The relative abundance ratio of Bacteroidetes to Firmicutes (B/F ratio) was associated with DR, the B/F ratio was increased in GM of DR patients (Khan et al., 2021). Based on this study, the increased B/F ratio of the GM in T2DM patients may be a potential biomarker for sight-threatening DR. DR patients showed ecological imbalance from the phylum to the genus level (Bai et al., 2022). At the phylum level, Firmicutes decreased, Bacteroidota and Desulfbacterota enriched in DR patients. At the genus level, Bacteroidetes are the most abundant. This is consistent with previous studies. A Mendelian random study showed that Chritensenellaceae and Peptococcaceae were protective factors for DR at the family level, and Ruminococcaceae and Eubacterium were associated with the risk of DR at the genus level (Liu et al., 2022b). We note that these floras belong to Firmicutes in this study. Consequently, this result potentially suggests that we should not only focus on major categories but also changes in specific floras when studying GM flora. Another study found that the Pasteurellaceae flora can be used as a potential marker to distinguish DR and DM (high accuracy AUC: 0.74) (Huang et al., 2021). Furthermore, whether this flora can be used clinically to monitor the development of DM/DR patients remains to be explored, but we expect that the potential biomarkers of these florae can be used as a powerful point for us to monitor the disease in the future.

There are significant differences in GM between DR patients and normal people (Das et al., 2021). Through 16SrRNA sequencing analysis, the gut microecology of DR patients has been disordered. In addition, the anti-inflammatory bacteria (Roseburia, Anaerostipes, Coprococcus, Streptococcus, Phascolarctobacterium, Faeculibacterium, Ruminococcus) and probiotics (Bifidobacterium, Lactobacillus) were found to be reduced. There was a decrease in the pro-inflammatory bacteria Sutterella, but an increase was observed in Shigella. The pathogenic bacteria such as Haemophilus and Rothia were reduced, but Escherichia and Enterobacter were greatly increased. The GM flora in the organism can regulate the inflammatory response by changing the levels of pro-inflammatory factors, anti-inflammatory factors, chemokines, and proteins. For example, the Faecalibacterium flora inhibits the activity of IL-6, IL-17, and HsCRP (van den Munckhof et al., 2018). Lactobacillus and Bifidobacterium are classified as beneficial bacteria, which can play a variety of health-promoting effects on the organism (Hill et al., 2014). They can reduce inflammation and islet β cell dysfunction by reducing the expression of cytokines IL-6, IL-1β, IL-8, and circulating LPS levels (Tian et al., 2016). Escherichia and Enterobacter produce misfolded amyloid proteins that affect the development of DM (Tetz et al., 2019). In summary, with the increase and decrease of the flora in these categories of anti-inflammatory, pro-inflammatory, and pathogenic, the gut microecological balance is disturbed, which promotes the disruption of the GM flora and induces an inflammatory response. This result is slightly different from the theoretical assumption of DR chronic inflammation, which further leads to the consideration of whether the internal imbalance of anti-inflammatory and pro-inflammatory flora is the cause of DR chronic inflammation.

The structure of GM flora is not only at the level of each phylum but also includes its diversity, similarity, and abundance. α-diversity refers to the GM biodiversity within the organism, including the richness index (Ace, Chao, OTUs), the diversity index (Shannon, Simpson, Invsimpson), and the evenness index (Heip, Shannoneven). β-diversity refers to the diversity differences among individuals, which are generally detected by UniFra, PERMANOVA, Anosim analysis and similarity analysis by UPGMA clustering after principal coordinate analysis (PCoA). DR patients with gut microecology disorders will show more complex pathological diversity (Huang et al., 2021). DR patients exhibited lower α-diversity and β-diversity in comparison to healthy individuals. Additionally, the probiotics Bifidobacterium and Lactobacillus showed an increase in abundance, as well as a decrease in the relative abundance of Firmicutes, and enrichment of Bacteroidetes, Desulfbacterota. This result is quite different from the above studies. The reason may be that some DR patients are affected by metformin or other drugs during the course of the disease. As studies have found that metformin and AMC (herbal formula) can significantly change the GM flora structure of T2DM patients (de la Cuesta-Zuluaga et al., 2017; Tong et al., 2018). In addition, compared with DM patients without DR, the α-diversity of PDR patients decreased (manifested as decreased OTUs, Chao 1, Shannon, Simpson), and the β-diversity based on PCoA showed a significant separation of the GM flora between the two groups (Ye et al., 2021). A study carried out in India demonstrated an increase in α-diversity in patients with DR (Das et al., 2023). This also further reminds us that in the study of the association between GM flora and DR/DM or other metabolic diseases, apart from looking at the effects of drugs, we should also pay attention to different species of populations (Figure 1; Table 1).

