Blog
August 22, 2026
Written for eye-care professionals. This is the referenced evidence review. A plain-language version for patients and families is published separately: MSM Eye Drops: What Patients Should Know Before Using Them.
Printable edition. This bulletin is also available as a three-page PDF, formatted for print: Eye Notes, Issue 1 (PDF, 3 pages).
This bulletin is general education, not medical advice. If you are being treated for glaucoma, macular degeneration, diabetic retinal disease or any other eye condition, continue the care your ophthalmologist has prescribed and speak with them before adding anything.
Patients who have watched their vision change, and who have been told there is nothing further to do, tend to arrive with a bottle in hand. More often than not it is a bottle of MSM eye drops. The question they ask is whether the ingredient in it works. That is the wrong question, and answering it honestly means starting somewhere less satisfying.
The right question has three parts. Has this specific, sterile product been shown to reach the tissue it is meant to act on? Does it improve an outcome that matters to a person, not a marker in a dish? And does it stay safe with repeated use over years, not weeks?
Most eye products sold directly to consumers, MSM drops among them, have never been asked any of the three. That is not an accusation. It is a description of how the market works, and understanding it is more useful than any verdict on a single ingredient.
A drop bypasses defenses the rest of the body has. It sits on living tissue that does not regenerate the way skin does, and it does so repeatedly, often for years. A safe ophthalmic product must therefore do considerably more than contain a chemically interesting molecule.
It must be sterile through its entire shelf life and every use, with controlled particulate and endotoxin burden. It needs tolerable pH and osmolality, excipients that are ophthalmically appropriate rather than merely food-grade, a reproducible delivered dose, and a container-closure system that protects the formulation from the first drop to the last.
None of that is bureaucratic detail. Published studies repeatedly document microbial contamination of multidose ophthalmic containers in both clinical and patient-use settings, and contaminated solutions have been linked to serious corneal infection since at least the 1980s.
The 2023 outbreak of extensively drug-resistant Pseudomonas aeruginosa traced to a preservative-free imported artificial tear is the modern reference case. The final CDC figures, confirmed in the peer-reviewed outbreak analysis, were 81 patients across 18 states, four deaths within thirty days among those with clinical cultures, four eyes removed, and fourteen patients with vision loss, eleven of them legally blind at thirty days.
That product contained no exotic active ingredient. It failed on manufacturing.
A product can be entirely reasonable in a capsule and entirely unsuitable in a dropper bottle. Route is not a detail.
Even a well-made drop faces a delivery problem that is easy to underestimate. Tear turnover, blinking, nasolacrimal drainage, the corneal epithelial barrier, conjunctival absorption into the systemic circulation, and the blood-ocular barriers all work against a topically applied molecule reaching the retina or optic nerve at a meaningful concentration.
A drop can coat the ocular surface, feel soothing, and never approach a therapeutic concentration in the tissue it is advertised to help. An entire field of ophthalmic formulation research, including nanocarriers, in-situ gels, punctal plugs and suprachoroidal delivery, exists precisely because this is difficult.
So a claim that a simple topical solution reaches the macula, the optic nerve, the lens, or the vitreous is a pharmacokinetic claim. It requires pharmacokinetic evidence. Analogy to what the same molecule does when swallowed is not evidence.
Preservatives reduce microbial growth in multidose containers. Chronic exposure to some of them, benzalkonium chloride above all, can adversely affect the ocular surface in susceptible patients, which is why preservative-free glaucoma medications exist.
But preservative-free systems do not remove the contamination problem. They relocate it, to single-use packaging or to validated multidose closure technology. Preservative-free on a label is therefore not a safety claim. It describes which of two problems the manufacturer chose to solve, and says nothing about whether they solved it.
Being available for purchase does not establish that a product is FDA-approved, legally marketed, sterile, or clinically tested.
A company can place a product into commerce before the FDA has inspected it or taken any action. Agency oversight combines premarket requirements with postmarket surveillance, inspections, sampling, adverse-event reports, warning letters, import actions, recalls and seizures. Those resources are allocated by risk, and enforcement is neither instantaneous nor evidence that a product was previously lawful.
Labelling also does not change the category. A product intended to treat, mitigate or prevent disease is regulated as a drug regardless of whether it is described as natural, homeopathic, traditional, or a supplement. Dietary-supplement rules govern things that are swallowed. They do not govern products placed in the eye.
Methylsulfonylmethane, or MSM, also known as dimethyl sulfone, is a small organosulfur compound. Patients ask about it more than any other ingredient in this category, and it is worth walking through carefully, both because the answer matters and because the reasoning transfers to whatever ingredient is popular next year.
