer and the average values over the past 3 months were calculated. Dialysis adequacy was defined by creatinine clearance and total Kt/V. Residual renal function was defined as the average of 24-h residual urine creatinine and urea clearance.
Assessment of cognitive function
Each patient underwent cognitive function assessment in a separate room, accompanied by medical staff. Additionally, two healthcare personnel observed in the same room. All healthcare personnel received training on how to accurately conduct the complete assessment.
The Modified Mini-Mental State Examination (3MS) [14] comprises the full battery to tests for evaluating the overall cognitive function. Since the 3MS mean scores vary by education level [15], overall cognitive impairment was defined as a score lower than 75 for patients with an education level lower than high school and a score less than 80 for those with a high school education.
Assessment of retinopathy and retinal vessel geometric parameters
Clinical examinations and retinal photography were performed for 107 participants. In this observational study, a double-blind approach was used, where neither the ophthalmologists nor the participants were aware of the cognitive function scores. Experienced ophthalmologists conducted comprehensive ophthalmic examinations and retinal evaluations. Retinopathy was defined by an ophthalmologist after consultation. Retinal arterial and venous diameters, tortuosity, and fractal dimension (FD) were also investigated. A non-mydriatic digital fundus camera was used to take two fundus photos of 45 degrees in two fields (one centered on the macula and the other on the optic nerve) in both eyes of the patients. The vessels were segmented from the color fundus images by using U-Net [16]. The position of the optic disk was detected using the Faster-RCNN algorithm [17]. The U-Net [16] algorithm was then used again to precisely segment the optic disk boundary within the detected region. This method was used to extract the skeletons of the vessels, and a support vector machine (SVM) was used to classify vessels into arterioles and venules [18].
The manifestations of retinopathy included the following findings: soft or hard exudates, blot- or flame-shaped retinal hemorrhages, disk swelling, intraretinal microvascular abnormalities, macular edema, microaneurysms, new vessels, vitreous hemorrhage, venous beading, or laser photocoagulation scars [19]. The retinal arteriolar and venular diameters were quantified by a computer-assisted technique. We digitized the images and measured the individual diameters of specific veins and arteries passing through a defined area around the optic disk (within 1–2 disk diameters from the center of the disk) as mean diameter and calculated the arteriole-to-venule ratio (AVR).
Retinal arteriolar and venular curvature tortuosity were normalized on the basis of the total path length and derived from the integral of the curvature square along the vessel path. Smaller tortuosity values indicate straighter vessels [20]. Since the curvature is defined as the inverse of the radius of curvature, which is measured in the unit of length, the unit for curvature is inverse-length (mm−1 is used here). Using the box-counting method, the line of retinal vessels was evaluated to determine the vascular FD of the retina. In this method, each photo was divided into a series of squares with different side lengths [21].
Statistical analysis
Demographic data and retinal measurements were compared between patients with and without retinopathy. The continuous variables that showed a normal distribution were analyzed by Student’s t-tests. Continuous variables with a non-normal distribution were analyzed by Mann-Whitney U test, and categorical variables were analyzed by the chi-squared test. Linear regression was applied to calculate the coefficient of retinopathy-related cognitive impairment after adjusted the findings for age, sex, education, mean arterial pressure, and Charlson index. Pre-specified subgroup analyses were performed with stratification by age, sex, level of education, Charlson index, mean arterial pressure, PD vintage, and diabetes. In the subgroup analysis, it was assessed whether the exposure factors associated with retinopathy interfered with other factors, such as age, sex, and level of education, by using a linear regression model. Furthermore, a sensitivity analysis was conducted using a multivariable linear mixed regression model to investigate the relationship between other retinal microvascular characteristics, such as FD, microaneurysms, retinal hemorrhage, soft exudates, and hard exudates, and 3MS scores. R version 4.0.2/RStudio version 1.1.463 was used for all analyses. Statistical testing was two-tailed with the significance level set at 0.05.