Several limiting factors must be taken into account when considering the aforementioned changes in the structure of the GM flora in DR. Firstly, the imbalance in the GM flora will be affected by some confounding factors such as race, age, gender, sample size, dietary habits, lifestyle, drugs, and fecal sampling methods. Secondly, on the one hand, most of the subjects were admitted to the hospital. There was some variation in the admission rate and selective bias. On the other hand, based on these cross-sectional studies, the causal relationship between the differences in the GM flora could not be precisely revealed. Finally, it is imperative to obtain additional evidence to confirm the restoration of the disordered gut microecology in disease treatment. Despite the various limitations of these studies, we also have a broad and preliminary understanding of gut microecology in DR. In the future, researchers should take these limitations fully into account in their research. The potential for bias can be reduced by using multi-center and large-sampled cohort studies. In addition, scientists should not limit themselves to clinical research, which should be supplemented by animal/cell experiments to improve the relevant mechanism of action, and further elucidate the causal relationship between GM flora and DR.

GM flora produces metabolites such as lipopolysaccharide (LPS), trimethylamine oxide (TMAO), BAs, and SCFAs. These metabolites are absorbed by the colon and released into the blood through the portal vein, where they are further released into the peripheral organs, such as the eyes neovascularization (Koh et al., 2016; Schroeder and Bäckhed, 2016). This gut-blood-retinal axis circulatory process is also accompanied by the regulation of cytokines, proteins and other modifications, such as promoting or inhibiting the production of inflammatory factors, activating immune regulation, leading to chronic persistent low-level inflammatory response, thus participating in the development and progression of DR (Figure 2).

Protective effects on DR from consumption of fruit, vegetables, fiber, fish, tea, oleic acid, and Mediterranean diet (Shah et al., 2022). For example, a high-fiber diet may enhance the serum metabolic level of T2DM patients by altering the GM flora (Chen et al., 2023). Docosahexaenoic acid (DHA) exists in most marine foods, such as fish, shrimp, and shellfish. It has a regulatory effect on the GM flora and immune function, such as the inhibition of neovascularization and ischaemic injury (Chiang et al., 2022). DHA-acidified Cur ester treatment can prevent the TMAO-mediated PT3K-NF-κB signaling pathway, thereby inhibiting inflammation, apoptosis, and oxidative stress (Shi et al., 2022). Therefore, a targeted diet may become a potential dietary pattern to prevent and treat the occurrence and development of related diseases, which still needs more research support.

At present, metformin, acarbose, and glipizide are the preferred drugs for T2DM. Regular medication can not only reduce the patient’s blood glucose but also play a protective role in microvascular complications to delay the development of DM. In recent years, studies have gradually found that these drugs can also improve metabolic function through the structure of GM flora and its metabolites. After acarbose treatment, Bifidobacterium, Lactobacillus, and other beneficial bacteria were enriched in T2DM patients, and the abundance of Bacteroides, Clostridium, and other bacteria decreased (Takewaki et al., 2021). It is worth noting that the study also found that these changes in flora may be closely related to habitual dietary intake. Furthermore, in addition to the above effects of GM flora, researchers have found that acarbose can improve insulin resistance by increasing the ratio of primary BAs/secondary BAs in plasma (Gu et al., 2017), and can also reduce the expression of cytokines such as LPS, prothrombin activator inhibitor-1 (PAI-1), monocyte chemoattractant protein-1 (MCP-1) in serum (Su et al., 2015). The effect of metformin on GM flora is not repeated here (see 3.1). In conclusion, these positive effects also suggest the importance of avoiding confounding factors during the clinical study of gut microecology in patients with DM/DR. In the future, the structure and metabolism of GM flora may provide a perspective when we choose drug strategies, as well as the intervention of dietary guidance may play a complementary role.