The FDA position is a matter of public record: there are no legally marketed ophthalmic drugs that contain MSM as an active ingredient. That statement appears in the 2023 consumer alert and remains on the consumer eye-drops page, which was last refreshed in June 2026.
In August 2023 the FDA warned consumers not to use two specific marketed MSM eye-drop products after agency testing found bacterial contamination in both and fungal contamination in one. Warning letters to both firms followed. Separately, a third company had already recalled its MSM drops earlier that year for non-sterility, a recall the FDA classified in its most serious category.
They condemn the products that were tested and recalled, and they illustrate what inadequately controlled ophthalmic manufacturing produces. They do not establish that every bottle containing MSM is contaminated, and nobody should say that they do.
The durable point is the category-level one: no MSM ophthalmic product has been through an approval process in which effectiveness, safety, quality and labelling were established.
It is worth pausing on the permeability finding, because it is often presented as though it were a benefit in itself. Increasing how much of a formulation crosses into the eye also increases exposure of sensitive tissue to everything else in the bottle: the active, the excipients, any impurities, and any contaminants.
Whether that is helpful or harmful depends on the entire sterile formulation, the dose, the osmolality, the pH, the container-closure system, where the material distributes, and a validated manufacturing process. Enhanced penetration of a contaminated solution is not an advantage.
The honest case for clinical oversight is not that a supervised programme has been proven to work better than a bottle. It has not been, and any such claim would require evidence that does not exist. The case is narrower, checkable, and about process.
A clinician can confirm the diagnosis and its urgency, and screen for red flags that need same-day attention. They can review every prescribed drug, over-the-counter product and supplement for interactions, including anticoagulant and antiplatelet risk, pregnancy, liver or kidney disease, and planned surgery. They can establish baseline measurements and repeat them on a defined schedule.
And they can interpret those measurements against what is known about their variability, which is where the value actually sits.
In the Ocular Hypertension Treatment Study, 86% of visual field abnormalities detected during follow-up were not confirmed on repeat testing. Nothing counts until it is reproduced. For structural imaging, the intervisit tolerance limit for average retinal nerve fibre layer thickness is roughly 3.9 microns, and a change of about 4 to 5 microns is the threshold at which a decrease begins to be meaningful rather than noise.
Those numbers cut in both directions. They are why a patient should not be alarmed by one bad field, and equally why a patient should not be reassured by one good one, or by feeling that something has helped after a week. A programme that treats every fluctuation as a response is not measuring anything.
The practical benefit of prespecified monitoring is that it can distinguish real change from test-retest variation, detect non-response, catch a safety signal, and identify when urgent referral is needed. It also makes it possible to stop something that is not working, which self-directed treatment rarely does.
A bottle cannot examine you, cannot read your chart, and cannot tell you to stop.
This bulletin is published by a clinic that provides integrative eye care, so the standard applied above has to be applied here too.
The Cochrane review of acupuncture for glaucoma concluded that available studies were insufficient to determine effectiveness, with very low-certainty evidence. It was published in 2020, but its literature search closed in November 2018, so its evidence base is now roughly eight years old. A systematic review of acupuncture for age-related macular degeneration rated the evidence low to very low and found no high-quality evidence of effectiveness. Dry-eye studies are more numerous and somewhat more encouraging on symptoms, but blinding, sham controls, heterogeneity and durability remain real limitations.
Botanical research relevant to retinal and optic-nerve disease, including astragaloside IV, Salvia miltiorrhiza and Lycium barbarum, is largely preclinical: cells, and rodent models of ocular hypertension, ischaemia and photoreceptor degeneration. That work is useful for prioritising compounds and designing trials. It is not a substitute for product-specific clinical research, and an herb name does not specify a drug: species, plant part, extraction solvent, marker compounds, contaminants, dose and batch consistency all change what is actually being given.
Herbal products also interact with medicines. Goji, or Lycium barbarum, is the subject of published case reports of potentiated warfarin effect. Acupuncture has a low rate of serious harm in trained hands but is not free of adverse events, and one prospective glaucoma trial reported a transient rise in intraocular pressure immediately after eye-point sessions.
Any preparation intended to contact the eye, whether an oil, a wash, a ghee-based preparation, a botanical, or anything else, faces the same ophthalmic standard set out at the start of this bulletin. The FDA position is that there are no FDA-approved Ayurvedic products, and that Ayurvedic products marketed to treat disease are illegally marketed in the United States; the agency has documented heavy metals and alkaloids in marketed preparations. Natural is not a synonym for sterile, ophthalmically appropriate, or tested.
Five tests, in the order that eliminates the most claims fastest.
One, route. Was the research done by the same route as the product? A trial of an oral supplement says nothing about a drop. This single test disposes of most of what circulates.
