Choroidal Thickness in Non-proliferative Diabetic Retinopathy vs Healthy Eyes using SS-OCT

2.1. Patient Enrollment

Patients diagnosed with T2DM with NPDR were recruited from a university hospital in Chengdu, China, from June 2022 to December 2023. The study adhered to the principles of the Declaration of Helsinki, and written consent for participation was obtained from the patients or their immediate family members. The study was approved by the Ethics Committee of the Affiliated Chengdu 363 Hospital of Southwest Medical University.

2.2. Inclusion and Exclusion Criteria

The inclusion criteria were: patients aged 45–75 years, diagnosed with NPDR by two professional ophthalmologists (work more than 10 years in retinal disease). The exclusion criteria were: having undergone ocular treatments, such as ocular surgery or retinal laser therapy, and those who had received ocular injections of medications, including anti-vascular endothelial growth factor (VEGF) or OZURDEX. Patients with systemic conditions affecting choroidal circulation (hypertension history: blood pressure >140/90 mmHg, diabetic nephropathy, or cardiovascular disease) were excluded. Meanwhile, patients with eye diseases (such as uveitis, glaucoma, corneal diseases, or retinal diseases) and those with media opacity affecting retinal image clarity were also excluded, as were individuals with refractive errors exceeding ± 6 diopters. The experimental group (the NPDR group) consisted of patients meeting these criteria. A control group was formed with healthy individuals without other ocular or systemic diseases.

2.3. Ophthalmic Examination

All patients underwent a comprehensive ophthalmic examination, including slit-lamp and fundus examinations, fundus photography, best-corrected VA assessment, and measurement of intraocular pressure (IOP). Additionally, the patients' medical history and duration of diabetes were recorded. Swept-source OCT (SS-OCT) (VG200S, SVision Imaging; Henan, China) was utilized to examine all included eyes. Compared to conventional SD-OCT, SS-OCT offers three key advantages for choroidal imaging: enhanced penetration, high-speed scanning and automated segmentation. These technical advancements ensure precise and objective CT measurements, particularly in diabetic eyes with potential edema or structural disruption. It also integrates eye tracking to generate thickness maps of the retina and choroid while maintaining central fixation. The macular region (6×6 mm) was evaluated.

2.4. Measurement of CT

CT was defined as the distance between the choro-scleral interface and the outer border of the retinal pigment epithelium (RPE). An automated built-in calibration software was employed to calculate CT for each sector. The examination grid was divided into nine sectors: the central foveal region (0–1 mm) and the inner ring extended 1–3 mm from the foveal center. The outer ring extended 3–6 mm from the foveal center. The inner and outer rings were further divided into four quadrants (superior, inferior, nasal, and temporal), as depicted in Fig. (1). All SS-OCT assessments were conducted by experienced technicians, and scans of poor image quality were excluded.

2.5. Statistical Analysis

Data normality was assessed using the Kolmogorov–Smirnov test and Q-Q plots. The independent t-test was employed to evaluate statistically significant differences in CT between the experimental and control groups. The statistical analysis was conducted using IBM SPSS Statistics 21.0; effect sizes and confidence intervals (CIs) were calculated with the threshold for statistical significance set at P < 0.05.

3.1. Patient Characteristics

This study included 52 patients with NPDR, ranging in age from 45 to 75 years, amounting to a total of 104 eyes for analysis. For the control group, 62 eyes from 31 healthy volunteers were selected. The NPDR group had 16 female and 36 male patients, while the control group had 16 female and 15 male patients. The average age of patients was 56.00 ± 9.08 group and 38.94 ± 10.92 years in the NPDR and control groups, respectively. The average duration of diabetes was 4.67 ± 1.41 years in the NPDR group. In both groups, the IOP values fell within the normal range. Table 1 presents a comprehensive overview of the demographic characteristics of both groups.

3.2. CT in the Central Foveal Region (0-1 mm)

In the central foveal region (0-1 mm), the mean CT was 343.21 ± 107.84 µm and 282.56 ± 92.81 µm in the NPDR and control groups, respectively. The mean CT was significantly higher in the NPDR group than in the control group (P < 0.001) (Table 2 and Fig. 2).

3.3. CT in the Inner Ring Region (1-3 mm)

Within the inner ring spanning 1-3 mm from the foveal center, the mean CT in the superior quadrant was higher in the NPDR group (333.71 ± 110.01 µm) than that in the control group (294.98 ± 97.47 µm) (P = 0.023). Similarly, in the inferior quadrant, the mean CT was higher in the NPDR group (334.71 ± 107.92 µm) than in the control group (284.48 ± 97.40 µm) (P = 0.003). The mean CT was also higher in both the nasal (326.00 ± 124.51 µm) and temporal quadrants (339.75 ± 96.61 µm) in the NPDR group than in the control group (nasal: 261.16 ± 96.75 µm, temporal: 292.97 ± 92.12 µm) (P = 0.001, P = 0.003) (Table 2 and Fig. 2).