As outlined in Section 3.1, besides confounding factors like medication, most studies indicate reduced levels of probiotics, such as Bifidobacterium and Lactobacillus, in patients with DR. The intervention of probiotics and prebiotics is viewed as an effective supplement, which represents a new strategy with significant potential in the development of DM/DR. In a double-blind randomized controlled trial, patients with T1DM were treated with insulin and capsules containing three probiotic strains (Wang et al., 2022). The treatment increased the Bifidobacterium, Lactobacillus salivarius, and Akkermansia muciniphila flora in the body, which helped regulate immune cytokine levels, resulting in a more stable blood glucose level. This study focuses on the combined effect of multi-strain probiotic supplementation. However, individual effects of single probiotic strains, as shown in subsequent studies, need further consideration. T2DM rats were supplemented with the probiotic Lactobacillus casei Q14, which was derived from the fermented dairy product yak yogurt (Qu et al., 2018). The study found a significant change in the composition of GM flora, as well as selective enrichment of SCFAs-producing flora, leading to increased secretion of SCFAs and GLP-1 for the regulation of DM development. In contrast, in clinical trials of individuals with metabolic syndrome, who were administered daily oral doses of live or pasteurized Akkermansia muciniphila flora, it was observed that the overall structure of the GM remained unchanged (Depommier et al., 2019). Nevertheless, this study also found that the probiotics could improve gut barrier function, accompanied by a decrease in plasma LPS levels, further improving the host’s inflammatory response and insulinemia. Similarly, prebiotic supplementation in obese and DM mice resulted in improved inflammatory tension and gut barrier function. These effects were found to be altered by GLP-2-dependent mechanisms (Cani et al., 2009). It was also accompanied by increased abundance of the Bifidobacterium spp., Lactobacillus spp., and Eubacterium rectale clusters. Furthermore, research has demonstrated that probiotics could serve as a drug delivery vehicle for the management of DR in the future (Prasad et al., 2023). Although some positive results have been documented in these studies, issues related to the availability, formulation composition, and quality control of probiotic strains remain challenging. Due to the complex nature of GM flora and individual variations, we may also explore the possibility of personalized probiotic supplementation in the future.

Could GM flora metabolites play an interventional role alongside the supplementation of effective strains? Based on the above 3.2 parts, we have learned that the GM flora and its metabolites were disturbed in the DM/DR. A logistic regression analysis suggests that BAs might function protectively, inhibiting the progression of T2DM into DR (Su et al., 2021). Consequently, BAs present a potential therapy target for retinal diseases. TUDCA is a metabolite of BAs with anti-inflammatory effects. Following treatment with TUDCA, DM mice were found to alleviate retinal inflammation, vascular permeability, and endoplasmic reticulum stress via the TGR5 receptor signaling pathway, ultimately resulting in improved visual function (Lenin et al., 2023). Ursodeoxycholic acid (UDCA) is a hydrolyzed form of TUDCA. Furthermore, the treatment of DR mice with UDCA leads to improved retinal vascular integrity and a reduction in pericyte loss (Chung et al., 2017).

