Introduction: The aim of the study was to identify retinal microvascular changes using optical coherence tomography angiography (OCTA) in type 2 diabetes (T2D) patients with preclinical retinopathy identified by ultra-widefield fundus photography (UWF-FP). Methods: This is a cross-sectional observational study. All patients underwent UWF-FP 200° examinations with OPTOS California (Optos, Dunfermline, UK) and Cirrus AngioPlex® spectral-domain (SD)-OCTA 3 × 3 mm acquisitions (ZEISS, Dublin, CA, USA). The absence of visible lesions was identified using UWF-FP. Results: One hundred and ninety three eyes of individuals with T2D with no visible lesions in the fundus and identified in a screening setting were included in the study. Skeletonized vessel density (SVD), perfusion density (PD), and areas of capillary nonperfusion (CNP) values on SD-OCTA were significantly decreased when compared with healthy population (p < 0.001). SVD and CNP values of the superficial capillary plexus (SCP) were more frequently decreased (35% and 45%, respectively) than SVD values of the deep capillary plexus (DCP) (9% and 15%, respectively), demonstrating that diabetic microvascular changes occur earlier in the SCP than in the DCP. The ischemic phenotype, identified by a definite decrease in SVD or CNP in the SCP may, therefore, be identified in the preclinical stage of diabetic retinal disease. Conclusions: Retinal capillary nonperfusion detected by OCTA metrics of SVD and CNP can be identified in the central retina in eyes with T2D before development of visible lesions in the retina. Our findings confirm the relevance of OCTA to identify macular microvascular changes in the initial stages of diabetic retinopathy, allowing the identification of its ischemic phenotype very early in the disease process.

Diabetes mellitus is now regarded as a global epidemic. In 2017, it was estimated that 425 million people worldwide were affected by diabetes and this number is expected to rise to 692 million by 2045 [1]. Consequently, the occurrence of diabetic retinopathy (DR) is also expected to escalate. It has been estimated that a third of the people with diabetes have signs of DR but only 10% of these will present signs of vision-threatening retinopathy, one of the leading causes of blindness in the working age population [2]. It is crucial, therefore, to identify which patients are at risk of progression and development of vision loss in their lifetime.

Optical coherence tomography angiography (OCTA) is an imaging technique that allows non-invasive visualization and quantification of the retinal vasculature in the different capillary networks. It is nowadays essential in the study of vascular diseases like DR, to detect changes in skeletonized vessel and perfusion densities which are indicators of microvascular changes and capillary closure. Recent studies [3, 4] have shown that OCTA and optical coherence tomography (OCT) can detect and distinguish different stages and mechanisms of DR progression, improving the characterization and follow-up of these patients.

Monitoring response time to treatment through serial examinations of the retina appears now to be an attainable goal, offering the opportunity to lower the risk of diabetic damage to the eye as well as other microvascular and macrovascular end-organ complications [5, 6]. Identification of preclinical biomarkers of microvascular disease in the diabetic retina is clearly a desirable goal. The hope is that these markers will allow clinicians an earlier diagnosis and stratification of patients according to their risk of end-organ complications.

OCTA, taking advantage of high-speed OCT and image processing algorithms, offers the opportunity to identify in the clinical environment the earliest microvascular alterations occurring in the diabetic retina [7]. In this study, we have examined individuals with type 2 diabetes (T2D) in a screening setting, looking at microvascular and neurodegenerative changes occurring in the preclinical stage of diabetic retinal disease, i.e., before lesions are visible in the retina using ophthalmoscopy and fundus photography.

Study Population

This is a cross-sectional observational study of patients with T2D. The patients included underwent a full ophthalmological examination during their visit which included an ophthalmoscopic examination by slit lamp performed by an ophthalmologist, best-corrected visual acuity, iCare tonometry, ultra-widefield fundus photography (UWF-FP), OCT, and OCTA at the Association for Innovation and Biomedical Research on Light and Image (AIBILI) within the scope of a screening program for patients with T2D (age mean ± standard deviation (SD): 66.04 ± 9.99 years) that occurred between 2020 and 2022. All patients signed an informed consent, and all study procedures occurred in accordance with the terms of the Declaration of Helsinki.

Seventy-six eyes of healthy individuals (age mean ± SD: 68.32 ± 3.50 years) and 193 eyes of patients with T2D without visible DR were considered for this analysis. The following exclusion criteria were implemented: (1) presence of vitreomacular traction or epiretinal membrane, (2) presence of macular edema, (3) glaucoma, (4) active intraocular inflammation, and (5) media opacities or other conditions that could compromise the quality of the images. Demographic data and medical history information including diabetes duration, hemoglobin levels, concomitant medication or treatment for diabetes, blood pressure, systemic diseases such as hypertension, history of strokes, or heart attacks along with smoking and alcohol habits were provided by the primary health care units with consent of the patients as part of the established screening program.

