Introduction

A global public health concern today is type 2 diabetes mellitus (T2DM), which has increased in prevalence due to the world’s rapid economic and social development. According to the International Diabetes Federation (IDF), the number of people with diabetes mellitus will grow to 552 million by 20301. Diabetes mellitus causes a variety of eye complications, including diabetic retinopathy (DR), cataracts, dry eye, corneal hyperalgesia, and diabetic iridopathy2. Nowadays, a lot of clinical studies uses retinal vasculature to assess the effects of diabetes on the eye3,4. Moreover, retinal vasculature is also used to analyze the effect of hypoxia in many other systemic diseases such as COVID-19 and sleep apnoea5,6,7. The retinal vasculature cannot be seen in diabetic patients with certain types of optic media clouding, such as cataracts and vitreous clouding. This makes it difficult to assess the ocular vasculature in these patients and to perform long-term follow-up. Therefore, better methods for visualizing ocular vascular alterations in diabetes individuals are required.

The pathophysiological alterations of ischemia and hypoxia in diabetes mellitus affect the entire vascular system of the eye8,9. Previous studies have shown fluorescein leakage at the pupil margin by anterior segment fluorescein angiography in diabetic patients without DR, suggesting that abnormalities of the iris vascular system precede the onset of abnormalities of the retinal vascular system in diabetic patients10,11,12. Diabetic iridopathy eventually develops into neovascular glaucoma, a diabetic ocular complication that presents a major risk to vision and even blindness13,14. Thus, observing the iris vasculature can provide full information about ocular vascular changes in diabetes patients, which is crucial for early prevention, screening, and follow-up evaluation of ocular vascular diseases. However, previous studies based on anterior segment fluorescein angiography of the iris vasculature in diabetic patients could not visualize and quantify the microstructure of the iris vasculature and were unable to better assess the micro alterations of the iris vascular system in diabetic patients. Anterior segment optical coherence tomography angiography (AS-OCTA) provides a non-invasive, rapid, and quantitative view of the iris vasculature. It may provide thorough iris vasculature imaging with fast scanning rates that enable 360° imaging of the iris, which gives extensive information about the iris vasculature15. Therefore, AS-OCTA appears to be a potential novel approach for studying the iris vascular system in diabetes patients.

Therefore, this study aimed to evaluate early changes of iris vasculature in T2DM patients without DR, as well as potential predictors contributing to these alterations.

Methods

Study participants

This was a cross-sectional observational study. This study comprised patients who visited the Department of Ophthalmology at the Affiliated Hospital of Yangzhou University between February 2024 and July 2024. Approval for data collection and analysis was obtained from the ethics committee of the Affiliated Hospital of Yangzhou University and adhered to the tenets set forth in the Declaration of Helsinki. All participants signed informed consent before the examination. We included adults with T2DM and healthy controls without underlying systemic diseases who had good fixation and cooperation in examinations. Participants were excluded from the study if they had systemic diseases other than T2DM; had DR changes in the fundus; had other ischemic retinal diseases in the fundus; inability to fixate or ptosis on examination; presence of significant iris nodules, atrophy or neovascularization; had any history of ocular trauma, surgery or laser treatment; and had low quality of fundus photographs and AS-OCTA images.

Clinical examinations

Systemic and ophthalmic examinations were performed on each participant. Demographic data collected were age, sex, height, weight, systolic/diastolic blood pressure (SBP/DBP), duration of diabetes and fasting blood glucose (FBG). The formula for calculating body mass index (BMI) was weight in kilograms divided by height in square meters. All participants had extensive ophthalmologic examinations. Slit-lamp biomicroscopy(SL-1800, Nidek, Tokyo, Japan) examined the eye’s anterior segment, which includes the conjunctiva, cornea, sclera, anterior chamber, iris, pupil, and lens. Intraocular pressure (IOP) was measured using a noncontact air-puff tonometer (TX-F, Canon, Tokyo, Japan). Fundus photographs were performed using an ultra-wide angle fundus camera (Mirante SLO, Nidek, Tokyo, Japan), which uses laser technology to scan the fundus of the eye and displays 3.2 times the typical extent of the retina as compared to a traditional fundus camera with a 7-field of view. Two expert ophthalmologists assessed fundus photographs to rule out DR and other retinal ischemic diseases, such as retinal vein occlusion. In the event of a disagreement, an assessment by a third qualified ophthalmologist was requested.

