Notable RGC loss is reported under chronic diabetic situations. High glucose impairs retrograde axonal transport and leads to glutamate accumulation, which induces neurotoxicity. It also activates p38 MARK, ERK1/2 cascade, and JNK cascade, which promotes apoptosis and inflammation (Potilinski et al., 2020). One of GLP-1RA’s most dominant benefits on eye diseases is its direct effect on RGCs’ function, rescue, and survival. According to current studies, the molecular pathways involved in GLP-1RAs’ impact on RGCs’ rescue can be categorized into two groups (Figure 2).

As a premise, we know that the pathology of many neurodegenerative diseases and eye diseases are interconnected. For example, the elevated IOP in glaucoma eyes exerts mechanical stress at the optic nerve head, a key site for RGC damage, and leads to neurodegeneration and visual impairment. Metabolic stress, calcium dysregulation and neuroinflammation mediated by glial cells like astrocytes and microglia further exacerbates RGC loss (Bou Ghanem et al., 2024). Animal studies also discovered the neuroprotective effects of neurosteroids, allopregnanolone and its enantiomer, in treating glaucoma by activating GABA receptors and thus preventing apoptotic cell death of RGC (Ishikawa et al., 2022; Ishikawa et al., 2021). Although not yet directly examined in the eyes of human patients, these shared pathology encourage researchers to draw lessons from GLP-1RAs’ success in neuroprotection and transfer it into treating eye diseases.

In many eye diseases, oxidative stress and mitochondrial dysfunction are key contributors to RGC degeneration. According to our current understanding of GLP-1RAs’ mitochondrial regulatory effect and oxidative stress suppression ability on eye diseases, the molecular pathways mainly involve the PINK1/Parkin-mediated mitophagy pathway and the GLP-1R-ERK1/2-HDAC6 signaling pathway (see Figure 3). These two pathways are connected by focusing on mitochondrial homeostasis and the cellular response to stress, specifically oxidative stress and mitochondrial damage. Both pathways protect retinal cells (including RGCs) from damage induced by high glucose or ischemic conditions.

The role of inflammation lies at the center of various eye diseases’ pathology. Microglia/macrophages (CD11b+ cells) respond to elevated IOP, the clinical feature of glaucoma, and produce three main pro-inflammatory cytokines: IL-1α, TNF-α, and C1q. These cytokines are sufficient and necessary for inducing both astrocyte transformation into neurotoxic A1 and blood-retina barrier destruction, resulting in the subsequent retinal ganglia cell death and various ophthalmopathy (Guttenplan et al., 2020; Liddelow et al., 2017). Apart from microglia/macrophages, Müller cell activation and gliosis are observed in animal models of glaucoma and diabetic retinopathy. As a subtype of glial cells vital for retina structural support, cellular homeostasis maintenance, and blood-retina barrier preservation, its reactivation would decrease l‐glutamic acid uptake, which results in excitotoxicity, as well as sabotage the integrity of the blood-retina barrier, leading to ultimate retinal ganglia cell death and vision damage (Ren et al., 2020). The anti-inflammation effect of GLP-1RAs is shown in Figure 4.

Recent discoveries have revealed the remodeling capabilities of GLP-1RAs in the retina. The following four molecular pathways collectively describe how GLP-1RAs facilitate retinal remodeling, primarily by enhancing vascular function and restoring ECM homeostasis. Each pathway targets specific aspects of vascular and cellular integrity, contributing to the overall protection and remodeling of the retinal vasculature in conditions like DR and ischemic retinal diseases (see Figure 5).

Abnormal angiogenesis plays a significant role in the pathogenesis of several eye diseases, particularly those involving retinal ischemia and neovascularization. Improper blood vessels are prone to leak, bleed, and scar, leading to vision loss and other retinal complications. Several studies have investigated the anti-angiogenic effect of GLP-1RAs in ischemic retinopathies and choroidal neovascularization.

Zhou L. et al. (2021) examined the molecular pathways in GLP-1R modulation of microglia. Their work describes how microglial activation can directly promote angiogenesis by secreting VEGF and other proangiogenic cytokines in ischemic retinopathies. When GLP-1R is activated, microglial activation is suppressed, reducing the levels of VEGF and other cytokines involved in neovascularization. GLP-1R activation also inhibits NF-κB activation through the cAMP/PKA signaling pathway, suppressing inflammatory and angiogenic gene expression in retinal cells, including VEGF.

In the context of wet age-related macular degeneration (AMD), geniposide (GEN), a GLP-1RA, is found to downregulate HB-EGF (heparin-binding epidermal growth factor) secretion from retinal pigment epithelium (RPE) cells under hypoxic conditions through the miR-145-5p/NF-κB signaling axis. GEN enhances the expression of miR-145-5p, which inhibits the NF-κB pathway and thus suppresses inflammatory responses and angiogenesis (Gu et al., 2021).