Color Fundus Photography

All patients underwent a 200° non-mydriatic UWF-FP with OPTOS California (Optos plc, Dunfermline, UK). This system is a confocal scanning laser ophthalmoscope with 2 light sources (532 nm green laser and a 633 nm red laser) with an ellipsoidal mirror to enable a 200° pseudo-color image of the retina fundus acquired in 0.4 s [8, 9]. After acquisition, UWF-FP images were analyzed by the OPTOS Advance Review Software (version 4.3.31). Any diabetic lesion was considered, such as microaneurysms, hemorrhages, intraretinal microvascular abnormalities, hard or soft exudates, venous beading, or presence of neovascularization. In case of uncertainty, the graders used the color channel separation feature provided by the review software to correctly identify the presence of retinal lesions.

OCT and OCTA Acquisitions

Retinal structural and vascular information was obtained from all patients using the Cirrus HD-OCT 5000 AngioPlex (Zeiss, Dublin, CA, USA). Retina structure was assessed using the macular cube 512 × 128 acquisition protocol, composed of 128 B-scans of 512 A-scans, to collect central retinal thickness (CRT) and ganglion cell layer + inner plexiform layer (GCL + IPL) average layer thickness using the standard device reports. To collect thickness measurements from the outer segment layer (OS), the structural OCT data were segmented using a segmentation software developed in-house [10]. The results from the automated segmentation were validated by one masked grader.

Retinal vasculature was imaged using angiography 3 × 3 mm acquisition protocol, which consists of 245 clusters of 4 B-scan repetitions, where each B-scan is composed of 245 A-scans. The OCTA acquisitions were processed by the Carl Zeiss Meditec Density Exerciser (version: 10.0.12787) to compute OCTA metrics, namely, foveal avascular zone (FAZ), skeletonized vessel density (SVD), and perfusion density (PD) for the superficial and deep capillary plexus (SCP and DCP) and full retina (FR). The area of CNP was identified by measuring the area of intercapillary spaces (AIS) calculated using the methodology previously described by our group [11]. This methodology applies morphological operations to the binary enface slabs of the SCP, DCP, and FR to identify abnormal spaces between capillaries.

An experienced grader reviewed the OCT and OCTA acquisitions for quality assessment. FAZ measurements with poor delimitation were discarded from the FAZ analysis. In addition, the grader reviewed each inner ring quadrant for the SCP, DCP, and FR to ensure reliable metrics (Fig. 1). Criteria for measurement reliability were well-focused capillary networks and absence of cropped scans, vessel duplication, and opaque artifacts. Single low-quality inner ring quadrants were not considered in the analysis. If more than one inner ring quadrant was classified as low quality then all measurements for that slab were discarded.

Fig. 1.

a, b Capillary nonperfusion areas for the superficial capillary plexus (SCP) measured using areas of intercapillary spaces (AIS) with the ETDRS area grid overlaid in gray. On the left, the SCP layer from a 70-year-old healthy control volunteer, with identified AIS in white. On the right, the SCP layer from a 68-year-old diabetic patient without visible color fundus photography lesions in the retina, with AIS identified in white.

Fig. 1.

a, b Capillary nonperfusion areas for the superficial capillary plexus (SCP) measured using areas of intercapillary spaces (AIS) with the ETDRS area grid overlaid in gray. On the left, the SCP layer from a 70-year-old healthy control volunteer, with identified AIS in white. On the right, the SCP layer from a 68-year-old diabetic patient without visible color fundus photography lesions in the retina, with AIS identified in white.

Close modal

Characterization of DR Phenotypes

The three different DR phenotypes for nonproliferative DR, previously described by our group [4, 12], were identified based on the values of SVD in the SCP and CRT according to the following rules: phenotype C was identified by decreased values of SVD of SCP ≥2 standard deviations (SDs) of a reference healthy population, with or without increased CRT; phenotype B was characterized by SVD decreased in the SCP <2 SD and increased values of CRT (≥260 µm in women and ≥275 µm in men); phenotype A was identified by SVD decrease in the SCP <2 SD of a reference healthy population and normal CRT values (<260 µm in women and <275 µm in men). CRT reference values presented in this study are the reference for Zeiss Cirrus HD-OCT 5000.

