Ocular Effects of
GLP-1 Receptor Agonists

GLP-1 Receptor Agonist Agents

The GLP-1 RA class encompasses several structurally diverse agents with varying receptor selectivity, pharmacokinetics, and routes of administration. Understanding these differences is critical to interpreting ocular safety signals, as agents differ in their degree of HbA1c lowering, rapidity of glucose reduction, and receptor distribution in ocular tissues.

Structural Classification

GLP-1 Receptors in Ocular Tissues

GLP-1 receptors (GLP-1Rs) are expressed throughout the eye, establishing a biological basis for direct ocular effects beyond systemic glycaemic changes. Key sites of expression include:

• Retinal ganglion cells (RGCs): GLP-1R activation promotes neuroprotection and modulates intraocular pressure regulation
• Retinal pigment epithelium (RPE): Involved in VEGF signalling modulation and barrier function
• Müller glial cells: Central to retinal homeostasis; GLP-1R signalling affects inflammatory cascades
• Optic nerve head: Potential site mediating NAION risk through vascular tone alterations
• Corneal epithelium and stroma: May explain tear film and corneal sensitivity changes
• Ciliary body: Role in aqueous humour dynamics and IOP modulation

Mechanisms of Ocular Effect

The ocular effects of GLP-1 RAs arise through multiple interrelated pathways. These can be broadly categorised into indirect effects (mediated through systemic metabolic changes) and direct effects (mediated through GLP-1R activation in ocular tissues).

Insulin-Mediated IGF-1 Pathways

An important additional mechanism involves the interaction between GLP-1 RAs and insulin secretion. Improved β-cell function increases endogenous insulin levels, which in turn stimulates hepatic IGF-1 production. In the retina, IGF-1 acts on IGF-1R to upregulate VEGF expression. This is particularly relevant in patients with pre-existing ischaemic retinopathy, where IGF-1-driven VEGF may tip an ischaemic retina toward neovascularisation.

Direct GLP-1R-Mediated Neuroprotection

Paradoxically, GLP-1R activation in retinal ganglion cells and glial cells exerts neuroprotective effects through several pathways:

• Activation of the cAMP/PKA pathway reduces apoptotic signalling in RGCs
• Inhibition of NF-κB-mediated neuroinflammation in Müller cells
• Reduction of oxidative stress via Nrf2 activation
• Upregulation of BDNF and CNTF (neurotrophic factors protective for RGCs)
• Anti-angiogenic effects via reduced VEGF expression in non-ischaemic retina

Vascular Mechanisms — NAION Pathway

The mechanism by which GLP-1 RAs may precipitate NAION is distinct from the retinopathy mechanism and involves:

• Optic nerve head vasculature: GLP-1Rs in posterior ciliary arteries modulate vasoconstriction; receptor activation may reduce perfusion pressure at the optic nerve head
• Rapid weight loss-related hypotension: Particularly with semaglutide and tirzepatide causing 10–15%+ body weight reduction, nocturnal hypotension in anatomically predisposed optic disc configurations ("disc at risk") may precipitate ischaemia
• Haematological changes: GLP-1 RAs reduce red cell mass and haematocrit in some patients, potentially reducing oxygen-carrying capacity to already-vulnerable optic nerve head
• Anti-platelet effects: GLP-1 RAs modulate platelet aggregation, which may alter microthrombus formation at the optic nerve head

Diabetic Retinopathy & GLP-1 RAs

The SUSTAIN-6 Trial — The Pivotal Study

The SUSTAIN-6 cardiovascular outcomes trial (Marso et al., NEJM 2016) was the seminal study raising concern about GLP-1 RA-associated retinopathy. This double-blind RCT randomised 3,297 patients with T2DM at high cardiovascular risk to subcutaneous semaglutide (0.5 mg or 1.0 mg weekly) versus placebo for 104 weeks.

Critically, the risk was concentrated in patients with pre-existing retinopathy (37% had this at baseline) and in those who experienced the most rapid HbA1c reduction. Patients without baseline retinopathy did not show a significantly increased risk.