Results
Patient characteristics
Among the 200 collected patients, 153 (76.5%) were included in the study, and the remaining 47 (23.5%) were excluded because they did not meet the inclusion criteria, but only 107 patients underwent cognitive testing and retinal examination (Figure 1). The average age of the enrolled patients was 45.52 ± 12.99 years, and they had PD vintages of 30 (10.50–72.00) months. There was no significant difference in male PD vintage, age, education, BMI, sex, HbA1c, and total kt/v. Among these patients, 18.87% had diabetes, and the proportions of patients with diabetes were significantly different in the groups with and without retinopathy (Table 1).
Figure 1.
Flow chart of research subject screening.
Table 1.
Patient characteristics.
| Total | Without Retinopathy | With Retinopathy | p value | |
|---|---|---|---|---|
| Number of patients, n (%) | 107 | 61 (57.01) | 46 (42.99) | |
| Age, year | 45.52 ± 12.99 | 44.64 ± 13.18 | 46.70 ± 12.77 | 0.418 |
| Men, n (%) | 48 (44.86) | 29 (47.54) | 19 (41.30) | 0.656 |
| PD vintage, months | 30.00 (10.50–72.00) | 30.00 (7.00–72.00) | 38.00 (12.00–72.00) | 0.603 |
| Diabetes patients, n (%) | 20 (18.87) | 6 (10.00) | 14 (30.43) | 0.016 |
| Charlson index | 5.00 (4.00–8.00) | 5.00 (4.00–7.00) | 6.00 (5.00-9.00) | 0.007 |
| Level of education, n (%) | – | 0.320 | ||
| ≤Elementary school | 13 (12.15) | 6 (9.84) | 7 (15.22) | |
| Middle school | 29 (27.10) | 14 (22.95) | 15 (32.61) | |
| High school | 38 (35.51) | 22 (36.07) | 16 (34.78) | |
| >High school | 27 (25.23) | 19 (31.15) | 8 (17.39) | |
| BMI, kg/m2 | 21.61 ± 2.95 | 21.36 ± 3.07 | 21.94 ± 2.78 | 0.307 |
| MAP, mmHg | 109.97 ± 15.01 | 107.01 ± 13.19 | 113.90 ± 16.47 | 0.022 |
| HbA1c, % | 5.40 (4.93–5.70) | 5.40 (4.90–5.60) | 5.50 (5.00–5.90) | 0.140 |
| Hemoglobin, g/L | 107.00 (91.50–117.50) | 109.00 (98.00–121.00) | 100.00 (87.25–112.00) | 0.052 |
| Serum albumin, g/L | 35.72 ± 4.73 | 36.44 ± 4.85 | 34.74 ± 4.43 | 0.064 |
| Triglyceride, mmol/L | 1.41 (1.08–2.14) | 1.60 (1.17–2.10) | 1.28 (1.02–2.19) | 0.370 |
| Total cholesterol, mmol/L | 4.60 (3.80–5.30) | 4.34 (3.74–5.00) | 4.84 (4.13–5.41) | 0.094 |
| hs-CRP, mg/dl | 2.00 (0.64–5.75) | 1.89 (0.54–5.97) | 2.04 (0.8–-5.65) | 0.684 |
| Total weekly kt/v | 2.00 (1.80–2.30) | 2.00 (1.85–2.45) | 1.99 (1.77–2.21) | 0.531 |
| Total Clcr, L/wk/1.73 m2 | 50.42 ± 11.79 | 51.17 ± 14.15 | 49.37 ± 7.41 | 0.446 |
Values are expressed as mean ± standard deviation, median (interquartile range), or n (%), as appropriate. PD: peritoneal dialysis; BMI: body mass index; MAP: mean arterial pressure; hs-CRP: high-sensitivity C-reactive protein; Kt/V: urea clearance per week.
Evaluation of retinal microvascular characteristics
In terms of retinal microvascular abnormalities, the prevalence of retinopathy, microaneurysms, retinal hemorrhage, soft exudates, and hard exudates was 43%, 30%, 25%, 7%, and 14%, respectively. Retinal hemorrhage, soft exudates, and hard exudates were more common in the retinopathy group than in the group without retinopathy (p < 0.05). The mean arteriole and venule calibers were 63.2 and 78.5 µm, respectively. The mean tortuosity of arterioles and venules was 37.7 ± 3.6 and 37.2 ± 3.0 mm−1, respectively. The mean FD was 1.49. The mean FD in the retinopathy group was sparser than that in the group without retinopathy (Table 2).