Statistical Analysis

Statistical analysis was performed using Stata 16.1 (StataCorp LLC, College Station, TX, USA), and p values <0.05 were considered statistically significant. Normal distribution was assessed with the Shapiro-Wilk test and graphically verified by histogram distributions. Demographic, systemic, and ocular characteristics were summarized as means and corresponding SDs for continuous variables. The Mann-Whitney U test was performed to compare control population with diabetic patients without visible retinal lesions.

Data from 193 eyes from individuals with T2D and with preclinical retinopathy, i.e., without visible lesions in the fundus on ophthalmoscopy and UWF-FP were included in the analysis. Patients’ demographic data are summarized in Table 1 comparing diabetic patients with preclinical retinal disease and a control population matched for age and sex. The eyes with diabetic preclinical retinal disease showed significant decrease in SVD and PD, and increase in CNP in both SCP (p < 0.001) and DCP (p < 0.006), in the central inner ring, demonstrating the presence of retinal nonperfusion in diabetic eyes without visible lesions. Areas of CNP were measured using AIS to identify retinal CNP in these eyes with preclinical retinopathy. No significant changes were detected in FAZ metrics, namely, area, perimeter, or circularity.

Table 1.

Comparison of demographic, systemic, and ocular characteristics between healthy eyes and diabetic patients without visible lesions

Healthy eyes (N = 76)Diabetic eyes without retinal lesions (N = 193)p value
Demographic characteristics, mean±SD 
 Age, years 68.32±3.50 66.04±9.99 0.112 
 Diabetes duration, years 9.31±5.83  
Systemic characteristics, mean±SD 
 Diastolic blood pressure, mm Hg 75.47±9.10  
 Systolic blood pressure, mm Hg 135.26±14.06  
 HbA1c (%) 6.86±1.05  
Ocular characteristics, mean±SD 
 Best-corrected visual acuity, letters 81.24±6.98  
 Intra ocular pressure, mm Hg 14.69±3.20  
 Central retinal thickness, µm 270.52±18.07 267.58±21.92 0.284 
 GCL + IPL thickness, µm 82.53±5.71 78.86±8.08 <0.001 
 OS CSF thickness, µm 46.07±3.22 40.12±6.31 <0.001 
 FAZ area, mm2 0.23±0.10 0.25±0.08 0.265 
 FAZ perimeter, mm 2.06±0.46 2.17±0.40 0.070 
 FAZ circularity, a.u. 0.66±0.07 0.65±0.09 0.372 
 SVD – InR – SCP, mm−1 22.11±0.71 20.98±1.45 <0.001 
 SVD – InR – DCP, mm−1 16.98±2.01 16.14±2.48 0.006 
 SVD – InR – FR, mm−1 23.53±0.69 22.69±1.29 <0.001 
 PD – InR – SCP, a.u. 0.40±0.01 0.38±0.02 <0.001 
 PD – InR – DCP, a.u. 0.32±0.03 0.31±0.04 0.006 
 PD – InR – FR, a.u. 0.42±0.01 0.40±0.02 <0.001 
 CNP – SCP, ×1,000 a.u. 11.94±3.24 19.74±10.23 <0.001 
 CNP – DCP, ×1,000 a.u. 29.05±9.93 36.48±17.06 0.001 
 CNP – FR, ×1,000 a.u. 7.66±2.54 12.01±6.67 <0.001 
Healthy eyes (N = 76)Diabetic eyes without retinal lesions (N = 193)p value
Demographic characteristics, mean±SD 
 Age, years 68.32±3.50 66.04±9.99 0.112 
 Diabetes duration, years 9.31±5.83  
Systemic characteristics, mean±SD 
 Diastolic blood pressure, mm Hg 75.47±9.10  
 Systolic blood pressure, mm Hg 135.26±14.06  
 HbA1c (%) 6.86±1.05  
Ocular characteristics, mean±SD 
 Best-corrected visual acuity, letters 81.24±6.98  
 Intra ocular pressure, mm Hg 14.69±3.20  
 Central retinal thickness, µm 270.52±18.07 267.58±21.92 0.284 
 GCL + IPL thickness, µm 82.53±5.71 78.86±8.08 <0.001 
 OS CSF thickness, µm 46.07±3.22 40.12±6.31 <0.001 
 FAZ area, mm2 0.23±0.10 0.25±0.08 0.265 
 FAZ perimeter, mm 2.06±0.46 2.17±0.40 0.070 
 FAZ circularity, a.u. 0.66±0.07 0.65±0.09 0.372 
 SVD – InR – SCP, mm−1 22.11±0.71 20.98±1.45 <0.001 
 SVD – InR – DCP, mm−1 16.98±2.01 16.14±2.48 0.006 
 SVD – InR – FR, mm−1 23.53±0.69 22.69±1.29 <0.001 
 PD – InR – SCP, a.u. 0.40±0.01 0.38±0.02 <0.001 
 PD – InR – DCP, a.u. 0.32±0.03 0.31±0.04 0.006 
 PD – InR – FR, a.u. 0.42±0.01 0.40±0.02 <0.001 
 CNP – SCP, ×1,000 a.u. 11.94±3.24 19.74±10.23 <0.001 
 CNP – DCP, ×1,000 a.u. 29.05±9.93 36.48±17.06 0.001 
 CNP – FR, ×1,000 a.u. 7.66±2.54 12.01±6.67 <0.001 