AS-OCTA iris blood flow quantification

Both eyes of all the enrolled subjects used swept-source anterior segment optical coherence tomography (SS-ASOCT) (YG-100 KPRO, TowardPi Medical Technology, Beijing, China) to obtain the iris AS-OCTA images in the same indoor examination room with north-facing windows, avoiding direct sunlight exposure. Fixed ceiling Light Emitting Diode (LED) panels (4000 K color temperature, 80 color rendering index) operated at maximum brightness were used as the sole artificial light source. All measurements were taken from 9:00 a.m. to 11:00 a.m. The device uses a swept laser source with a wavelength centered at 1,060 nm and a scan rate of 100,000 A-scans per second. With a bandwidth of 100 nm, the optical resolution is 3.8 μm axially and 10 μm transversely. AS-OCTA obtained a 12 mm × 12 mm scan mode (Fig. 1A). The cross-sectional image of iris was shown (Fig. 1B). TowardPi built-in signal strength system was used to obtain image quality. This study included images with quality scores of 7 or higher. Given that the binocular iris exhibits uniformity, we selected monocular eyes with better imaging quality for inclusion in further analysis. The whole iris blood flow included the full circle area from the pupillary margin to the corneoscleral limbus (Fig. 1C). SS-ASOCT built-in software was used to quantify the entire iris vessel density (IVD) and pupil diameter (PD) (Fig. 1D). The pupillary margin was manually adjusted if the software failed to detect it correctly. The iris vascular density was also performed in quadrant sectors, that is nasal superior (NS), nasal inferior (NI), temporal superior (TS), and temporal inferior (TI) sectors (Fig. 1D).

Fig. 1
figure 1

Representative iris AS-OCTA image and corresponding pseudo-color map of a right eye. The scan mode of AS-OCTA was 12 mm × 12 mm (A). Iris in a SS-ASOCT cross-sectional image (B). Iris AS-OCTA image (C) and corresponding pseudo-color map (D) of a right eye. The built-in software automatically detected the boundaries of the iris and pupil (denoted in white). The iris was divided into four quadrant sectors, that is nasal superior (NS), nasal inferior (NI), temporal superior (TS), and temporal inferior (TI) sectors. PD, pupil diameter.

Full size image

Statistical analysis

SPSS software (version 22.0; SPSS, Inc., Chicago, IL, USA) was used for analyses.

The Kolmogorov-Smirnov test was used to test that all measurement data were normal.

The t-test or Wilcoxon rank-sum test was performed based on the results to evaluate IVD between the two groups. Pearson correlation analysis or Spearman correlation analysis was used to assess the relationship between the iris blood flow parameters and ocular or systemic factors. Parameters with significant correlations were further included in the regression analysis. The means ± standard deviations are presented. A value of P < 0.05 was considered statistically significant.

Results

Thirty-six patients with T2DM and thirty-nine healthy controls were recruited in our study. All T2DM patients did not have DR in the fundi. One eye of each subject was chosen based on the quality threshold of the AS-OCTA image. If both eyes had the same quality, one was chosen at random. Their characteristics are shown in Table 1. Except for fasting blood glucose, there was no difference in age, sex, BMI, blood pressure, IOP, PD and quality index (QI) between T2DM patients and healthy controls. Fasting blood glucose in T2DM patients was 9.42 ± 4.58 mmol/L, while healthy controls had 5.97 ± 0.74 mmol/L (P < 0.001).