LEADER Trial — Liraglutide

The LEADER cardiovascular outcomes trial (Marso et al., NEJM 2016) evaluated liraglutide in 9,340 patients with T2DM and high cardiovascular risk. Unlike SUSTAIN-6, LEADER did not demonstrate a statistically significant increase in retinopathy complications. However, the liraglutide-associated HbA1c reduction was more gradual (−0.4% versus −0.9% for semaglutide at 36 weeks), supporting the "rate of change" hypothesis.

Other Cardiovascular Outcomes Trials

The "Early Worsening" Phenomenon — Temporal Course

This phenomenon — well-recognised from insulin therapy — occurs within the first 3–12 months of treatment. Key temporal characteristics include:

• Onset: Typically 3–6 months after initiation of therapy or dose escalation
• Peak risk window: 6–12 months post-initiation
• Resolution: In many patients, retinopathy stabilises or improves after 12–24 months as sustained glycaemic control provides long-term benefit
• Severity: Most concerning in patients with moderate-to-severe NPDR or early PDR at baseline

The early worsening of retinopathy with rapid HbA1c normalisation is a well-described phenomenon first identified with intensive insulin therapy in the DCCT trial. GLP-1 RAs appear to trigger the same pathophysiological cascade, albeit at varying rates depending on the potency and dose of the agent used. — Adapted from Diabetes Care, 2023

Classification of Diabetic Retinopathy — Baseline Risk Stratification

Long-Term Protective Effects

Despite the short-term worsening risk, accumulating data suggest that in patients who tolerate GLP-1 RA therapy and achieve sustained HbA1c control, the long-term trajectory of diabetic retinopathy is favourable. Animal models demonstrate GLP-1R-mediated reduction in retinal pericyte loss, microaneurysm formation, and VEGF expression under normoglycaemic conditions. Long-term observational data from registries (follow-up >3 years) show reduced rates of retinopathy progression and need for retinal laser therapy compared to patients treated with sulfonylureas at equivalent HbA1c targets.

Non-Arteritic Anterior Ischaemic Optic Neuropathy (NAION)

What is NAION?

Non-arteritic anterior ischaemic optic neuropathy (NAION) is the most common acute optic nerve disorder in adults over 50. It is caused by infarction of the short posterior ciliary arteries supplying the anterior portion of the optic nerve, resulting in acute, painless, monocular vision loss.

Clinical Features

• Acute, painless monocular visual loss — typically present on waking
• Altitudinal or arcuate visual field defect (most commonly inferior field loss)
• Relative afferent pupillary defect (RAPD)
• Hyperaemic, oedematous optic disc with peripapillary flame haemorrhages
• Characteristic "disc at risk" — small optic disc with no physiological cup (cup-to-disc ratio <0.2)
• No effective treatment — supportive management only

Epidemiology — Background Rate

• Incidence: approximately 2.3–10.3 per 100,000 persons per year in the general population
• Higher in diabetic patients (estimated 2–3× baseline risk)
• Bilateral involvement: contralateral eye affected in 15–25% within 5 years
• No proven effective treatment; spontaneous partial recovery in ~40% of cases

The Harvard/Mass Eye & Ear Study (2024)

This retrospective cohort study included patients from a large ophthalmology practice with T2DM or obesity, comparing semaglutide users to non-semaglutide users (matched controls on insulin, non-GLP-1 antidiabetics, and non-GLP-1 weight-loss medications).

Mechanistic Hypotheses for NAION

Several mechanisms have been proposed, though none definitively confirmed:

1. Nocturnal Hypotension from Rapid Weight Loss

Patients using semaglutide or tirzepatide for obesity treatment can lose 10–15% of body weight within months. This rapid weight reduction is associated with significant reductions in blood pressure (both daytime and nocturnal). In patients with a "disc at risk" anatomy — a small crowded optic disc with minimal physiological cup — even modest reductions in nocturnal perfusion pressure may cause critical hypoperfusion of the optic nerve head during sleep-associated drops in blood pressure.