Bold values represent statistically significant alterations with p < 0.05 using Mann-Whitney U test to compare healthy control group with diabetic group without visible lesions.

N, number of participants; SD, standard deviation; HbA1c, glycated hemoglobin; GCL + IPL, ganglion cell layer + inner plexiform layer; OS, outer segment; CSF, central subfield; FAZ, foveal avascular zone; SVD, skeletonized vessel density; PD, perfusion density; a.u., arbitrary units; InR, inner ring; SCP, superficial capillary plexus; DCP, deep capillary plexus; FR, full retina; CNP, areas of capillary nonperfusion.

Thirty-five percent of the eyes included in the analyses showed definite retinal nonperfusion in SCP (defined by decreased SVD values ≥2 SD of the control population). Definite retinal nonperfusion in the DCP (decrease ≥2 SD of the control population) was much less frequent, occurring only in 9% of the eyes. Similar results were observed when using areas of retinal CNP (measured using AIS) as an OCTA metric. Increases in CNP in the SCP were identified in the eyes with definite closure on SCP but even with higher frequency (45%) (Table 2).

Table 2.

Characterization of groups according to decreases in SVD or CNP

No visible retinal lesionsNo visible retinal lesions
Decrease in SVD in SCP ≥2 SD of healthy controls 67/193 35% Increase in CNP in SCP ≥2 SD of healthy controls 86/193 45% 
Decrease in SVD in DCP ≥2 SD of healthy controls 18/193 9% Increase in CNP in DCP ≥2 SD of healthy controls 29/193 15% 
Decrease in SVD in SCP <2 SD of healthy controls and decrease in SVD in DCP ≥2 SD of healthy controls 4/193 2% Increase in CNP in SCP <2 SD of healthy controls and increase in CNP in DCP ≥2 SD of healthy controls 6/193 3% 
No visible retinal lesionsNo visible retinal lesions
Decrease in SVD in SCP ≥2 SD of healthy controls 67/193 35% Increase in CNP in SCP ≥2 SD of healthy controls 86/193 45% 
Decrease in SVD in DCP ≥2 SD of healthy controls 18/193 9% Increase in CNP in DCP ≥2 SD of healthy controls 29/193 15% 
Decrease in SVD in SCP <2 SD of healthy controls and decrease in SVD in DCP ≥2 SD of healthy controls 4/193 2% Increase in CNP in SCP <2 SD of healthy controls and increase in CNP in DCP ≥2 SD of healthy controls 6/193 3% 

SD, standard deviation; SVD, skeletonized vessel density; SCP, superficial capillary plexus; DCP, deep capillary plexus; CNP, capillary nonperfusion.

Retinal neurodegeneration identified by thinning of the GCL + IPL was present in the eyes of patients with T2D without visible lesions (p < 0.001). This increased thinning of the GCL + IPL is associated with the presence of retinal nonperfusion identified by SVD and CNP in the SCP (p < 0.001) (Table 3). The presence of retinal nonperfusion of the SCP is also associated with photoreceptor damage evidenced by thinning of the OS layer of the retina (SVD: p = 0.030, CNP: p = 0.026). The systemic parameters show that the eyes examined belong to patients under relatively good metabolic control, presenting values of HbA1c and blood pressure within the normal range.

Table 3.

