Cannabis and intraocular pressure: why lowering eye pressure is not enough to treat glaucoma

“I’ve been told that cannabis lowers eye pressure. Is that true?” The honest answer is yes. The effect was first measured in 1971, it has been reproduced since, and no one seriously disputes it. What is missing is not the effect. It is everything else: the duration, the route of administration, tolerance, the effects on the rest of the body, and the slightest demonstration that a visual field has ever been preserved.

UNDERSTANDING

The effect is real, it lasts a few hours, and that is where everything is decided

Yes, cannabis does lower intraocular pressure. This is not waiting-room gossip: it is a finding published in 1971 in eleven volunteers, with a reduction of about 30% after inhalation. The authors already noted, in that same article, that pressure returned within a few hours. Half a century later, it is this second sentence that has become the more important of the two.

The misunderstanding comes from the place the phrase “eye pressure” occupies in conversations about glaucoma. Intraocular pressure is the only risk factor we know how to act on, and lowering it remains the basis of every existing treatment. But it is not the disease. The disease is the progressive loss of optic nerve fibres, which continues night and day, without the slightest symptom until an already advanced stage.

A treatment for glaucoma is therefore not a substance that lowers pressure. It is a substance that keeps it low permanently, for decades, without damaging anything else along the way. An eye drop taken in the morning is still working the next morning. It is this property, not the strength of the effect, that makes something a medicine.

The whole question lies in this gap. A product whose action lasts three to four hours leaves, every day, twenty hours during which pressure has returned to its starting level. Reviews published over the past ten years converge on this precise point, whatever their nuances elsewhere.

By what mechanism does THC lower pressure?

Through receptors that occur naturally in the eye. The endocannabinoid system was not invented to respond to cannabis: it is present in the ciliary body, which produces aqueous humour, and in the trabecular meshwork, which drains it. Tetrahydrocannabinol (THC) binds there and affects both sides of the balance, slowing production slightly and, above all, facilitating outflow.

Three receptor families are involved. CB1 is the best documented, in the ciliary body. CB2 is present in the trabecular meshwork, where it modulates cell migration and matrix organisation. A third, GPR18, has been identified in mice, where it regulates pressure independently of the first two. CB1 also interacts with the beta-2 adrenergic receptor, the very one targeted by beta-blocker eye drops.

One clarification is needed, and it holds for a good half of the literature on this subject: these mechanisms have been described in mice, in rabbits, or in cultured trabecular meshwork cells. Topical THC in mice lowers pressure by about 28% for eight hours, and this effect even depends on the animal’s sex. This is an elegant finding, but it does not allow us to predict either the duration or the magnitude of what would happen in a patient.

This caveat is not merely theoretical. A retrospective analysis of more than 37,000 UK Biobank participants looked for the genetic signature of these receptors in the pressure reduction observed in users. It did not find one: neither CB1 nor GPR18 explained the effect measured in humans. Its authors concluded that the case for a cannabis-based glaucoma treatment remained doubtful.

Three to four hours, then tolerance

Kinetics are the main obstacle, and they have been known since the earliest studies. Taken orally, pressure reaches its minimum at around two hours and the effect has gone by the fourth. Given intravenously, the reduction peaks between thirty and ninety minutes. Applied topically, a synthetic cannabinoid agonist tested in humans lowered pressure by about 20 to 30% at one hour, with no significant difference from placebo by the second hour.

In practical terms, this means six to eight doses to cover a single day, every day, for the rest of a patient’s life. No medical protocol has ever worked at that pace, and this is not a question of discipline: it is a question of cumulative dose.

One trial tried. Nine patients received oral THC, 2.5 to 20 mg four times a day. Tolerance developed in all nine, meaning that the same dose produced a progressively weaker effect. Only one patient completed the planned thirty-six weeks; the other eight stopped between the third and the twenty-eighth week. Only four had reached the target pressure, with reductions of 1 to 5 mmHg.

This order of magnitude is worth comparing with current treatments. Prostaglandin analogues, given as a single daily instillation, lower pressure by about 5.6 mmHg; carbonic anhydrase inhibitors, by about 2.5 mmHg. The studies from the 1970s were comparing cannabis with a therapeutic arsenal that has since been completely renewed. That work has not been disproved: it has been superseded.