Studies examining the effect of GLP-1RAs on glaucoma risk in diabetes patients is a research hotspot that emerged in recent years. Separate retrospective cohort studies carried out by Sterling et al. (2023) and Chuang et al. (2024) in different populations both exhibited significantly lower incidences of open-angle glaucoma in diabetic patient groups treated with GLP-1RAs medication compared with the control group. More specifically speaking, Sterling’s group observed a decreased percentage of newly diagnosed glaucoma after GLP-1RAs exposure from 1.33% to 0.51%, and Chuang’s research discovered lowered glaucoma onset with a significant adjusted hazard ratio (aHR) (0.712, 95% CI: 0.533–0.936. P = 0.0025). More recent research has also looked into GLP-1RAs’ beneficial influence on glaucoma but focused on this effect at different time ranges. The nationwide, nested case-control study carried out by Niazi et al. (2024) observed a strong correlation between GLP-1RAs treatment with a 19% decreased likelihood of glaucoma onset, which was particularly notable in patients receiving the agonists for periods extending beyond 3 years. However, another study found a significantly lowered risk of developing glaucoma and ocular hypertension in all patient groups receiving the treatment at 1, 2, and 3 years (41%, 50%, and 41% risk reduction, respectively) (Muayad et al., 2024). The controversy surrounding whether a shorter treatment duration could lead to significant protection results may stem from the fact that the earlier study primarily enrolled patients in second-line diabetes treatments. This approach could reduce the dosage of individual drugs, such as GLP-1RAs, compared to the latter study. A recent retrospective cohort also observed that, even in overweight patients without diabetes, the risk of primary open-angle glaucoma and ocular hypertension were significantly lower in the GLP-1RA group at both 3 and 5 years of follow-up (Vasu et al., 2025), suggesting the benefit of GLP-1RAs in the eye as an independent factor apart from blood glucose control.

Apart from this, previous studies have also examined the specific influence of GLP-1RAs on IOP changes, RGCs’ survival, etc. Hallaj and colleagues discovered that exposure to GLP-1RAs was strongly associated with reduced IOP in diabetic patients after glaucoma surgery (Hallaj et al., 2024). Another study by Lawrence’s team confirmed the effect of GLP-1RA drugs on RGCs’ rescue in hypertensive mouse models of glaucoma (Lawrence et al., 2023). The findings strongly added to the potential benefit of GLP-1RAs in reducing glaucoma risks. They urged researchers to dive deeper into the mechanism behind this correlation to better understand glaucoma pathology, hoping to develop novel management approaches toward the disease.

The number of trials and research examining the relationship between GLP-1RAs and DR is growing rapidly in recent years. A meta-analysis in 2023 examined 93 clinical trials from ClinicalTrials.gov that compared GLP-1RAs’ direct or indirect impact on DR with placebo, insulin, or oral antidiabetic medications (Kapoor et al., 2023). The results suggested that the use of GLP-1 RAs was notably linked to a higher risk of early-stage DR and early retinal adverse events when compared to placebo. However, when compared to insulin, GLP-1 RAs provided protection against advanced-stage DR. What’s more, its demographic analysis showed significant differences between the GLP-1 RA and comparator groups in factors such as age, HbA1c, body weight, BMI, duration of diabetes, sex, race, and so on, suggesting that the effect of GLP-1RAs on DR is yet controversial and that it may vary based on the specific GLP-1RA used, as well as the patient’s demographic and clinical characteristics.

A Swedish cohort study examined the impact of multiple GLP-1RAs on the incidence of DR in 2,390 diabetic patients. It matched patients treated with GLP-1RAs with those who did not use the medication and found that GLP-1RAs were associated with a significantly lower risk of DR, with a hazard ratio (HR) of 0.42. Mendelian randomization analysis also showed an inverse relationship between GLP-1 receptor expression and severe DR, indicating the potential of GLP-1RAs in reducing the risk of DR in type 2 diabetes patients (Zheng et al., 2023). However, the SUSTAIN six trial evaluated semaglutide, a GLP-1 analog, in patients with type 2 diabetes and high cardiovascular risk and discovered that semaglutide was associated with a higher incidence of diabetic retinopathy complications compared to placebo, with a hazard ratio (HR) of 1.76, particularly in patients with pre-existing DR (Vilsbøll et al., 2018). This suggests that semaglutide may cause early worsening of DR, especially in patients with poor glycemic control, highlighting the importance of monitoring DR during GLP-1 RA therapy. Apart from trials supporting the two opinions mentioned above, other results denied the possible impact of GLP-1RAs on DR. The EXSCEL trial (Bethel et al., 2020) and LEADER trial (Marso et al., 2016) investigated the effect of exenatide and liraglutide, respectively, on microvascular and cardiovascular outcomes in patients with type 2 diabetes. Despite the drugs’ cardiovascular benefits, both trials found no significant difference in retinopathy outcomes between the GLP-1RAs and placebo groups.

This seemingly controversial impact of GLP-1RAs on DR may be because most trials focused on microvascular and cardiovascular outcomes rather than DR. Thus, the direct influence of GLP-1RAs on DR still calls for future investigation.

