Diabetic retinopathy (DR) is one of the leading causes of visual impairment worldwide and the principal cause of blindness among adults aged 25–74 years. Its pathogenesis is multifactorial, driven primarily by the metabolic consequences of chronic hyperglycaemia, which lead to microvascular injury, blood–retinal barrier breakdown and progressive retinal ischaemia. Conventional risk factors for the onset and progression of DR include age, duration of diabetes, glycaemic control, hypertension and dyslipidaemia. Considerable research has focused on identifying accessible biomarkers that could flag patients at high risk of DR before irreversible visual loss occurs.

Red blood cells (RBCs) possess unique rheological properties — aggregability, deformability and potential adherence to the vascular endothelium — that influence microcirculatory flow. Enhanced RBC adherence to endothelial cells has been implicated in vaso-occlusive pathology in conditions such as sickle cell disease, thalassemia and diabetes mellitus, where glycated RBC membrane proteins (notably Band 3) interact with the endothelial receptor for advanced glycation end-products (RAGE), triggering oxidative stress and further endothelial injury. Despite this mechanistic plausibility, few studies have examined RBC adhesive properties specifically in the context of DR.

Red cell distribution width (RDW), a routinely reported index of variability in circulating erythrocyte size, has emerged as a marker of chronic inflammation and oxidative stress in cardiovascular disease, heart failure and, more recently, diabetes mellitus and its microvascular complications. Elevated RDW is thought to reflect impaired erythropoiesis and accelerated erythrocyte turnover secondary to sustained oxidative stress.

A thorough review of the literature revealed no prior study assessing both RBC adhesive properties and RDW within the same population of patients with DR. This study was therefore designed to evaluate the association of these two parameters — one representing a cellular adhesion phenomenon and the other a global marker of erythrocyte heterogeneity — with the presence and severity of DR, alongside the influence of established conventional risk factors.

OBJECTIVES

1. To assess the association of RDW and RBC adherence with the presence and severity of diabetic retinopathy.
2. To assess the influence of conventional risk factors (age, duration of diabetes, glycaemic status [HbA1c], hypertension and dyslipidaemia) on the presence and severity of diabetic retinopathy.

Study design and setting: This was a case–control study conducted among inpatients and outpatients attending the medicine and ophthalmology departments of a tertiary care hospital, over a total study period of 30 months, after approval from the Institutional Ethics Committee. Study participants: Patients with type 2 diabetes mellitus aged more than 18 years were eligible. Patients younger than 18 years, those with a history of haematological disorders, iron, vitamin B12 or folate deficiency, chronic liver disease or malignancy were excluded. Written informed consent was obtained from all participants. Thirty patients were enrolled and divided into three equal groups of 10 patients each: Group I – type 2 diabetes without DR (controls); Group II – type 2 diabetes with non-proliferative DR (NPDR); Group III – type 2 diabetes with proliferative DR (PDR). As this was a pilot study, an arbitrary sample size of 30 patients (10 per group) was chosen, informed by prior studies of RDW alone (sample size ~150) and of RBC adherence alone (sample size ~40). Clinical evaluation: All patients underwent a detailed clinical interview and examination, including demographic details, duration of diabetes, mode of treatment, and co-morbidities (hypertension, dyslipidaemia, peripheral neuropathy, diabetic nephropathy, stroke and peripheral vascular disease). Fundus examination was performed by the department of ophthalmology to determine the presence and type of DR, with fundus photography performed wherever possible. Laboratory investigations: Venous blood samples were collected for a complete haemogram (including RDW), renal function tests, fasting lipid profile and HbA1c. The reference range for RDW-SD in this laboratory was 40.0–55.0 fL. RBC adherence assay: Blood was collected and RBCs separated from whole blood by centrifugation in phosphate-buffered saline. Separated RBCs were incubated on an uncoated glass surface (fabricated as a microfluidic para-film channel sandwiched between glass coverslips) for one hour. Non-adherent cells were then washed away under an initial flow of 40 µL/min, and adherent cells were subjected to a stepwise increase in flow rate (up to 600 µL/min), with video recorded at each flow rate. Images were extracted and analysed using a LabVIEW-based image-analysis program (NI Vision Assistant) to quantify the number of adherent cells at each flow rate. The flow rate (converted to shear stress) at which 50% of cells detached was taken as the measure of RBC adherence. Statistical analysis: Data were entered in Microsoft Excel and analysed using SPSS. Descriptive data were expressed as frequencies and percentages. Normally distributed continuous variables were compared across groups using one-way ANOVA (with Bonferroni post-hoc testing); non-normally distributed variables (RBC adherence) were compared using the Kruskal–Wallis test. Categorical variables were compared using the chi-square test. Correlations were assessed using Pearson's or Spearman's correlation coefficients as appropriate. A p-value <0.05 was considered statistically significant.