2. Direct Vascular Effects on Posterior Ciliary Arteries

GLP-1Rs expressed in vascular smooth muscle of the posterior ciliary arteries may alter vascular tone. In animal models, GLP-1R activation has been shown to modulate both vasodilatory (via cAMP/NO pathway) and vasoconstrictive responses, potentially creating conditions of intermittent relative ischaemia at the optic nerve head.

3. Haematological Changes

GLP-1 RAs, particularly with weight loss, are associated with reductions in red cell mass and haematocrit. Combined with lower blood pressure, this may reduce oxygen delivery to the already watershed-zone-dependent optic nerve head vasculature.

4. Anti-Platelet Effects

GLP-1 RAs have demonstrated modest anti-platelet effects in some studies — potentially altering microthrombus dynamics at the optic nerve head, either protective or harmful depending on context.

Risk Factors for NAION in GLP-1 RA Users

Regulatory Status (2024–2025)

• FDA: Reviewing safety data; issued a Drug Safety Communication noting the JAMA Ophthalmology study but has not yet modified labelling pending further data
• EMA: Pharmacovigilance review ongoing; no formal label change as of early 2025
• MHRA (UK): Yellow Card signals reviewed; clinicians advised to report any suspected cases
• Novo Nordisk: Conducting post-marketing surveillance studies; issued guidance for healthcare professionals to counsel patients about NAION risk

Differential Diagnosis — NAION vs Arteritic ION

Other Ocular Effects

Diabetic Macular Oedema (DME)

DME represents a common cause of visual loss in diabetes. Its relationship with GLP-1 RAs is complex:

• Short-term risk: Rapid HbA1c reduction may transiently worsen DME through haemodynamic and inflammatory changes — similar to the early worsening phenomenon for retinopathy
• Long-term benefit: Sustained glycaemic control reduces DME incidence and recurrence
• Anti-VEGF interaction: GLP-1 RA-associated VEGF modulation may alter response to intravitreal anti-VEGF therapy (bevacizumab, ranibizumab, aflibercept) — an area of active investigation
• Liraglutide trial data: LEADER showed no significant change in DME rates; however, OCT-based sub-studies showed potential reduction in central subfield thickness with prolonged therapy

Dry Eye Disease (DED)

Several pharmacovigilance datasets (FDA FAERS, EudraVigilance) have identified increased reporting of dry eye symptoms with GLP-1 RA use, particularly semaglutide. Proposed mechanisms include:

• GLP-1R expression in lacrimal glands: Direct receptor-mediated reduction in aqueous tear production
• Weight loss-associated hormonal changes: Reductions in androgen levels (important for meibomian gland function) with significant weight loss
• Reduced blink rate: Anecdotal reports of reduced blink frequency associated with GLP-1 RA-related satiety signals
• Nutritional changes: Dietary restriction may reduce omega-3 intake relevant to tear film quality

The clinical significance of this association remains uncertain, and DED in patients with long-standing diabetes is common at baseline. Prospective, controlled studies are needed.

Corneal Effects

Diabetic corneal neuropathy (reduced corneal nerve density and sensitivity) is common in T2DM and may be modified by GLP-1 RA therapy:

• GLP-1R activation in corneal epithelial cells promotes epithelial healing and reduces apoptosis in vitro
• Animal models demonstrate GLP-1R-mediated improvement in corneal sensitivity and nerve regeneration
• Small human studies (confocal microscopy) show improved corneal nerve fibre density with liraglutide treatment at 12 months
• Clinical implications for contact lens wear, LASIK suitability, and post-surgical healing in GLP-1 RA users warrant further investigation

Uveitis

Several case reports and one pharmacovigilance study have identified an association between GLP-1 RA use and anterior uveitis. The proposed mechanism involves immune modulation:

• GLP-1Rs expressed on immune cells (T-lymphocytes, macrophages) may alter cytokine profiles
• Rapid metabolic changes associated with weight loss alter the systemic immune milieu
• The absolute signal is very low (estimated OR ~1.5–2.0) and may represent confounding by underlying autoimmune conditions common in T2DM