Table 2.
Evaluation of retinal microvascular characteristics.
| Total | Without Retinopathy | With Retinopathy | p value | |
|---|---|---|---|---|
| Entries | 214 | 122 (57.01) | 92 (42.99) | |
| Mean caliber of arteriole (μm) | 63.20 (55.68–69.80) | 64.19 (57.71–70.00) | 61.45 (55.56–68.80) | 0.350 |
| Mean caliber of venule (μm) | 78.53 (71.46–88.11) | 78.15 (69.54–88.51) | 79.03 (72.42–86.24) | 0.874 |
| AVR | 0.80 ± 0.13 | 0.81 ± 0.13 | 0.80 ± 0.12 | 0.589 |
| Mean tortuosity of arteriole (mm-1) | 37.74 ± 3.60 | 37.34 ± 3.36 | 38.27 ± 3.86 | 0.067 |
| Mean tortuosity of venule (mm-1) | 37.24 ± 2.98 | 37.32 ± 2.89 | 37.12 ± 3.11 | 0.631 |
| Fractal dimension | 1.49 (1.47–1.51) | 1.50 (1.48–1.51) | 1.48 (1.45–1.51) | <0.001 |
| Microaneurysms, n (%) | 65 (30.37) | 23 (18.85) | 42 (45.65) | <0.001 |
| Retinal hemorrhage, n (%) | 54 (25.23) | 12 (9.84) | 42 (45.65) | <0.001 |
| Drusen, n (%) | 9 (4.21) | 5 (4.10) | 4 (4.35) | >0.99 |
| Soft exudates, n (%) | 14 (6.54) | 1 (0.82) | 13 (14.13) | <0.001 |
| Hard exudate, n(%) | 30 (14.02) | 7 (5.74) | 23 (25.00) | <0.001 |
| AV nicking, n(%) | 53 (24.77) | 25 (20.49) | 28 (30.43) | 0.1315 |
AVR: arteriole-to-venule ratio; AV nicking: arteriovenous nicking.
Evaluation of global cognitive function
Among the participants who underwent cognitive assessments, the likelihood of cognitive impairment evidenced by 3MS scores was higher in those with retinopathy (Figure 3). The crude linear regression model showed that retinopathy was significantly associated with global cognitive impairment (3MS scores: regression coefficient, −4.20; 95% CI, −7.90 to −0.50; p = 0.03). After adjusting for age, sex, education, mean arterial pressure, and Charlson Index, retinopathy remained significantly associated with the risk of global cognitive impairment (regression coefficient, −3.71; 95% CI: −7.09 to −0.33; p = 0.03) (Figure 4).
Figure 3.
Box plot of evaluation of global cognitive function using t-test, p = 0.032.
Figure 4.
Association between retinopathy and 3MaS scores by multivariable linear regression analysis. Scatterplot of association between retinopathy and 3MS scores by multivariable linear regression analysis (sample size 107). The Solid line represents the regression line with a coefficient of −4.20. The area within the dotted lines represents the Confidence Interval at 95% (−7.90 ∼ −0.50, p = 0.026).
Subgroup and sensitivity analyses
Subgroup analysis showed no interaction between retinopathy and age, sex, level of education, Charlson index, mean arterial pressure, PD vintage, or diabetes in the linear regression models (Figure 2). Sensitivity analyses using multivariable linear regression models showed no association between retinal microvascular characteristics, including FD, microaneurysms, retinal hemorrhage, soft exudates, and hard exudates, and 3MS scores (Table 3).
Figure 2.
Subgroup analysis of associations between retinopathy and 3MS scores.
The squares and horizontal bars represent estimates and 95% confidence intervals (CIs).
Table 3.