Baseline characteristics: The mean age of participants in Groups I, II and III was 63.2 ± 12.0, 61.8 ± 10.1 and 57.4 ± 11.9 years respectively, with no significant difference between groups (ANOVA, p=0.10). Of the 30 patients, 17 (56.7%) were male and 13 (43.3%) were female; gender distribution did not differ significantly across groups (chi-square, p=0.98).

The duration of diabetes was significantly longer in patients with DR: 70% of patients with PDR and 60% with NPDR had diabetes for more than 10 years, compared with only 10% of patients without DR (ANOVA, p=0.005). Among treatment modalities, patients without DR were predominantly managed with combination oral hypoglycaemic agents (90%), whereas insulin use (alone or combined with oral agents) was more frequent in the NPDR (70%) and PDR (80%) groups.

Hypertension was present in all patients with NPDR and PDR (100%), compared with 30% of patients without DR — a highly significant association (chi-square, p<0.001). Dyslipidaemia was present in 50% of patients in Groups I and II and 70% in Group III, but this difference was not statistically significant (p=0.08). Other diabetes-related complications — peripheral neuropathy, diabetic nephropathy and stroke — were more frequent in patients with DR, although this association did not reach statistical significance (p=0.05). HbA1c was numerically higher in patients with DR (>11% in 40–50% of Groups II and III) but did not differ significantly between the three groups (ANOVA, p=0.05).

Table 1. Comparison of RDW across study groups

RDW rose progressively from Group I to Group III, and this difference was highly significant. Post-hoc (Bonferroni) analysis confirmed significant differences between Group I and Group III, and between Group II and Group III. RDW showed a moderate, significant positive correlation with duration of diabetes (r=0.38, p=0.04) and with serum creatinine (r=0.38, p=0.04), and hypertension had a significant effect on RDW (independent t-test, p=0.001). The correlation between RDW and HbA1c was weak and non-significant (r=0.15, p=0.42), and diabetes-related complications as a group had no significant independent effect on RDW.

Table 2. Comparison of RBC adherence across study groups

Unlike RDW, RBC adherence did not differ significantly between the three groups (Kruskal–Wallis, p=0.40, and p=0.64/0.60 on repeat analysis). Duration of diabetes showed a weak, non-significant correlation with RBC adherence (r=0.10, p=0.60), as did HbA1c (r=0.18, p=0.33). Neither hypertension, treatment modality, nor other diabetes-related complications had a significant effect on RBC adherence.

In this case–control study of patients with type 2 diabetes mellitus, we found no significant association between age or gender and the presence or severity of DR, consistent with prior reports. Duration of diabetes, however, emerged as a robust and significant risk factor for DR, in keeping with large cohort studies such as those by Forga et al. and the epidemiological analysis by Younis et al., both of which demonstrated that the risk of DR rises steadily with increasing diabetes duration. HbA1c did not differ significantly between groups in our cohort, in contrast to landmark trials such as the ACCORD trial and the UKPDS, which demonstrated that intensive glycaemic control reduces the incidence and progression of retinopathy. This discrepancy likely reflects the cross-sectional design of our study, which captured glycaemic status at a single time point rather than cumulative glycaemic exposure over the years preceding DR onset — a limitation also noted by other single-timepoint studies. Hypertension showed a strong and highly significant association with the presence and severity of DR, corroborating findings from the UKPDS and the ABCD trial, both of which showed that tighter blood pressure control reduces the risk of retinopathy progression. Dyslipidaemia, although numerically more common in the PDR group, was not significantly associated with DR severity in our study, differing from the DCCT/EDIC cohort and the Early Treatment Diabetic Retinopathy Study, which found positive associations between elevated lipid levels and retinopathy-related outcomes; this discrepancy may reflect our limited sample size. RDW: Our principal finding was a significant, stepwise increase in RDW with increasing severity of DR, with the highest values seen in patients with PDR. This is consistent with the findings of Yin et al. and Kurtul et al., who reported that higher RDW was associated with poorer glycaemic control and more advanced diabetic retinopathy. In our cohort, RDW also correlated significantly with serum creatinine and hypertension, mirroring observations by Ece Kurtul et al. and Magri and Fava, who linked elevated RDW to diabetic nephropathy and other microvascular complications. Chronic inflammation and oxidative stress — both hallmarks of poorly controlled diabetes — are thought to underlie this relationship by damaging erythrocytes and shortening their survival, thereby widening the distribution of circulating red cell volumes. Taken together, these findings support RDW as an inexpensive, widely available marker that may help identify patients at higher risk of both microvascular retinal and renal complications. RBC adherence: In contrast, RBC adherence showed no significant association with the presence or severity of DR in our study. This finding differs from earlier work by Wautier et al., who demonstrated significantly higher RBC–endothelial adhesion in diabetic patients compared with controls, and correlated adhesion intensity with the extent of vascular complications and glycated haemoglobin. Our contrasting result may be explained by the small sample size of this pilot study, methodological differences in the adherence assay (a flow-based microfluidic assay against an uncoated glass surface rather than a cultured endothelial monolayer), or the possibility that RBC adhesive changes are more closely linked to systemic vascular disease in general rather than to retinal microvascular disease specifically. Further studies with larger sample sizes and endothelial cell-based adhesion assays are warranted to clarify this relationship. Strengths and limitations: To our knowledge, this is among the first studies to assess both RDW and RBC adherence within the same cohort of patients with varying severity of DR. Limitations include the small sample size (10 patients per group), the cross-sectional design (precluding assessment of cumulative glycaemic exposure), and the absence of a standardised endothelial substrate for the RBC adherence assay. Larger, prospective, multi-centric studies are needed to validate these findings and to determine whether RDW independently predicts incident DR or its progression over time.