Intraocular Pressure (IOP) & Glaucoma

Preclinical and early human data suggest GLP-1R activation in the ciliary body and trabecular meshwork may reduce IOP. Animal studies show 15–25% IOP reduction with direct intravitreal GLP-1R agonist administration. Systemic GLP-1 RAs show more modest IOP effects (<1–2 mmHg reduction in small trials). Ongoing clinical trials are exploring GLP-1R agonists as a novel class of topical glaucoma therapy — a potentially exciting neuroprotective application.

Lens & Cataract

Osmotic changes in lens epithelial cells from rapid glucose normalisation may transiently alter lens refraction — a phenomenon well-recognised in newly diagnosed T2DM patients. Patients may experience transient blurred vision during the first weeks of GLP-1 RA therapy, which typically resolves with stable glycaemic control. Whether GLP-1 RAs modify long-term cataract risk (beyond glycaemic effects) remains unknown.

Myopia Progression

An emerging and controversial area involves observations that GLP-1 RA-associated weight loss may alter axial length and refraction in some patients. Reduced adipose tissue around the orbit may change orbital pressure dynamics. These reports are primarily case-based and require systematic investigation before clinical significance can be established.

Optic Nerve Neuroprotection

Beyond NAION risk, GLP-1R agonists have demonstrated significant neuroprotective effects on retinal ganglion cells in animal models of traumatic, ischaemic, and toxic optic nerve injury. These effects are mediated through:

• cAMP-mediated inhibition of RGC apoptosis
• Reduction of glutamate excitotoxicity
• Anti-inflammatory modulation in astrocytes and Müller cells
• Enhanced axonal transport and mitochondrial function

These findings have prompted research into GLP-1 RAs for conditions including glaucoma, optic neuritis, and traumatic optic neuropathy — representing a potential therapeutic rather than adverse role for these agents in optic nerve disease.

Risk Stratification & Patient Selection

Identifying patients at elevated ocular risk prior to GLP-1 RA initiation is critical. The following framework assists clinicians in pre-treatment risk assessment:

High-Risk Patient Profile for Retinopathy Worsening

High-Risk Profile for NAION

Protective Factors & Lower-Risk Scenarios

• No baseline retinopathy on screening
• Short duration of diabetes (<5 years)
• Normal optic disc morphology with physiological cup
• Gradual titration protocol (slower HbA1c reduction expected)
• Choosing lower-potency agent (exenatide, lixisenatide) if high-risk
• Normal blood pressure with no antihypertensive therapy
• Treated sleep apnoea

Management Strategies

Pre-Treatment Evaluation

Titration Strategies to Reduce Ocular Risk

• Slow titration in high-risk patients: Extend standard titration schedule — e.g., semaglutide 0.25 mg for 8 weeks instead of 4 weeks before dose escalation
• Target-based HbA1c reduction rate: Aim for HbA1c reduction of no more than 0.5–1% per 3-month period in patients with severe NPDR/PDR — discuss with endocrinologist
• Agent selection: Consider lower-potency agents (liraglutide, dulaglutide) in patients with high baseline retinopathy risk, accepting slower but safer glycaemic normalisation
• Combination with retinal treatment: If significant active DME or PDR — consider initiating anti-VEGF therapy or laser first, stabilising retina, then starting GLP-1 RA

Managing Established Retinopathy on GLP-1 RAs

If retinopathy worsening is detected after GLP-1 RA initiation:

• Do not automatically discontinue — discuss risk-benefit with ophthalmology
• For mild worsening (e.g., progression from mild to moderate NPDR) — continue with enhanced monitoring (every 3 months)
• For significant worsening (new PDR, vitreous haemorrhage, significant DME) — consider dose reduction or switch to lower-potency agent while initiating ophthalmic treatment
• Anti-VEGF therapy (intravitreal ranibizumab, aflibercept, bevacizumab) for DME and PDR should proceed on standard criteria regardless of GLP-1 RA use
• Panretinal photocoagulation (PRP) for high-risk PDR — standard of care; continue GLP-1 RA unless actively unstable