Retinal CNP associated with ganglion cell thinning and photoreceptor damage measured

Healthy controlsDiabetic patients without decrease in SVD in SCP ≥2 SD (m±SD)Diabetic patients with decrease in SVD in SCP ≥2 SD (m±SD)p value (DR patients with vs. without decrease in SVD in SCP)
According to SVD 
 GCL + IPL  thickness, µm 82.53±5.71 80.33±7.69 75.97±8.12 p < 0.001 
  Versus healthy controls p = 0.057 Versus healthy controls p < 0.001  
 OS layer  thickness, µm 46.07±3.22 40.80±6.37 38.85±6.10 p = 0.030 
  Versus healthy controls p < 0.001 Versus healthy controls p < 0.001  
Healthy controlsDiabetic patients without decrease in SVD in SCP ≥2 SD (m±SD)Diabetic patients with decrease in SVD in SCP ≥2 SD (m±SD)p value (DR patients with vs. without decrease in SVD in SCP)
According to SVD 
 GCL + IPL  thickness, µm 82.53±5.71 80.33±7.69 75.97±8.12 p < 0.001 
  Versus healthy controls p = 0.057 Versus healthy controls p < 0.001  
 OS layer  thickness, µm 46.07±3.22 40.80±6.37 38.85±6.10 p = 0.030 
  Versus healthy controls p < 0.001 Versus healthy controls p < 0.001  
Healthy controlsDiabetic patients without increase in CNP in SCP ≥2 SD (m±SD)Diabetic patients with increase in CNP in SCP ≥2 SD (m±SD)p value (DR patients with vs. without decrease in SVD in SCP)
According to areas of CNP  
 GCL + IPL  thickness, µm 82.53±5.71 81.53±6.95 75.41±8.17 p < 0.001 
  Versus healthy controls p = 0.347 Versus healthy controls p < 0.001  
 OS layer  thickness, µm 46.07±3.22 40.95±6.49 39.06±5.95 p = 0.026 
  Versus healthy controls p < 0.001 Versus healthy controls p < 0.001  
Healthy controlsDiabetic patients without increase in CNP in SCP ≥2 SD (m±SD)Diabetic patients with increase in CNP in SCP ≥2 SD (m±SD)p value (DR patients with vs. without decrease in SVD in SCP)
According to areas of CNP  
 GCL + IPL  thickness, µm 82.53±5.71 81.53±6.95 75.41±8.17 p < 0.001 
  Versus healthy controls p = 0.347 Versus healthy controls p < 0.001  
 OS layer  thickness, µm 46.07±3.22 40.95±6.49 39.06±5.95 p = 0.026 
  Versus healthy controls p < 0.001 Versus healthy controls p < 0.001  

DR, diabetic retinopathy; SVD-skeletonized vessel density; SCP, superficial capillary plexus; SD, standard deviation; m, mean; GCL + IPL, ganglion cell layer + inner plexiform layer; OS, outer segment; CNP, capillary nonperfusion; AISs, area of intercapillary spaces.

Eyes of individuals with T2D without visible lesions in the fundus corresponding to preclinical retinopathy show the presence of retinal capillary closure or capillary nonperfusion identified by definite decrease in SVD, PD, and increased CNP occurring early and predominantly in the SCP. These findings point to capillary closure or decreased blood flow in the SCP as the initial microvascular alteration occurring in the retina in T2D, confirming the results of De Carlo et al. [7].

The identification of retinal capillary nonperfusion as the initial microvascular damage in diabetic individuals without visible lesions in the fundus has been reported before [13‒18]. Retinal capillary nonperfusion is shown to occur in variable degrees and appears to be a highly relevant microvascular alteration in the diabetic retina. One of the most important lesions in experimental DR is the occurrence of closed or acellular capillaries. It may represent a failed response to the photoreceptor and neuronal changes caused by the hostile hyperglycemic environment [19]. Indeed, relative tissue hypoxia has consistently been considered as the most probable primary inciting event in the development of the retinal lesions in diabetes.

Development of retinal capillary nonperfusion appears to be initiated at the SCP involving only the DCP later and may result from a failed microvascular autoregulation response. This capillary nonperfusion appears to be the precursor alteration that leads to collateral formation and development of the thoroughfare channels proposed by Cogan and Kuwabara [20]. These microvascular changes lead to progressive increase in the areas of retinal nonperfusion which characterize the ischemic phenotype associated with progression to the major vision-threatening complications of diabetes [21]. Retinal capillary nonperfusion of the SCP in preclinical retinopathy may therefore be the biomarker that identifies the eyes at risk of progression to clinical retinopathy and potential vision loss.

It is interesting to note that contrary to previous reports, the FAZ does not appear to be reliably affected in earliest stages of DR. It is also particularly relevant that retinal CNP is associated with photoreceptor damage, identified by thinning of the OS retinal layer and presence of neurodegenerative changes demonstrated by thinning of the GCL + IPL layers. These observations indicate an early alteration of the neurovascular coupling unit in diabetic retinal disease and suggest an initial toxic neurodegenerative response to the chronic hyperglycemia of diabetes [22]. Previous studies have documented subtle changes in color perception and contrast sensitivity in diabetic eyes without clinical retinopathy [23, 24].