Table 1 Demographic characteristics of all participants.
Full size table

The iris vessel density of all participants is shown in Table 2; Fig. 2. IVD in T2DM groups was lower than in healthy groups, particularly in the whole iris, nasal inferior, and temporal inferior sectors of the iris. In the whole iris, IVD was significantly lower in diabetes (37.25 ± 9.22) compared with healthy groups (42.56 ± 7.66) (P = 0.016; Table 2). In the nasal inferior sector, IVD was significantly lower of diabetes (37.44 ± 11.74) compared with healthy groups (43.51 ± 9.05) (P = 0.022; Table 2). While in the temporal inferior sector, IVD was significantly lower in diabetes (38.69 ± 9.48) compared with healthy groups (45.46 ± 8.87) (P = 0.007; Table 2).

Table 2 Iris vessel density in all participants.
Full size table
Fig. 2
figure 2

Iris vessel density of all participants. * P < 0.05. HC, healthy controls; T2DM, type 2 diabetes mellitus; IVD, iris vessel density; NS, nasal superior sector; NI, nasal inferior sector; TS, temporal superior sector; TI, temporal inferior sector.

Full size image

The relationships between iris vessel density and systemic and ocular factors in healthy participants are shown in Table 3. In healthy populations, sex was significantly positively associated with whole IVD (r = 0.487, P = 0.002; Table 3), NI IVD (r = 0.402, P = 0.011; Table 3), TS IVD (r = 0.375, P = 0.019; Table 3) and TI IVD (r = 0.465, P = 0.003; Table 3). Body mass index was significantly negatively related to TI IVD (r = −0.395, P = 0.013; Table 3). While fasting blood glucose was significantly negatively related to TI IVD (r = −0.366, P = 0.019; Table 3). Age, blood pressure, IOP and PD were not associated with IVD (P > 0.05; Table 3).

Table 3 Correlation analysis of iris vessel density in healthy participants.
Full size table

The relationships between iris vessel density and systemic and ocular factors in T2DM patients are shown in Table 4. In T2DM patients, age was significantly positively associated with NS IVD (r = 0.522, P = 0.038; Table 4). Sex was significantly positively associated with NS IVD (r = 0.493, P = 0.023; Table 4) and TS IVD (r = 0.493, P = 0.042; Table 4). Body mass index was significantly negatively related to whole IVD (r = −0.334, P = 0.027; Table 4) and TI IVD (r = −0.362, P = 0.022; Table 4). While fasting blood glucose was significantly negatively related to whole IVD (r = −0.339, P = 0.036; Table 4) and NI IVD (r = −0.466, P = 0.044; Table 4). Blood pressure, IOP, duration of diabetes and PD were not associated with IVD (P > 0.05; Table 4).

Table 4 Correlation analysis of iris vessel density in T2DM patients.
Full size table

We analyzed the differences in IVD between males and females. Males had higher IVD than females in whole (43.71 ± 7.07 VS 36.30 ± 7.07, P < 0.001), NS (40.91 ± 3.89 VS 33.60 ± 10.72, P = 0.003), NI (44.29 ± 9.19 VS 37.30 ± 8.48, P = 0.007), TS (43.26 ± 6.36 VS 33.65 ± 7.78, P = 0.002), TI (46.57 ± 4.95 VS 38.10 ± 5.66, P < 0.001).

Multiple regression analysis of iris vessel density are shown in Table 5. The iris vascular density was chosen as the dependent variable, with age, sex, BMI, and FBG as the independent variables. Male gender was a positive independent predictor of whole IVD (β = 0.408, P = 0.001; Table 5) and TI IVD (β = 0.248, P = 0.032; Table 5). BMI was a negative independent predictor of NI IVD (β = −0.270, P = 0.038; Table 5). While FBG was a negative independent predictor of whole IVD (β = −0.305, P = 0.024; Table 5) and NI IVD (β = −0.271, P = 0.041; Table 5).

Table 5 Regression analysis of iris vessel density in all participants.
Full size table

Discussion

Previous studies demonstrated iris fluorescein leakage by fluorescein angiography in diabetic patients10,11,12. However, research on quantifying iris vascularization in diabetic patients is poor. To the best of our knowledge, this study is the first to observe the whole iris vascular system in T2DM patients without DR using AS-OCTA. We found that iris vascular density was lower in T2DM patients without DR compared to those without T2DM, especially in the whole, NI and TI sectors of iris. Moreover, male gender, BMI, and blood glucose were associated with iris vascular density.