Apart from glaucoma and DR, the association between GLP-1RAs and other ocular events has also been investigated in recent years. Hao et al. (2022) observed promising therapeutical effects of SM102, a novel GLP-1RA, on hyperglycemia, weight control, and diabetic dry eye syndrome via its antioxidant properties. A cohort study by Fan’s team (Fan et al., 2023) also confirmed a significantly reduced risk of dry eye syndrome, superficial keratitis, and infectious keratitis in diabetic patients treated by GLP-1 compared with placebo. A pathology research observed inflammation, fibrosis, reduced tear secretion and decreased level of GLP-1R in the lacrimal gland of diabetic dry eye. Liraglutide restored proper autophagic flux by alleviating the accumulation of P62 (an autophagy dysfunction marker) in the lacrimal gland, reduced oxidative damage by enhancing clearance of ROS and promoting cellular repair mechanisms, and potentially addressed dry eye syndrome (Sun et al., 2025). What’s more, a recent retrospective, multicenter cohort study denied higher occurrence of nonarteritic anterior ischemic optic neuropathy (NAION) in GLP-1RAs users and suggested that its potential benefits for blood glucose control and cardiovascular health likely outweigh its risks (Chou et al., 2025), despite that a few research have reported association between semaglutide prescription and a higher risk of NAION(61). Regarding age-related macular degeneration (AMD), a recently published retrospective cohort study reported that GLP-1RAs were associated with reduced hazard of non-exudative and exudative AMD compared to metformin, insulin and statins (Allan et al., 2025). Although no mechanism hypothesis has yet been posed upon GLP-1RAs’ impact on most adverse ocular events, these primary findings still require that clinical workers monitor GLP-1RA users’ eye condition. The information of current clinical studies and trials studying GLP-1RAs’ effect on various eye diseases is summarized in Table 1.

Despite progress in understanding molecular pathways such as calcium channel modulation and mitochondrial mitophagy, the relative contributions of these mechanisms to specific diseases remain unclear. GLP-1RAs inhibit RGC apoptosis in glaucoma through GABAergic signaling, but their role in AMD, characterized by oxidative stress and choroidal neovascularization, needs further investigation (Shao et al., 2024). Additionally, most mechanistic studies rely on animal models, raising concerns about translatability to humans due to interspecies differences in GLP-1R distribution and signaling (Drucker, 2022). Human retinal organoids and patient-derived cell lines could bridge this gap by providing disease-relevant platforms for mechanistic interrogation.

The latest diabetic retinopathy preferred practice guidelines from the American Academy of Ophthalmology indicate that the rising use of GLP-1RAs may lead to quick glycemic control and severe hypoglycemia and has been reported to cause higher rates of retinopathy complications, especially in patients with early-stage DR (Lim et al., 2025). Thus, the clinical application of GLP-1RAs on ocular diseases still requires caution and further research.

Systemic administration of GLP-1RAs reduces their availability in the retina and can lead to off-target effects, particularly gastrointestinal issues. Therefore, localized delivery methods, such as intravitreal implants, nanoparticle carriers, or gene therapy aimed at retinal GLP-1R expression, are essential for maximizing therapeutic effectiveness while minimizing systemic exposure. Recent advancements in sustained-release formulations, like PLGA nanoparticles for delivering exendin-4, show promise in preclinical models but still require validation in human studies (Wang et al., 2023).

Given the promising potential of GLP-1RA in ocular therapeutics, combing it with other current medicine could shed light on novel therapeutic approaches. Sampedro’s team (Sampedro et al., 2019) discovered in mice that treatment with GLP-1RAs resulted in a significant decrease in VEGF levels in the retina, kidneys, and systemic circulation. This finding correlates with reduced tissue damage and improved vascular integrity, and also suggests a synergistic effect with current ocular medications, such as anti-VEGF drugs. A retrospective study explored the impact of prior GLP-1RA treatment on the response of DR patients to subsequent intravitreal anti-VEGF injections and other conventional ophthalmic therapy (Michaeli et al., 2024). The results demonstrated that patients who had previously received GLP-1RA treatment had a more favorable response to intravitreal injections, with improved visual acuity outcomes and reduced central retinal thickness. Additionally, the study highlighted that the need for repeated anti-VEGF injections was modestly lower in the GLP-1RA group, indicating a potential priming effect that enhances the efficacy of conventional ocular treatments. However, since the clinical efficacy of GLP-1RA drugs in ocular diseases is a newly emerging area of research, direct clinical evidence and higher-level trials are still lacking. Combining GLP-1RAs with anti-VEGF agents or IOP-lowering drugs may enhance efficacy but pose unresolved challenges. In DR, GLP-1RAs’ vascular stabilization effects (via RhoA/ROCK inhibition) (Wei et al., 2022) could complement anti-VEGF therapy, yet their interaction in modulating angiogenesis remains unexplored. Administering GLP-1RAs with neuroprotective agents like brimonidine in glaucoma also requires careful dose optimization to prevent additive side effects.

Emerging reports link GLP-1RAs to rare ocular adverse events, including NAION (Hathaway et al., 2024). It is still uncertain whether these risks are caused by direct mechanisms such as changes in blood flow or are influenced by confounding factors like preexisting vascular conditions. Longitudinal registries and post-marketing surveillance studies are essential for assessing long-term eye safety, particularly in older populations with multiple health issues.