It is important to keep in mind that diabetes is not a single disease, but rather a group of conditions broadly categorized by a single diagnostic criterion, hyperglycaemia, on which disparate metabolic derangements converge [25]. Increasingly, there is evidence suggesting that T2D – the predominant diabetes subtype making up 90–95% of cases – is itself heterogeneous in terms of both mechanisms of action and relationships with health outcomes. Clustering approaches using clinical or genetic biomarkers have identified subtypes of T2D that are clinically distinct and differentially associated with diabetic complications [26, 27].

Our group has proposed the characterization of DR progression in three different phenotypes C, B, and A, according to the presence of definite capillary nonperfusion, phenotype C; presence of subclinical macular edema without definite CNP, phenotype B; and presence of neurodegeneration without edema or capillary nonperfusion, phenotype A [12]. It would be of particular interest to explore if the presence of phenotype C, associated with ischemia and identified by the presence of ≥2 SD decrease in SVD in the central macula determined very early in diabetic retinal disease may offer the opportunity to identify very early in the disease process the eyes at risk for progression. Phenotype C, i.e., presence of definite retinal CNP, may indeed identify very early the eyes at risk.

Similarly, phenotypes B and A may also be identifiable very early in the disease process, offering the possibility of early intervention directed at the dominant disease pathway, ischemia, neurodegeneration, and edema [12]. Following previous published work and taking into consideration that measurement of SVD on SCP is obtained directly from the instrument (Zeiss Cirrus AngioPlex device), we report on SVD on SCP, and use it to define the ischemic phenotype, although in this study, CNP measured by AIS procedure appears to be more sensitive to identify retinal CNP.

The Optos California UWF-FP device presents some limitations since it employs warping to represent the concave retina in the imaging plane, resulting in significant different magnifications between the central and peripheral retina. It is also common to image patient’s eyelashes which may obfuscate findings in the superior or inferior peripheral retina. In addition, it represents the retinal fundus in pseudo-color, and special care must be taken to identify retinal abnormalities. Nonetheless, several independent groups have shown that devices using the same technology are able to provide good agreement in ETDRS severity grading when compared to the standard 30° color fundus photography procedure [28‒30].

Our findings may be particularly relevant for improved management of diabetic retinal disease. Glucose control may be vital in delaying or preventing the onset of DR but definite retinal capillary nonperfusion once established may be the point of no return, a stage where even if ideal glucose control is achieved the retinopathy continues to progress [31]. This early stage could be the first evidence of definitive retinal CNP in central retina, occurring before any clinically detectable pathology. Furthermore, monitoring the rate of progression of retinal CNP may be the most appropriate monitor of DR progression and development of vision-threatening complications.

The tenets of the Declaration of Helsinki were followed, and approval was obtained from the AIBILI – Association for Innovation and Biomedical Research on Light and Image Ethics Committee for Health with the number CEC/009/17. A written informed consent was signed by each participant, agreeing to participate in the study, after all procedures were explained.

Torcato Santos, Ana Rita Santos, Ana Catarina Almeida, Ana Cláudia Rocha, Débora Reste-Ferreira, Inês Marques, António Cunha-Vaz Martinho, and Luís Mendes declare no conflicts of interest. Katharina Foote is a Carl Zeiss Meditec employee. José Cunha-Vaz reports grants from Carl Zeiss Meditec, Bayer and Boehringer Ingelheim and is a consultant for Alimera Sciences, Bayer, Boehringer Ingelheim, Carl Zeiss Meditec and Roche.

This work was supported by AIBILI and the Fundo de Inovação, Tecnologia e Economia Circular (FITEC)—Programa Interface (FITEC/CIT/2018/2).

Torcato Santos, Ana Rita Santos, Ana Catarina Almeida, Ana Cláudia Rocha, Débora Reste-Ferreira, Inês Marques, António Cunha-Vaz Martinho, and Luís Mendes collected data, analyzed, interpreted, wrote, reviewed and edited the manuscript. Katharina Foote analyzed and critically reviewed the manuscript. José Cunha-Vaz assisted in the analysis and interpretation of the data, wrote the manuscript. José Cunha-Vaz is the guarantor of this work and, as such, had full access to all the data in the study and took responsibility for the integrity of the data and the accuracy of the data analysis. All authors have read and agreed to the published version of the manuscript.