Association between retinal microvascular characteristic measures and 3MS scores as determined by multivariable linear mixed regression analysis.
| Without characteristics | With characteristics | Beta (95% CI) | p-value | |
|---|---|---|---|---|
| Fractal dimension | 107 | 107 | 1.77 (−0.46 ∼ 3.99) | 0.126 |
| Microaneurysms | 149 | 65 | 1.01 (−1.36 ∼ 3.38) | 0.409 |
| Retinal hemorrhage | 160 | 54 | 0.51 (−2.05 ∼ 3.08) | 0.699 |
| Soft exudates | 200 | 14 | 1.54 (−2.86 ∼ 5.93) | 0.499 |
| Hard exudates | 184 | 30 | −0.16 (−3.30 ∼ 2.98) | 0.922 |
Adjusted model: adjusted for demographic and clinical measures (including age, sex, level of education, Charlson index, and mean arterial pressure). The dependent variable of the model was 3MS score.
Discussion
In this observational study, PD patients with retinopathy showed a higher possibility of poorer overall cognition. However, the geometric parameters associated with the retinal vasculature, including calibers, tortuosity, and FD, were not associated with cognitive impairment.
Retinopathy has been relatively understudied in PD patients. At the fifth visit in the Atherosclerosis Risk in Communities (ARIC) study, 6% of the participants had mild retinopathy and 2% had moderate/severe retinopathy (n = 2624) [22]. A longitudinal study of this cohort performed using data from the third visit also found that retinopathy and retinal hemorrhage were more associated with a 20-year cognitive decline [23]. In the Chronic Renal Insufficiency Cohort Study, the prevalence of retinopathy in patients with CKD and cognitive impairment was 30% and 14%, respectively [24]. In comparison with participants without retinopathy, those with retinopathy were associated with a higher possibility of cognitive impairment in naming and execution after multivariable adjustment. To date, there has been only one nationwide multicenter study of retinopathy in PD patients; it showed the prevalence of DR in these patients was 19.1% [6]. These discrepancies could be attributed to differences in the different populations. In the current study, the prevalence of retinopathy was 42.99% and higher than that in previous studies, which was related to the high incidence of comorbidity such as diabetes and hypertension in this study population. Nevertheless, our subgroup analysis showed no difference in retinopathy and overall cognitive function between diabetes and non-diabetes patients. This result was consistent with the finding in the ARIC study that the association of retinal signs with 20-year cognitive change was similar in both diabetes and non-diabetes patients [23].
The eye is the window for direct observation of blood vessels and neuropathy, and retinal microvascular abnormalities and microvascular changes in cognitive dysfunction may be linked due to similarities in the blood–brain and the blood–retina barrier, which may be a sign of cerebrovascular lesions [25]. A previous study showed that retinopathy was significantly associated with cognitive impairment in several domains, including executive function, attention, and naming in CKD patients [24], indicating that evaluation of the abnormalities of the retinal microvasculature could be a prospective tool to identify CKD patients at an increased risk of cognitive impairment.
The relationship between global cognitive impairment and retinopathy is a topic of interest, and some common biological pathways and risk factors, such as age, may underlie the link between retinopathy and neurodegenerative diseases, both of which are known to affect cognition. However, adjustments for risk factors did not weaken the associations between retinopathy and 3MS test results in the present study.
A recent meta-analysis evaluated the association between cognitive impairment and retinal vascular parameters such as retinal vascular caliber, tortuosity, and FD, and found that cognitive impairment might be related to a smaller retinal microvascular FD [26]. Similarly, retina calibers and tortuosity were not associated with cognitive impairment in our study. However, we did not observe the relationship between FD and cognitive impairment. Our results are consistent with those of Cai et al. [27]. FD is a novel parameter and a “global” measure of the overall geometric complexity of the retinal vascular network [28] and has been linked to stroke, diabetes, hypertension, and chronic kidney disease [29]. A previous study employed the Abbreviated Mental Test to evaluate cognitive function and showed that reduced FD was related to cognitive dysfunction [29]. The FD value in our study was slightly higher than that of the population. The major differences in results were related to the sample size, different cognitive scales, and different software algorithms and calculation techniques used. In addition, glucose exposure [30,31], microinflammatory state [32], and uremia per se in PD patients may trigger the proangiogenesis response [33]. An additional question addressed was whether these changes occur in the retina. Further research is needed to determine whether FD is related to this response in PD patients.
