This study found that history of hypertension, LDL-C, PLR, and SII were independent risk factors for developing DR in patients with T2DM. History of hypertension, UA, HbA1c, TyG, SII, total CSVD load, WMHs, and LI were significantly different in mild and moderate-to-severe DR, among which UA, TyG, SII, total CSVD load, and WMHs were independent risk factors for the progression of DR.

DR is one of the most common chronic microvascular complications of T2DM, and its basic pathological change is retinal neovasculogenesis, which has become the main cause of irreversible visual impairment [17]. At present, DR has caused a heavy economic and social burden, affecting the quality of life of people aged 50 years and older [18]. The onset of DR is insidious. For patients with T2DM, it is not enough to only control known risk factors such as blood glucose, blood pressure, and lipid levels. Early identification of specific risk indicators and implementation of timely interventions can prevent blindness caused by DR [19].

Hypertension greatly accelerates the process of retinal microvascular disease by causing damage to microvessels, resulting in hemodynamic changes, which significantly increases the risk of DR in patients with T2DM [20, 21]. Its mechanism mainly involves increased vascular permeability, retinal edema, ischemic injury, neovascularization, oxidative stress, and inflammatory response [22]. Gong et al. found that systolic blood pressure was an independent risk factor for vision in patients with DR [23].

Previous studies showed that metabolic level was closely related to the progression of DR, in which LDL-C, TyG, and UA were important influencing factors [24]. High levels of LDL-C might accelerate the development of DR by promoting lipid deposition in retinal microvascular endothelial cells and triggering an inflammatory response [25]. The TyG index assessed FBG and triglyceride levels, revealing a synergistic effect of carbohydrate and lipid metabolism [26]. The increase of TyG index could predict the development of T2DM, the occurrence of diabetic nephropathy and coronary artery blockage, and the adverse trend of prognosis [27]. Other studies pointed out that the increase of TyG index was proportional to the severity of DR [28]. It was a risk indicator of DR and could predict the progress of DR [29]. Our study found a significant negative correlation between MoCA and TyG, suggesting that the higher the TyG, the more severe the cognitive impairment. UA had the dual effect of promoting oxidative stress and antioxidant activity, which could lead to inflammation, endothelial dysfunction, and vasoconstriction [30]. A meta-analysis found that UA levels were significantly higher in patients with DR than in diabetic controls, suggesting that UA might be a potential biomarker for identifying disease severity in patients with DR [31]. Under the influence of long-term high UA, the severity of DR would be further aggravated, which was consistent with the results of our study. Gong et al. found that LDL-C was also a risk factor for retinopathy [23]. Therefore, regular monitoring of LDL-C, TyG index, and UA levels is essential to prevent and slow the progression of DR. Effective management of these metabolic markers can help reduce the risk of retinal damage, improve visual health and quality of life in people with diabetes, and provide a strong basis for individualized treatment.

Inflammatory factors played a significant role in T2DM, which were key factors affecting DR [32]. NLR, PLR, LMR, and SII were novel markers that could reflect the balance and persistent inflammatory status between leukocyte subtypes [33]. These ratio indicators had higher diagnostic value than single inflammatory indicators, and were not susceptible to physiological changes in blood samples in the body. Therefore, they could provide more stable and accurate results [34]. PLR and SII have been shown to be significantly associated with the development of DR. PLR reflected the inflammatory state in vivo and the level of platelet activation. The inflammatory factors released by platelets could cause vascular endothelial cell injury and neovascularization, and promote the occurrence of DR in patients with T2DM [35]. Zeng et al. found that PLR was an independent risk factor for DR [36]. SII reflected the systemic inflammatory state, which also led to vascular endothelial cell damage and inflammatory cell infiltration. Thus, the process of retinal microangiopathy was intensified and the development of DR was accelerated [37]. This study found a significant negative correlation between MoCA and SII, suggesting that the higher the SII the more severe the cognitive impairment. There was a significant negative correlation between MoCA and PLR, suggesting that the higher the PLR the more serious the cognitive impairment. In addition, the study revealed that SII was an independent risk factor for the occurrence and development of DR, further validating the key role of inflammation in DR. It also contributed to our in-depth understanding of the pathogenesis of DR and provided new predictors and potential therapeutic targets for the clinic.