If NAION Is Suspected

Role of Ophthalmology in Shared Care

The prescribing of GLP-1 RAs has increasingly extended to primary care, obesity clinics, and general medicine — settings where ophthalmological expertise may not be immediately accessible. Effective management requires:

• Clear communication pathways between prescribing physicians and ophthalmology departments
• Standardised pre-prescribing checklists incorporating ophthalmic risk factors
• Patient-facing information on visual symptoms requiring urgent attention
• Integration with existing diabetic eye screening programmes
• Collaborative guidelines from ophthalmology and endocrinology societies — several are in development (2024–2025)

Monitoring Protocol

Ophthalmological Monitoring Schedule

Symptoms Requiring Urgent Ophthalmological Assessment

All patients initiating GLP-1 RA therapy should receive written information advising urgent review for:

• Any sudden loss of vision in one or both eyes
• Visual field defect or loss of peripheral vision
• Floaters, flashing lights (photopsia), or a curtain/shadow across vision (possible retinal detachment)
• New onset blurred vision not explained by blood glucose fluctuation
• Eye pain or redness with visual change (possible uveitis)
• Diplopia (double vision)

HbA1c Monitoring & Rate of Change

Monitoring the rate of HbA1c reduction is as important as the absolute target in patients with established retinopathy:

• Check HbA1c at baseline, 3 months, and 6 months after initiation
• In patients with moderate-severe NPDR: if HbA1c drops >2% in 3 months, consider dose reduction and immediate ophthalmology referral
• Target HbA1c reduction rate: <0.5–1% per 3-month interval in high-risk patients

Future Directions & Ongoing Research

Active Clinical Trials (2024–2026)

Topical GLP-1 RA — The Next Frontier

Perhaps the most exciting future direction is the development of topical (eye drop) formulations of GLP-1R agonists. If delivered directly to the eye, these agents could harness the neuroprotective, anti-inflammatory, and IOP-lowering properties of GLP-1R activation without systemic effects. Early preclinical data with topical exendin-4 and semaglutide analogues show:

• Significant IOP reduction (15–25% from baseline in animal models)
• Retinal ganglion cell preservation in glaucoma models
• Reduction in retinal neuroinflammation markers
• Good tolerability with no significant anterior segment adverse effects

GLP-1 RAs for Age-Related Macular Degeneration (AMD)?

An intriguing hypothesis posits that the anti-inflammatory and anti-angiogenic properties of GLP-1R activation could be relevant to AMD pathogenesis. Epidemiological data from T2DM cohorts suggest users of GLP-1 RAs may have lower rates of neovascular AMD progression compared to non-users — though causality has not been established. Preclinical models of choroidal neovascularisation show suppression of CNV lesion size with systemic GLP-1R agonist administration.

Tirzepatide & Triple Agonists — Pending Data

Tirzepatide (GLP-1/GIP dual agonist) achieves even greater HbA1c reduction and weight loss than semaglutide, raising the theoretical concern of more pronounced "early worsening" of retinopathy and greater NAION risk through more severe hypotension. Dedicated ocular safety analyses from SURPASS trials are expected in 2025. Retatrutide (triple GLP-1/GIP/glucagon agonist) in Phase III trials with >20% weight loss figures will require careful post-marketing ocular surveillance.

Pharmacogenomics & Personalised Risk

Emerging research is identifying genetic variants that modify GLP-1 RA ocular risk:

• Polymorphisms in the GLP1R gene affecting receptor sensitivity in ocular tissues
• VEGF gene variants modifying retinal angiogenic response to glucose normalisation
• Genetic determinants of optic disc size and anatomy predisposing to NAION

These findings may eventually enable pharmacogenomic pre-screening to identify individuals at highest ocular risk before GLP-1 RA initiation.

Summary — Key Clinical Takeaways