Additional Information

Torcato Santos and Ana Rita Santos contributed equally to this work.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.

1.
Federation
ID
International diabetes federation. DF diabetes atlas
.
2017
.
2.
Ogurtsova
K
,
da Rocha Fernandes
JD
,
Huang
Y
,
Linnenkamp
U
,
Guariguata
L
,
Cho
NH
.
IDF Diabetes Atlas: global estimates for the prevalence of diabetes for 2015 and 2040
.
Diabetes Res Clin Pract
.
2017
;
128
:
40
50
.
3.
Marques
IP
,
Alves
D
,
Santos
T
,
Mendes
L
,
Lobo
C
,
Santos
AR
.
Characterization of disease progression in the initial stages of retinopathy in type 2 diabetes: a 2-year longitudinal study
.
Invest Ophthalmol Vis Sci
.
2020
;
61
(
3
):
20
.
4.
Ribeiro
L
,
Marques
IP
,
Santos
T
,
Carvalho
S
,
Santos
AR
,
Mendes
L
.
Characterization of 2-year progression of different phenotypes of nonproliferative diabetic retinopathy
.
Ophthalmic Res
.
2022
;
66
(
1
):
228
37
.
5.
Diabetes ControlComplications Trial Research Group
.
Progression of retinopathy with intensive versus conventional treatment in the diabetes control and complications trial
.
Ophthalmology
.
1995
;
102
(
4
):
647
61
.
6.
UK Prospective Diabetes Study (UKPDS) Group
.
Intensive blood-glucose control with sulfonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes
.
Lancet
.
1998
;
352
(
9131
):
837
53
.
7.
De Carlo
TE
,
Chin
AT
,
Bonini Filho
MA
,
Adhi
M
,
Branchini
L
,
Salz
DA
.
Detection of microvascular changes in eyes of patients with diabetes but not clinical diabetic retinopathy using optical coherence tomography angiography
.
Retina
.
2015
;
35
(
11
):
2364
70
.
8.
Kato
Y
,
Inoue
M
,
Hirakata
A
.
Quantitative comparisons of ultra-widefield images of model eye obtained with Optos® 200Tx and Optos® California
.
BMC Ophthalmol
.
2019
;
19
(
1
):
115
.
9.
Nagiel
A
,
Lalane
RA
,
Sadda
SR
,
Schwartz
SD
.
Ultra-widefield fundus imaging: a review of clinical applications and future trends
.
Retina
.
2016
;
36
(
4
):
660
78
.
10.
Marques
IP
,
Madeira
MH
,
Messias
AL
,
Santos
T
,
Martinho
AC-V
,
Figueira
J
.
Retinopathy phenotypes in type 2 diabetes with different risks for macular edema and proliferative retinopathy
.
J Clin Med
.
2020
;
9
(
5
):
1433
.
11.
Mendes
L
,
Marques
IP
,
Cunha-Vaz
J
.
Comparison of different metrics for the identification of vascular changes in diabetic retinopathy using OCTA
.
Front Neurosci
.
2021
;
15
:
755730
.
12.
Marques
IP
,
Ribeiro
ML
,
Santos
TP
,
Mendes
LG
,
Reste-Ferreira
D
,
Santos
AR
.
Different risk profiles for progression of nonproliferative diabetic retinopathy: a 2-year study
.
Ophthalmol Ther
.
2023
;
12
(
1
):
485
500
.
13.
Kim
AY
,
Chu
Z
,
Shahidzadeh
A
,
Wang
RK
,
Puliafito
CA
,
Kashani
AH
.
Quantifying microvascular density and morphology in diabetic retinopathy using spectral-domain optical coherence tomography angiography
.
Invest Ophthalmol Vis Sci
.
2016
57
9
OCT362
70
.
14.
Simonett
JM
,
Scarinci
F
,
Picconi
F
,
Giorno
P
,
De Geronimo
D
,
Di Renzo
A
.
Early microvascular retinal changes in optical coherence tomography angiography in patients with type 1 diabetes mellitus
.
Acta Ophthalmol
.
2017
;
95
(
8
):
e751
5
.
15.
Scarinci
F
,
Picconi
F
,
Giorno
P
,
Boccassini
B
,
De Geronimo
D
,
Varano
M
.
Deep capillary plexus impairment in patients with type 1 diabetes mellitus with no signs of diabetic retinopathy revealed using optical coherence tomography angiography