Every route of administration has been tried

Smoked, oral, sublingual, intravenous, in eye drops: none has led anywhere. This is not for lack of trying, but the number of patients involved remains tiny. Since 2000, there have been only a handful of human studies on cannabinoids and intraocular pressure, with sample sizes of five to twenty-four participants. Ophthalmology reviews put it bluntly: this level of evidence does not support any recommendation.

The most frequently cited series remains one of eighteen glaucoma patients who inhaled a calibrated cigarette. Pressure fell by 4.1 mmHg at thirty minutes, with a maximum of 6.6 mmHg at ninety minutes. At the same time, five participants developed postural hypotension, eight had anxiety with a racing heart, nine had conjunctival redness with a drooping eyelid, and all eighteen had an altered level of consciousness. These effects are part of the outcome.

The topical route seemed the obvious solution: why expose the whole body when the target is the eye? It runs into a chemical obstacle. Cannabinoids are very poorly soluble in water, which for a long time required oily vehicles that were poorly tolerated by the ocular surface. Recent formulations markedly improve on this, but they have been evaluated in rabbits and mice.

Two cannabis derivatives do, however, exist in pharmaceutical form. Dronabinol, synthetic THC, lowers pressure in healthy volunteers at a dose of 7.5 mg. Nabilone produces a reduction of about 28%, peaking between two and four hours. Neither is approved for glaucoma, and neither has been the subject of a trial measuring change in visual field.

CBD is not a THC stripped of its drawbacks

This is the most widespread idea on the subject, and the most mistaken. Since cannabidiol has no psychoactive effect, it is tempting to assume it would offer the ocular benefit without the feeling of being under the influence. The only human trial to compare the two says the opposite. In six patients treated sublingually, 5 mg of THC significantly lowered pressure at two hours, while 20 and 40 mg of CBD did no better than placebo. The 40 mg dose even caused a transient rise at the fourth hour.

The experimental work that followed partly explains why. In mice, CBD not only fails to lower pressure, it blocks the effect of THC when the two are given together: the two molecules do not work in the same direction. A CBD nanoemulsion applied topically in mice reduced inflammation of the ocular surface without changing pressure.

Other preclinical findings point in a more favourable direction, and honesty requires citing them too. In a perfused porcine anterior segment, CBD increases aqueous humour outflow in under an hour. In normotensive rabbits, a CBD nanoemulsion lowered pressure by up to about 4.5 mmHg, peaking at around two and a half hours.

This work justifies continuing the research, and nothing more. None of it has been conducted in humans, and the only available human trial remains the one that saw pressure rise. A patient who replaced their eye drops with CBD oil would be swapping a treatment with a demonstrated effect for a product whose ocular effect, in humans, has not been demonstrated.

What cannabis does elsewhere while it is lowering pressure

One objection comes up regularly in the literature, more subtle than the one about psychoactive effects: cannabis also lowers blood pressure. Yet the optic nerve is supplied by vessels whose perfusion depends on the difference between blood pressure and intraocular pressure. Lowering both at the same time could, in theory, cancel out the benefit. This question has not been settled.

The human data do not all point the same way, either. In twenty-four healthy volunteers, 5 mg of dronabinol increased blood flow at the optic nerve head without changing pressure. This single finding argues against fears of hypoperfusion, but it involved twenty-four healthy people observed on one occasion, which does not allow any conclusion for a patient treated over twenty years.

The rest of the adverse effects are better understood. The benchmark meta-analysis on medical cannabis use, covering seventy-nine trials and more than six thousand participants, finds an increased risk of short-term adverse effects: dizziness, dry mouth, nausea, drowsiness, disorientation. In a 75-year-old patient who still drives and is already taking several medications, these effects are far from trivial.

There is also the question of smoke, when that is the chosen route. Combustion exposes the ocular surface and the airways to the same substances as tobacco smoke. Yet in the cohorts that have examined risk factors for glaucoma, it is smoking, not cannabis, that emerges as the associated factor.