Previous studies on diabetic patients’ ocular vasculature have always centered on the retina. Numerous studies have shown impairment to the retinal microcirculation and microstructure in diabetes patients4,16,17. However, the inability to observe the retina has been an issue in some diabetic patients with cataracts and vitreous clouding, making follow-up and screening for ocular complications in this population more difficult. The development of AS-OCTA has provided a novel method for quantification of iris blood flow. It was even shown that in ocular ischemia diseases, such as DR and retinal vein occlusion, iris vascular parameters are more sensitive than retinal, and it may be employed as a new predictor of ocular ischemic diseases18.

In this study, iris vascular density was reduced in T2DM patients without DR, which is consistent with the study by Cui et al19. However, the scan range of AS-OCTA in our research was 12 mm x 12 mm, and the whole iris was obtained for quantitative blood flow analysis. To our knowledge, this is the first full-range assessment of iris blood flow parameters. Our study found that a significant decline in iris vascular density not only in the whole iris, but also in the NI and TI sectors of iris. However, Cui et al. found that T2DM patients exhibited higher iris vessel density at the pupil margin20. This may probably be due to sectoral filling patterns revealed by iris fluorescein angiography21. Hayreh and Scott reported that such sectoral filling is physiological and is most probably because the various supply arteries to the iris arise at different points along the ophthalmic artery22. Moreover, Cui et al. reported that T2DM patients had smaller pupil diameters than healthy adults20. In our study, we found that although the pupil diameter in the diabetic group was smaller than that of the healthy group, there was no significant difference between two groups, which could be attributed to the varied methodologies employed to quantify pupil diameter.

Furthermore, in the present study, we found that male gender was a positive independent predictor of iris vascular density, whereas BMI and blood glucose were negative independent predictors of iris vascular density. According to a previous study, men have thicker iris than women23. While greater iris thickness may indicate denser blood flow, iris vascular density may be higher in males compared to females, nevertheless further investigations are needed to prove this. Higher BMI may have a negative predictor on iris vascular density. Visceral obesity is closely linked to insulin resistance, and earlier studies suggested that free fatty acids may be the cause of this relationship24. Obesity and metabolic syndrome are important contributors to vascular endothelial dysfunction, and both can cause insulin resistance and disturb vascular permeability25. This may lead to disruption of the integrity of the iris vascular system (i.e., blood-aqueous barrier). Hyperglycemia is a major risk factor in the development of endothelial dysfunction in diabetic patients. Advanced glycation end products (AGEs) formation is an important biochemical abnormality in diabetes mellitus. The interaction of AGEs with vascular wall components increases vascular permeability, expression of procoagulant activity, and reactive oxygen species production, resulting in endothelial leukocyte adhesion molecules with enhanced expression26,27,28,29. At the same time, the interaction between glucose, insulin, and increased triglycerides has a synergistic effect on atherosclerosis and plays a key role in the early pathophysiology of vascular lesions in diabetic patients30.

There are several limitations in the current study. First, we exclusively analyzed T2DM patients without DR, hoping to investigate early iris blood flow changes in T2DM patients, and subsequent inclusion of patients with DR of varying severity may be required in the future. Second, our study’s sample size was limited, therefore, larger sample sizes and more centers may be required in future research to validate the value of iris blood flow parameters. Third, we did not actively control ambient lighting in this study, but we did our best to reduce its potential impact using a variety of methods. In future studies, we will improve the ambient lighting control.

In conclusion, AS-OCTA provides a reliable method to quantify iris blood flow. Our study reveals the deterioration of the iris vascular system in patients with T2DM, even without DR. Male gender was a positive independent predictor of IVD, whereas body mass index and blood glucose were negative independent predictors of IVD. These findings highlight the significance of iris blood flow parameters in T2DM patients, which could be helpful in follow-up studies and future screening for ocular vascular diseases associated with T2DM.