.
Acta Ophthalmol
.
2018
;
96
(
2
):
e264
5
.
16.
Tang
FY
,
Ng
DS
,
Lam
A
,
Luk
F
,
Wong
R
,
Chan
C
.
Determinants of quantitative optical coherence tomography angiography metrics in patients with diabetes
.
Sci Rep
.
2017
;
7
(
1
):
2575
10
.
17.
Tang
F
,
Cheung
CY
.
Quantitative retinal optical coherence tomography angiography in patients with diabetes without diabetic retinopathy
.
Invest Ophthalmol Vis Sci
.
2017
;
58
(
3
):
1766
.
18.
Chen
Q
,
Ma
Q
,
Wu
C
,
Tan
F
,
Chen
F
,
Wu
Q
.
Macular vascular fractal dimension in the deep capillary layer as an early indicator of microvascular loss for retinopathy in type 2 diabetic patients
.
Invest Ophthalmol Vis Sci
.
2017
;
58
(
9
):
3785
94
.
19.
Ludovico
J
,
Bernardes
R
,
Pires
I
,
Figueira
J
,
Lobo
C
,
Cunha-Vaz
J
.
Alterations of retinal capillary blood flow in preclinical retinopathy in subjects with type 2 diabetes
.
Graefe’s Arch Clin Exp Ophthalmol
.
2003
;
241
(
3
):
181
6
.
20.
Cogan
DG
,
Kuwabara
T
.
Capillary shunts in the pathogenesis of diabetic retinopathy
.
Diabetes
.
1963
;
12
:
293
300
.
21.
Cunha-Vaz
J
.
A central role for ischemia and OCTA metrics to follow DR progression
.
J Clin Med
.
2021
;
10
(
9
):
1821
.
22.
Gardner
TW
,
Davila
JR
.
The neurovascular unit and the pathophysiologic basis of diabetic retinopathy
.
Graefe’s Arch Clin Exp Ophthalmol
.
2017
;
255
(
1
):
1
6
.
23.
Kurtenbach
A
,
Wagner
U
,
Neu
A
,
Schiefer
U
,
Ranke
MB
,
Zrenner
E
.
Brightness matching and colour discrimination in young diabetics without retinopathy
.
Vision Res
.
1994
;
34
(
1
):
115
22
.
24.
Trick
GL
,
Burde
RM
,
Gordon
MO
,
Santiago
JV
,
Kilo
C
.
The relationship between hue discrimination and contrast sensitivity deficits in patients with diabetes mellitus
.
Ophthalmology
.
1988
;
95
(
5
):
693
8
.
25.
Cole
JB
,
Florez
JC
.
Genetics of diabetes mellitus and diabetes complications
.
Nat Rev Nephrol
.
2020
;
16
(
7
):
377
90
.
26.
Udler
MS
,
Kim
J
,
von Grotthuss
M
,
Bonàs-Guarch
S
,
Cole
JB
,
Chiou
J
.
Type 2 diabetes genetic loci informed by multi-trait associations point to disease mechanisms and subtypes: a soft clustering analysis
.
PLoS Med
.
2018
;
15
(
9
):
e1002654
23
.
27.
Ahlqvist
E
,
Storm
P
,
Käräjämäki
A
,
Martinell
M
,
Dorkhan
M
,
Carlsson
A
.
Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables
.
Lancet Diabetes Endocrinol
.
2018
;
6
(
5
):
361
9
.
28.
Silva
PS
,
Cavallerano
JD
,
Sun
JK
,
Noble
J
,
Aiello
LM
,
Aiello
LP
.
Nonmydriatic ultrawide field retinal imaging compared with dilated standard 7-field 35-mm photography and retinal specialist examination for evaluation of diabetic retinopathy
.
Am J Ophthalmol
.
2012
;
154
(
3
):
549
59.e2
.
29.
Kernt
M
,
Hadi
I
,
Pinter
F
,
Seidensticker
F
,
Hirneiss
C
,
Haritoglou
C
.
Assessment of diabetic retinopathy using nonmydriatic ultra-widefield scanning laser ophthalmoscopy (Optomap) compared with ETDRS 7-field stereo photography
.
Diabetes Care
.
2012
;
35
(
12
):
2459
63
.
30.
Price
LD
,
Au
S
,
Chong
NV
.
Optomap ultrawide field imaging identifies additional retinal abnormalities in patients with diabetic retinopathy
.
Clin Ophthalmol
.
2015
;
9
:
527
31
.
31.
Alder
VA
,
Su
EN
,
Yu
DY
,
Cringle
SJ
,
Yu
PK
.
Diabetic retinopathy: early functional changes
.
Clin Exp Pharmacol Physiol
.
1997
24
9–10
785
8
.