Our study has potential clinical implications. Our finding suggests that retinopathy may provide information about cognitive impairment. Digital retinal image analysis is becoming more common because it is noninvasive and can be performed quickly in clinics. Artificial intelligence and computer testing programs can help analyze different aspects of retinopathy [34,35]. A scheme has been developed to integrate AI-based applications that combine diabetic retinopathy screening with multimodal features. Fundus images, including digital fundus photography and OCT, are combined with multimodal features to help identify various systemic diseases such as diabetes, nephropathy, cognitive disorder, and Cerebro- or cardiovascular disease. This integration assists in four key areas of medical practice [36]. With the help of emerging imaging techniques, the examination of the retinal microvascular system may contribute to the development of future noninvasive screening tools for the assessment of the risk of cognitive impairment.
The strength of our study was that it provided a comprehensive and standardized assessment of cognitive function and an objective assessment of retinal images using a validated computer-assisted program. However, this study also had some limitations. First, the cross-sectional design could not clearly determine whether retinopathy is a risk biomarker or a pathogenic factor. Longitudinal assessments of retinopathy and cognitive function may address this point more clearly. Second, we did not investigate the pathogenesis underlying retinopathy. Further studies are needed to investigate the potential effects of different pathogenetic mechanisms. Third, our findings, which were obtained from the small number of participants included in this study, cannot be generalized to other PD populations because of differences in the bioclinical, socioeconomic, and demographic characteristics among different PD patients. Further larger, more diverse samples in future research should aim to adequately address this issue. Nevertheless, taken together, our results show that retinopathy is not uncommon in patients with PD and that patients with retinopathy have low 3MS scores.
To our knowledge, this is the first study to report the association between retinopathy and cognition in PD patients by using retinal photography and indicates the potential of the approach as an adjunct for the detection of global cognitive impairment. Our finding suggests a need for regular assessment of the fundus and evaluate the risk for cognitive impairment and suggests likely areas for further research.
Acknowledgments
The authors wish to express their gratitude to the participants who volunteered for this study. The authors are thankful to Yi-Hong Wang (The University of Sydney) for his advice on the statistical analysis. The authors also acknowledge Prof. Jie Dong (Peking University First Hospital) for her guidance pertaining to the study design.
Funding Statement
Guangdong Provincial Medical Science and Technology Research Fund Project (No.: B2019025).
Ethics approval and consent to participate
This study was approved by the Medical Ethics Committee of Peking University Shenzhen Hospital (2019-048). The study conformed to the provisions of the Declaration of Helsinki (as revised in 2013).
Disclosure statement
No potential conflict of interest was reported by the author(s).
Data availability statement
The data used to support the findings of this study are available from the corresponding author upon request.
References
- 1.Xu R, Pi HC, Xiong ZY, et al. Hyponatremia and cognitive impairment in patients treated with peritoneal dialysis. Clin J Am Soc Nephrol. 2015;10(10):1–8. doi: 10.2215/CJN.02240215. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Neumann D, Mau W, Wienke A, et al. Peritoneal dialysis is associated with better cognitive function than hemodialysis over a one-year course. Kidney Int. 2018;93(2):430–438. doi: 10.1016/j.kint.2017.07.022. [DOI] [PubMed] [Google Scholar]
- 3.Liao JL, Zhang YH, Xiong ZB, et al. The association of cognitive impairment with peritoneal dialysis-related peritonitis. Perit Dial Int. 2019;39(3):229–235. doi: 10.3747/pdi.2018.00180. [DOI] [PubMed] [Google Scholar]