T2D, type two diabetes, inner ring region: 1–3 mm from the foveal center, outer ring region: 3–6 mm from the foveal center.

3.4. CT in the Outer Ring Region (3-6 mm)

The mean CT in the inferior quadrant (313.03 ± 94.87 µm) and nasal quadrant (282.10 ± 113.68 µm) were higher in the NPDR group than in the control group (281.31 ± 93.19 µm, 219.85 ± 88.94 µm) (P = 0.037, P < 0.001). However, while the NPDR group exhibited slightly higher mean CT in the superior and temporal quadrants, no significant differences were observed between the NPDR and control groups (Table 2 and Fig. 2).

3.5. CT in the all Regions with Age Adjustment

To account for the age disparity, an ANCOVA was performed with age as a covariate. After adjustment, the NPDR group maintained significantly thicker choroid in the central foveal region (P < 0.001), inner ring quadrants (P< 0.001), and outer ring quadrants (Superior: P <0.001, Inferior: P <0.001, Nasal: P <0.001; Temporal: P=0.003), confirming that CT differences are independent of age (Table 3).

4.1. Imaging Technology Heterogeneity

Studies using SD-OCT Querques et al. [16] reported generalized CT thinning, potentially due to limited choroidal visualization in diabetic eyes with edema or fibrosis. In contrast, SS-OCT’s deeper penetration (1,050 nm vs . 840 nm wavelength) and enhanced signal-to-noise ratio Copete et al. [13] enable precise CT measurement even in pathologic choroids, as demonstrated by our findings and Wang et al. [24]. Manual choroidal boundary delineation Unsal et al. [20] introduces inter-observer variability (±15–20 µm), whereas our automated software reduced the risk of human error and bias associated with manual measurements.

4.2. Participant Stratification

Studies with mixed DR stages Adhi et al. [19]: NPDR + PDR observed attenuated CT differences, whereas our strict focus on NPDR revealed pronounced thickening. Similarly, cohorts with longer diabetes duration Kim et al. [25]: 8.2 years showed thinning, suggesting CT dynamics shift from compensatory thickening to atrophy as DR progresses.

In summary, the apparent contradictions in CT literature arise from technological and population-level differences. Our use of SS-OCT, automated segmentation, and strict NPDR enrollment criteria provides a refined perspective on early choroidal changes, underscoring the need for standardized protocols in future studies.

The choroid, responsible for 95% of ocular blood flow, is under autonomic nervous system control and exhibits autoregulation patterns. In the initial stages of DR, sympathetic innervation activation leads to increased choroidal circulation, resulting in choroidal thickening. Improved repeatability of choroidal measurements can enhance our understanding of DR detection, progression, and response to treatment, as the choroid may play a crucial role. The thickening of the choroid in the stages of NPDR may be linked to choroidopathy secondary to T2DM, indicating a possible development of choroidopathy in diabetic patients [27]. Diabetic choroidopathy may contribute to RPE dysfunction and influence vascular permeability. The thickening of the choroidal layer might be attributed to overexpression of cytokines activated by inflammation, oxidative stress, and angiogenesis during the initial stages of DR. Studies have reported significant associations between choroidal thickening and cytokines such as monocyte chemotactic protein-1, platelet-derived growth factor, and VEGF [28]. Some authors suggest that microvasculopathy in the retina, choroid, and kidneys may share pathogenic mechanisms related to metabolic changes, inflammation, and endothelial dysfunction characteristic of diabetes [29]. Therefore, during the early stages of DR, preserving perfusion may exert a protective influence on photoreceptors and the RPE [22].

Our study has two key strengths that distinguish it from others in the field. First, we carefully selected participants at the early stage of NPDR, minimizing potential confounding factors that could skew our results. This approach strengthens the validity of our findings and enhances their relevance to the broader understanding of NPDR. Second, we utilized SS-OCT, a non-invasive imaging technology that provides high-resolution cross-sectional images of the retina and choroid for CT measurements. SS-OCT’s deeper penetration and automated segmentation enabled the detection of subtle CT changes in NPDR. This sensitivity is critical for early diagnosis, as choroidopathy may precede visible retinal lesions, reducing the risk of human error and bias associated with manual measurements.

Nevertheless, this study also has some limitations. First, we did not employ a longitudinal approach, which would be important for observing changes in choroidal blood flow in DR patients as the disease progresses from NPDR to PDR. Second, although the age distribution between the two groups was not perfectly matched, we regard age as a potential confounder and validate our conclusion through adjusted analyses. The result suggested that this age difference did not significantly impact our result. Future studies still should be designed to include two groups of age-matched patients to reduce the risk of such bias.