In regular users, the retina works less well

This is the least familiar side of the subject, and the most solidly established in humans. A series of studies conducted in France used electroretinography to compare the electrical activity of the retina in regular users with that of controls. The signal from the ganglion cells — the very cells that form the optic nerve — is delayed in users: about 98.6 milliseconds versus 88.4, a median lag of 8.4 milliseconds.

The same group found abnormalities at other stages of retinal processing: delayed cone responses at both the onset and the offset of the stimulus, and abnormal oscillatory potentials attributed to amacrine cells. The severity of these abnormalities varies with the level of use.

These delays are not visible and do not reduce the acuity measured on a letter chart: they reflect a slowing of signal processing, not damage. One finding is worth knowing: in forty users enrolled in a reduction programme, some parameters improved after eight to twelve weeks. The phenomenon therefore appears to be at least partly reversible.

Other studies, using a driving simulator, find a reduction in contrast sensitivity, stereoscopic vision and recovery after glare. Lastly, dry eye is a frequent reason for consultation among regular users; this is sometimes how the subject first comes up, well before any question about pressure. Our advice on dry eye then applies without any special modification.

Does cannabis cause glaucoma?

No, according to current data. The question deserves to be asked plainly, because it worries patients who do not dare bring it up. The most robust approach to answering it is Mendelian randomisation, which uses genetic variants associated with use as a substitute for random allocation. Applied to primary open-angle glaucoma, it finds no association: the odds ratio is 1.04 for lifetime use and 0.97 for cannabis use disorder, with confidence intervals spanning no effect.

The UK Biobank cohort reaches the same conclusion: cannabis use was not significantly associated with the occurrence of glaucoma, whereas age and smoking were. One secondary observation is nevertheless worth mentioning: participants who reported at least a hundred uses had received their diagnosis around ten years earlier than the others.

This finding does not mean that cannabis makes the disease appear earlier. It may reflect differences in age, lifestyle or healthcare-seeking behaviour between the groups, and the authors do not interpret it as a cause-and-effect relationship. I mention it because it is part of the data, and because an article that kept only the reassuring results would not be trustworthy.

The useful conclusion is therefore twofold. Nothing indicates that using cannabis causes glaucoma; nor does anything indicate that it protects against it, despite the pressure reduction observed over a few hours. A one-off effect on a risk factor does not automatically translate into an effect on a disease that plays out over thirty years.

Neuroprotection: a promising avenue, still confined to the laboratory

This is the most interesting argument in favour of cannabinoids, and the one that must be handled with the greatest caution. The hypothesis: independently of pressure, certain cannabinoids might directly protect retinal ganglion cells, through an action on oxidative stress, excitotoxicity and inflammation. If confirmed in humans, this would change the question, since it would target the disease itself rather than its risk factor.

The experimental findings are real. Activation of the endocannabinoid system preserves ganglion cells in animal models. Cannabinol, a minor derivative, combines neuroprotective action with a pressure reduction in these same models. A systematic review devoted to CB1, CB2 and TRPV1 receptors in animal models of glaucoma catalogues this entire body of work.

The important word in the previous sentence is “animal”. None of these experiments has been reproduced in humans, and the history of neuroprotection in ophthalmology is a long list of molecules that protected rodents perfectly well and changed nothing in clinical trials.

One finding, moreover, calls for caution in the opposite direction: in BALB/c mice, systemic administration of THC causes retinal damage. The molecule cannot, therefore, be presented as intrinsically protective of the retina; depending on the dose, the route and the species, it either protects or damages. Only a clinical trial can settle a question like this.

What the medical societies say, what the law says, and what I tell my patients

Ophthalmology reviews published over the past ten years converge: cannabis is not recommended as a treatment for glaucoma, because of the brevity of its effect, tolerance, systemic effects, and the absence of data on visual field. The American Academy of Ophthalmology and the American Glaucoma Society state this explicitly: they do not recommend cannabis, in any form whatsoever, to treat glaucoma. This position is consistent, including in publications that acknowledge the value of research into cannabinoids.

The gap with public perception is measurable. An analysis of seven hundred social media posts found 72% of opinions favourable to using cannabis against glaucoma, against 18% unfavourable, the latter coming overwhelmingly from ophthalmologists and medical media.