- 4.Griva K, Stygall J, Hankins M, et al. Cognitive impairment and 7-year mortality in dialysis patients. Am J Kidney Dis. 2010;56(4):693–703. doi: 10.1053/j.ajkd.2010.07.003. [DOI] [PubMed] [Google Scholar]
- 5.Pereira AA, Weiner DE, Scott T, et al. Cognitive function in dialysis patients. Am J Kidney Dis. 2005;45(3):448–462. doi: 10.1053/j.ajkd.2004.10.024. [DOI] [PubMed] [Google Scholar]
- 6.Liao JL, Xiong ZY, Yang ZK, et al. An association of cognitive impairment with diabetes and retinopathy in end stage renal disease patients under peritoneal dialysis. PLOS One. 2017;12(8):e0183965. doi: 10.1371/journal.pone.0183965. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Biessels GJ, Whitmer RA.. Cognitive dysfunction in diabetes: how to implement emerging guidelines. Diabetologia. 2020;63(1):3–9. doi: 10.1007/s00125-019-04977-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Crosby-Nwaobi RR, Sivaprasad S, Amiel S, et al. The relationship between diabetic retinopathy and cognitive impairment. Diabetes Care. 2013;36(10):3177–3186. doi: 10.2337/dc12-2141. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Cheng D, Zhao X, Yang S, et al. Association between diabetic retinopathy and cognitive impairment: a systematic review and meta-analysis. Front Aging Neurosci. 2021;13:692911. doi: 10.3389/fnagi.2021.692911. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Ding J, Strachan MW, Reynolds RM, et al. Diabetic retinopathy and cognitive decline in older people with type 2 diabetes: the Edinburgh type 2 diabetes study. Diabetes. 2010;59(11):2883–2889. doi: 10.2337/db10-0752. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Heringa SM, Walraven I, Moll AC, et al. Vascular retinopathy in relation to cognitive functioning in an older population–the hoorn study. J Am Geriatr Soc. 2014;62(5):977–979. doi: 10.1111/jgs.12815. [DOI] [PubMed] [Google Scholar]
- 12.Peng SY, Wu IW, Sun CC, et al. Investigation of possible correlation between retinal neurovascular biomarkers and early cognitive impairment in patients with chronic kidney disease. Transl Vis Sci Technol. 2021;10(14):9. doi: 10.1167/tvst.10.14.9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Charlson ME, Hollenberg JP, Hou J, et al. Realizing the potential of clinical judgment: a real-time strategy for predicting outcomes and cost for medical inpatients. Am J Med. 2000;109(3):189–195. doi: 10.1016/s0002-9343(00)00477-0. [DOI] [PubMed] [Google Scholar]
- 14.Teng EL, Chui HC.. The modified mini-mental state (3MS) examination. J Clin Psychiatry. 1987;48(8):314–318. [PubMed] [Google Scholar]
- 15.Kurella M, Chertow GM, Fried LF, et al. Chronic kidney disease and cognitive impairment in the elderly: the health, aging, and body composition study. J Am Soc Nephrol. 2005;16(7):2127–2133. doi: 10.1681/ASN.2005010005. [DOI] [PubMed] [Google Scholar]
- 16.Olaf Ronneberger PF, Brox T.. International Conference on Medical image computing and computer-assisted intervention; 2015. Cham (Switzerland): Springer; 2015. p. 234–241. [Google Scholar]
- 17.Dashtbozorg B, Mendonca AM, Campilho A.. An automatic graph-based approach for artery/vein classification in retinal images. IEEE Trans Image Process. 2014;23(3):1073–1083. doi: 10.1109/TIP.2013.2263809. [DOI] [PubMed] [Google Scholar]
- 18.Akbar S, Akram MU, Sharif M, et al. Decision support system for detection of hypertensive retinopathy using arteriovenous ratio. Artif Intell Med. 2018;90:15–24. doi: 10.1016/j.artmed.2018.06.004. [DOI] [PubMed] [Google Scholar]
- 19.Hubbard LD, Brothers RJ, King WN, et al. Methods for evaluation of retinal microvascular abnormalities associated with hypertension/sclerosis in the atherosclerosis risk in communities study. Ophthalmology. 1999;106(12):2269–2280. doi: 10.1016/s0161-6420(99)90525-0. [DOI] [PubMed] [Google Scholar]
- 20.Cheung CY, Zheng Y, Hsu W, et al. Retinal vascular tortuosity, blood pressure, and cardiovascular risk factors. Ophthalmology. 2011;118(5):812–818. doi: 10.1016/j.ophtha.2010.08.045. [DOI] [PubMed] [Google Scholar]