There was no significant association between RBC adherence and diabetic retinopathy in this cohort. However, RDW showed a significant, stepwise positive correlation with the presence and severity of diabetic retinopathy, and also correlated with hypertension and diabetic nephropathy. Duration of diabetes and hypertension were significant conventional risk factors for DR, whereas age, gender and dyslipidaemia were not significantly associated in this cohort. These findings suggest that higher RDW may serve as a strong, independent and readily accessible marker of poor glycaemic control and progression of diabetic retinopathy, and may warrant further evaluation as a screening adjunct in patients with type 2 diabetes mellitus.

Declarations

Conflict of interest: None declared.

Funding: None declared.

Acknowledgements: The authors thank the Departments of General Medicine and Ophthalmology, and the biostatistician, for their contribution to this work.

1. Tarr JM, Kaul K, Chopra M, Kohner EM, Chibber R. Pathophysiology of Diabetic Retinopathy. ISRN Ophthalmol. 2013;2013:343560.
2. Kohner EM, Patel V, Rassam SMB. Role of Blood Flow and Impaired Autoregulation in the Pathogenesis of Diabetic Retinopathy. Diabetes. 1995;44(6):603–7.
3. Duh EJ, Sun JK, Stitt AW. Diabetic retinopathy: current understanding, mechanisms, and treatment strategies. JCI Insight. 2017;2(14):e93751.
4. Wang W, Lo ACY. Diabetic Retinopathy: Pathophysiology and Treatments. Int J Mol Sci. 2018;19(6):1816.
5. Simó R, Stitt AW, Gardner TW. Neurodegeneration in diabetic retinopathy: does it really matter? Diabetologia. 2018;61(9):1902–12.
6. Wautier ML, Chappey O, Corda S, Stern DM, Schmidt AM, Wautier JL. Activation of NADPH oxidase by AGE links oxidant stress to altered gene expression via RAGE. Am J Physiol Endocrinol Metab. 2001;280(5):E685-694.
7. Wautier JL, Wautier MP. Increased adhesion of erythrocytes to endothelial cells in diabetes mellitus and its relation to vascular complications. N Engl J Med. 1981;305:237–42.
8. Al-Kindi SG, Refaat M, Jayyousi A, Asaad N, Al Suwaidi J, Abi Khalil C. Red Cell Distribution Width Is Associated with All-Cause and Cardiovascular Mortality in Patients with Diabetes. Biomed Res Int. 2017;2017:5843702.
9. Ece Kurtul B, İnal B, Altıaylık Özer P, Kabatas E. The Correlation between Red Cell Distribution Width and Diabetic Retinopathy in Patients with Type 2 Diabetes Mellitus. Türkiye Klinikleri Oftalmoloji Dergisi. 2016.
10. Malandrino N, Wu WC, Taveira TH, et al. Association between red cell distribution width and macrovascular and microvascular complications in diabetes. Diabetologia. 2012;55:226–35.
11. Yin Y, Ye S, Wang H, Li B, Wang A, Yan W, et al. Red blood cell distribution width and the risk of being in poor glycemic control among patients with established type 2 diabetes. Ther Clin Risk Manag. 2018.
12. Magri CJ, Fava S. Red blood cell distribution width and diabetes-associated complications. Diabetes Metab Syndr. 2014;8(1):13–7.
13. Forga L, Goñi MJ, Ibáñez B, Cambra K, García-Mouriz M, Iriarte A. Influence of Age at Diagnosis and Time-Dependent Risk Factors on the Development of Diabetic Retinopathy in Patients with Type 1 Diabetes. J Diabetes Res. 2016;2016:9898309.
14. Chew EY, Ambrosius WT, Davis MD, et al. Effects of medical therapies on retinopathy progression in type 2 diabetes (ACCORD Eye Study). N Engl J Med. 2010;363(3):233–44.
15. Stratton IM, Kohner EM, Aldington SJ, et al. UKPDS 50: risk factors for incidence and progression of retinopathy in Type 2 diabetes over 6 years from diagnosis. Diabetologia. 2001;44:156–63.
16. Younis N, Broadbent DM, Vora JP, Harding SP; Liverpool Diabetic Eye Study. Incidence of sight-threatening retinopathy in patients with type 2 diabetes in the Liverpool Diabetic Eye Study: a cohort study. Lancet. 2003;361:195–200.