Two, species. Cells and rodents establish plausibility and help design trials. They do not establish benefit in people. A phrase like shown to protect retinal cells almost always means cultured cells or mice, and the gap between that and a human outcome is where most promising compounds die.
Three, the exact product. An ingredient is not a product. Species, plant part, extraction method, dose, formulation and batch all change what is actually being given. A study of one extract is not evidence for a different preparation carrying the same name on the label.
Four, the endpoint. Did something change that a person would notice, such as acuity, visual field, reading, mobility or quality of life, or did a marker move? A biomarker shifting is a reason to run a trial. It is not a result.
Five, who is telling you. Ask whether the party making the claim sells the thing. That includes this bulletin.
Clinically proven is a claim about the state of the evidence, not about the product, which makes it checkable in a way that efficacy claims are not. If a specific product has been tested in a controlled human trial, that trial has investigators, a registration number, and a published result. Ask for all three.
A reference list of mechanism papers is not the same thing. A long one is not better than a short one; it is a different kind of claim wearing the same clothes.
The instinct to dismiss it is understandable and usually counterproductive. A patient who feels judged for trying something will simply stop mentioning what they are taking, which removes the interaction screen and the opportunity to catch a safety problem early.
A more useful sequence: ask what they are using, how often, and for how long, and record it in the chart alongside prescribed therapy. Check for interactions, particularly anticoagulant and antiplatelet risk with botanicals. Ask whether the product is single-use or multidose, and whether the bottle tip has contacted the eye or lid. Establish what they expect it to do, and what would tell them it was not working.
Then say plainly what is and is not known, and make clear that prescribed treatment and monitoring continue unchanged. Patients who are told the truth about the state of the evidence tend to stay in care. Patients who are told nothing tend to substitute.
Nothing described here substitutes for the care your ophthalmologist directs. Pressure-lowering therapy, anti-VEGF injections, laser treatment, surgery, urgent management of infection and scheduled monitoring all stay exactly where they are. Do not reduce or stop any of them unless the clinician who prescribed them changes the plan.
Sudden loss of vision. Flashes or a shower of new floaters, particularly with a curtain or shadow across the field. Severe eye pain. Trauma or chemical exposure. Acutely red eye with light sensitivity. Any new symptom after eye surgery.
Published by Netra Eye Institute, which provides integrative eye care as a paid clinical service. This is not independent third-party product research.
It is general education, not medical advice, and it is not a diagnosis or treatment recommendation for any individual.
No programme or product described here has been shown in controlled trials to change the course of any eye disease. Where evidence exists it is identified above, along with its limitations. Individual responses vary.
Regulatory statements are current as of 22 August 2026 and should be reconfirmed against the agency source before being relied upon.
FDA, What You Should Know About Eye Drops (content current 15 June 2026).
FDA, FDA Warns Consumers Not to Purchase or Use Certain Methylsulfonylmethane (MSM) Eye Drops Due to Contamination (22 August 2023; content current 30 August 2023).
Grossman MK, et al. Extensively drug-resistant Pseudomonas aeruginosa outbreak associated with artificial tears. Clin Infect Dis. 2024;79(1):6-14. Final case figures from CDC.
Zhang M, et al. Assessment of methylsulfonylmethane as a permeability enhancer for regional EDTA chelation therapy. Drug Deliv. 2009;16(5):243-248.
Liu P, et al. Metal chelator combined with permeability enhancer ameliorates oxidative stress-associated neurodegeneration in rat eyes with elevated intraocular pressure. Free Radic Biol Med. 2014;69:289-299.
Hwang JC, Khine KT, Lee JC, Boyer DS, Francis BA. Methyl-Sulfonyl-Methane (MSM)-induced acute angle closure. J Glaucoma. 2015;24(4):e28-e30.
Law SK, Wang L, Li T. Acupuncture for glaucoma. Cochrane Database Syst Rev. 2020;2(2):CD006030. Literature search closed November 2018.
Sun W, et al. Effects of acupuncture on age-related macular degeneration: a systematic review and meta-analysis. PLoS One. 2023;18(3):e0283375.
Baudouin C, et al. Preservatives in eyedrops: the good, the bad and the ugly. Prog Retin Eye Res. 2010;29(4):312-334.
Ahmed S, Amin MM, Sayed S. Ocular drug delivery: a comprehensive review. AAPS PharmSciTech. 2023;24(2):66.
FDA, FDA Warns About Heavy Metal Poisoning Associated With Certain Unapproved Ayurvedic Drug Products (updated 2 December 2025).
Full reference list, including all 84 peer-reviewed sources underlying this bulletin, available on request.