- 21.Wang SB, Mitchell P, Liew G, et al. A spectrum of retinal vasculature measures and coronary artery disease. Atherosclerosis. 2018;268:215–224. doi: 10.1016/j.atherosclerosis.2017.10.008. [DOI] [PubMed] [Google Scholar]
- 22.Lee MJ, Deal JA, Ramulu PY, et al. Prevalence of retinal signs and association with cognitive status: the ARIC neurocognitive study. J Am Geriatr Soc. 2019;67(6):1197–1203. doi: 10.1111/jgs.15795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Deal JA, Sharrett AR, Rawlings AM, et al. Retinal signs and 20-year cognitive decline in the atherosclerosis risk in communities study. Neurology. 2018;90(13):e1158–e1166. doi: 10.1212/WNL.0000000000005205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Yaffe K, Ackerson L, Hoang TD, et al. Retinopathy and cognitive impairment in adults with CKD. Am J Kidney Dis. 2013;61(2):219–227. doi: 10.1053/j.ajkd.2012.10.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Heringa SM, Bouvy WH, van den Berg E, et al. Associations between retinal microvascular changes and dementia, cognitive functioning, and brain imaging abnormalities: a systematic review. J Cereb Blood Flow Metab. 2013;33(7):983–995. doi: 10.1038/jcbfm.2013.58. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Wu H, Wang C, Chen C, et al. Association between retinal vascular geometric changes and cognitive impairment: a systematic review and meta-analysis. J Clin Neurol. 2020;16(1):19–28. doi: 10.3988/jcn.2020.16.1.19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Cai S, Gray C, Robbins CB, et al. Optical coherence tomography angiography fractal dimension in alzheimer’s disease, mild cognitive impairment and cognitively healthy controls. Investig Ophthalmol Vis Sci. 2020;61(7):5325–5325. [Google Scholar]
- 28.Lemmens S, Devulder A, Van Keer K, et al. Systematic review on fractal dimension of the retinal vasculature in neurodegeneration and stroke: assessment of a potential biomarker. Front Neurosci. 2020;14:16. doi: 10.3389/fnins.2020.00016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Cheung CY, Ong S, Ikram MK, et al. Retinal vascular fractal dimension is associated with cognitive dysfunction. J Stroke Cerebrovasc Dis. 2014;23(1):43–50. doi: 10.1016/j.jstrokecerebrovasdis.2012.09.002. [DOI] [PubMed] [Google Scholar]
- 30.Davies SJ, Phillips L, Naish PF, et al. Peritoneal glucose exposure and changes in membrane solute transport with time on peritoneal dialysis. J Am Soc Nephrol. 2001;12(5):1046–1051. doi: 10.1681/ASN.V1251046. [DOI] [PubMed] [Google Scholar]
- 31.Bartosova M, Schaefer B, Vondrak K, et al. Peritoneal dialysis vintage and glucose exposure but not peritonitis episodes drive peritoneal membrane transformation during the first years of PD. Front Physiol. 2019;10:356. doi: 10.3389/fphys.2019.00356. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Shi J, Yu M, Sheng M.. Angiogenesis and inflammation in peritoneal dialysis: the role of adipocytes. Kidney Blood Press Res. 2017;42(2):209–219. doi: 10.1159/000476017. [DOI] [PubMed] [Google Scholar]
- 33.Gao D, Zhao ZZ, Liang XH, et al. Effect of peritoneal dialysis on expression of vascular endothelial growth factor, basic fibroblast growth factor and endostatin of the peritoneum in peritoneal dialysis patients. Nephrology. 2011;16(8):736–742. doi: 10.1111/j.1440-1797.2011.01502.x. [DOI] [PubMed] [Google Scholar]
- 34.Gulshan V, Peng L, Coram M, et al. Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. JAMA. 2016;316(22):2402–2410. doi: 10.1001/jama.2016.17216. [DOI] [PubMed] [Google Scholar]
- 35.Bellemo V, Lim ZW, Lim G, et al. Artificial intelligence using deep learning to screen for referable and vision-threatening diabetic retinopathy in Africa: a clinical validation study. Lancet Digit Health. 2019;1(1):e35–e44. doi: 10.1016/S2589-7500(19)30004-4. [DOI] [PubMed] [Google Scholar]
- 36.Huang X, Wang H, She C, et al. Artificial intelligence promotes the diagnosis and screening of diabetic retinopathy. Front Endocrinol (Lausanne). 2022;13:946915. doi: 10.3389/fendo.2022.946915. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data used to support the findings of this study are available from the corresponding author upon request.