Several previous studies have shown a decreasing trend in the levels of SCFAs and the abundance of the GM flora responsible for producing SCFAs during DR. Exogenous supplementation of SCFAs can improve the metabolism of amino acids, vitamins, and energy. Besides increasing the strains that can produce SCFAs, it is possible that supplementing with derivative SCFAs may also impact DR. On the one hand, SCFAs have the potential to decrease retinal inflammation by suppressing the NF-κB pathway, thereby restoring retinal homeostasis in the case of DR. On the other hand, the butyrate found in SCFAs can bind to the GRP43 receptors situated on the surface of intestinal epithelial cells (IECs), thereby repairing and maintaining their homeostasis (Fujiwara et al., 2018). This process prevents harmful products such as TMAO and LPS from entering the bloodstream. Meanwhile, supplementation of DR mice with butyrate not only increased the expression of tight junction proteins (including ZO-1 and occludin proteins) to restore gut barrier function but also further improved the thinning retina and visual function of DR (Huang et al., 2023). Interestingly, supplementing SCFAs in vivo/vitro reduced intraocular inflammation induced by LPS, which ultimately affected the function of retinal glial cells in the eye (Chen et al., 2021). This finding could potentially indicate that gut microecology is a complex and dynamic system. In summary, the majority of studies depend on a single metabolite, which restrains our comprehensive view of the dynamic relationship between disease and gut microecology. When studying gut microecology and DR in the future, it is necessary to comprehensively consider GM and its metabolites. Subsequently, the potential mechanism of action between GM and its metabolites should be further explored through basic experiments.

The gut microecology and its host are in a dynamic equilibrium system. Perhaps we should focus on this whole system, rather than simply supplementing with effective strains and metabolic derivatives of GM flora. FMT plays a significant role in metabolic disorders, such as obesity and diabetes. Yet the evidence for DR and FMT is rather limited. Initially, in preclinical research: Mice that underwent FMT were fed in groups with obese and lean twins, based on the observed coprophagia in mice (Ridaura et al., 2013). It was found that the two affected each other to prevent weight gain and obesity. Obese mice that received FMT from metformin-treated T2DM patients showed changes in the abundance of GM and the expression level of the metabolic derivative TUDCA (Sun et al., 2018). This led to further improvement in glucose intolerance and insulin resistance by inhibiting the FXR signal transduction pathway. Secondly, in clinical research: FMT has been shown to prevent the decrease of insulin production in patients with T1DM, and no associated adverse events were reported during the follow-up period (de Groot et al., 2021). Simultaneously, the study revealed a close relationship between GM flora, its metabolites, and the functionality of residual β cells after FMT treatment. Moreover, the performance of autologous FMT was superior to that of donor FMT. Nonetheless, this study has certain limitations, including the lack of a placebo group for control purposes and no dietary intervention. The study design incorporated these limitations as part of a double-blind randomized controlled trial, which also showed that intervening in a patient’s lifestyle is exceedingly challenging (Ng et al., 2022). On the one hand, FMT treatment increased the abundance of GM in obese individuals with T2DM, particularly butyric acid-producing bacteria. Repeated FMT further improved the implantation of the gut flora. On the other hand, while combining FMT with lifestyle intervention might increase the relative abundance of Bifidobacterium and Lactobacillus, it would not affect the diversity of the GM flora. Finally, the gut-retinal axis has further promoted the research progress of FMT and DR. The expression of key cytokines associated with retinal inflammation can be reduced by transferring GM flora from young to old mice (Parker et al., 2022).

Despite the aforementioned limitations of FMT, it is necessary to consider the following aspects: (1) In animal experiments, antibiotics were generally used to treat participants before FMT was administered. The positive aspect was the ability to observe changes in the gut microecology of participants more objectively and clearly. Another consideration is whether this will lead to a bias in the research results. Furthermore, this method may result in adverse side effects for subjects, leading to ethical concerns and considerations for antibiotic management when applied in clinical studies. This enables us to concentrate on the delivery route of FMT, aiming to discover an implantation method that does not necessitate antibiotic pretreatment, among other options. More evidence is required. (2) Favorable tolerability in both the short and long term, and few adverse events, have been demonstrated in these studies. In order to prevent the spread of disease through FMT, the preparation of the treatment must follow strict standardization guidelines, and the selection of donors should undergo strict screening. (3) As with probiotics, the varying outcomes achieved through FMT may be associated with the particular composition, dosage, efficacy, and individual heterogeneity of the therapy. Moreover, variations in geography, ethnicity, dietary habits, and lifestyles may also contribute significantly. In summary, the mechanism behind the strategy of adjusting gut microecology in DM and its complications needs to be further clarified and proven clinically. Although faced with several challenges, the study of gut microecology is a comprehensive and dynamic exploration that has undeniably opened up a novel field of research.