As a microvascular complication of T2DM, DR is closely related to systemic microvascular disease. Studies have shown that patients with moderate-to-severe DR often had extensive microvascular and small vessel lesions, which not only affected the retina but also caused irreversible damage to the micro/small blood vessels in the brain [38]. CSVD was the most common chronic progressive vascular disease affecting the white matter, arterioles, capillaries, and venules of the brain. The onset of CSVD was insidious and there was no specific clinical manifestation in the early stage. It was the leading cause of vascular cognitive impairment (VCI) [39]. On the basis of anatomical and physiological characteristics, retinal blood vessels were recognized to reflect changes in brain microstructure to a certain extent [40]. Recent studies have found a significant correlation between CSVD and retinal vascular changes. An investigation showed that the enlargement of the retinal venules and the narrowing of the arterioles were associated with the scale of white matter injury [41]. Another study found that changes in arteriovenous crossing, extensive focal arteriolar stenosis, and retinal vessel curvature were common in patients with CSVD [42]. Our study found a significant negative correlation between MoCA and total MRI load, suggesting that the greater the total MRI load the greater the cognitive impairment. We integrated WMHs, LI, EPVS, and CMBs into the assessment measures to fully account for the cumulative impact of CSVD lesions. The usefulness and accuracy of this approach has been demonstrated in several studies.

This study found that total CSVD load and WMHs were independent risk factors for DR progression. The reasons might be as follows: (1) the cerebrovascular lesions caused by the increase of total CSVD load would lead to systemic hemodynamic changes, including blood pressure fluctuations and reduced cerebral blood flow. These changes might lead to retinal blood perfusion insufficiency, aggravated retinal ischemia and hypoxia, and further promote the progression of DR. (2) White matter hypersignaling and lacunar cerebral infarction caused by increased CSVD load could damage the integrity of neurovascular units, thus affecting the functions of the blood–brain barrier and blood–retina barrier, increasing the penetration of harmful substances and aggravating retinopathy. (3) Brain dysfunction caused by CSVD might affect the central nervous system’s regulation of metabolism, exacerbate diabetes-related metabolic problems, and further aggravate DR. (4) Cognitive impairment caused by CSVD would affect the memory, attention, and executive function of patients, resulting in poor management of diabetes, which would further aggravate retinal microangiopathopathy and lead to the progression of DR. In addition, our study found that total CSVD load and LI were independent risk factors for cognitive impairment in moderate-to-severe DR.

We constructed a multi-factor integrated predictive model and found that smoking history, TyG, SII, total CSVD load, and LI were independent risk factors for cognitive impairment in patients with moderate-to-severe nonproliferative DR. DR was a common complication of diabetes, and patients with moderate-to-severe nonproliferative DR were at higher risk of developing cognitive impairment. Early prediction and intervention of cognitive impairment is essential to improve the quality of life and prognosis of patients with DR.

There are still some limitations in this study: (1) the sample size was relatively small, and the research objects were mainly concentrated in a specific region or population, which might affect the universality and generalization of the results. Future studies should expand the sample size to cover different regions and populations to verify the broad applicability of the findings. (2) Although a variety of risk factors were considered in the study, there might still be confounding factors that were not fully controlled, such as lifestyle and genetic factors, which might have an impact on the occurrence of DR and cognitive impairment. Therefore, future studies should consider the influence of these potential confounding factors more comprehensively. (3) This study was mainly based on cross-sectional data and could not fully reveal causality. Future studies should follow the long-term follow-up data of patients through longitudinal study design to better understand the dynamic process and causality of DR progression and cognitive impairment.