Legally, in France, cannabis is classified as a narcotic. A regulatory framework governs the medical use of certain products derived from it, but it covers a limited number of clinical situations, of which glaucoma is not one, and no cannabinoid medicine holds an authorisation for this indication under it. By contrast, if you have a risk factor, glaucoma screening does follow a well-established pathway.

Frequently asked questions

Cachan practice · Tel. +33 1 45 47 08 11

Scientific sources

• Joshi N, et al. Cannabinoids for the Treatment of Glaucoma: A Review. Med Cannabis Cannabinoids 2024. PMID 39474241
• Passani A, et al. Cannabinoids in Glaucoma Patients: The Never-Ending Story. J Clin Med 2020. PMID 33302608
• Wang MTM, Danesh-Meyer HV. Cannabinoids and the eye. Surv Ophthalmol 2021. PMID 32763339
• Bondok M, et al. Adverse Ocular Impact and Emerging Therapeutic Potential of Cannabis and Cannabinoids: A Narrative Review. Clin Ophthalmol 2024. PMID 39629058
• Novack GD. Cannabinoids for treatment of glaucoma. Curr Opin Ophthalmol 2016. PMID 26840343
• Sun X, et al. Marijuana for Glaucoma: A Recipe for Disaster or Treatment? Yale J Biol Med 2015. PMID 26339209
• Rafuse P, Buys YM. Medical use of cannabis for glaucoma (editorial). Can J Ophthalmol 2019. PMID 30851777
• Elhusseiny AM, et al. Lifestyle modifications and environmental risk factors for glaucoma. Curr Opin Ophthalmol 2025. PMID 39446700
• Bhartiya S, et al. Complementary and Alternate Management of Glaucoma: The Verdict so Far. J Curr Glaucoma Pract 2014. PMID 26997809
• Panahi Y, et al. The arguments for and against cannabinoids application in glaucomatous retinopathy. Biomed Pharmacother 2017. PMID 28027538
• Pinar-Sueiro S, et al. Cannabinoid applications in glaucoma. Arch Soc Esp Oftalmol 2011. PMID 21414525
• Ige M, et al. Herbal Medicines in Glaucoma Treatment. Yale J Biol Med 2020. PMID 32607093
• Benbadis SR, et al. Medical marijuana in neurology. Expert Rev Neurother 2014. PMID 25427150
• Whiting PF, et al. Cannabinoids for Medical Use: A Systematic Review and Meta-analysis. JAMA 2015. PMID 26103030
• Katsimpris A, et al. Cannabis use and the risk of primary open-angle glaucoma: a Mendelian randomization study. Sci Rep 2023. PMID 37949880
• Lehrer S, Rheinstein PH. Cannabis smoking and glaucoma in the UK Biobank cohort. J Fr Ophtalmol 2022. PMID 35093258
• Lehrer S, Rheinstein PH. Cannabis, Intraocular Pressure, and the Growth Arrest-Specific 7 (GAS7) Gene: A Retrospective Analysis. Cureus 2022. PMID 35411287
• Yakobashvili D, et al. Public Perception of Marijuana Use for the Treatment of Glaucoma. J Glaucoma 2023. PMID 36897646
• Miller S, et al. Δ9-Tetrahydrocannabinol and Cannabidiol Differentially Regulate Intraocular Pressure. Invest Ophthalmol Vis Sci 2018. PMID 30550613
• Aebersold AS, et al. The Effects of Cannabidiol on Aqueous Humor Outflow and Trabecular Meshwork Cell Signaling. Cells 2022. PMID 36230968
• Senapati S, et al. Cannabidiol Loaded Topical Ophthalmic Nanoemulsion Lowers Intraocular Pressure in Normotensive Dutch-Belted Rabbits. Pharmaceutics 2022. PMID 36559077
• Rebibo L, et al. The dose-dependent effect of a stabilized cannabidiol nanoemulsion on ocular surface inflammation and intraocular pressure. Int J Pharm 2022. PMID 35245638
• Warjri GB, et al. Role of Endocannabinoids in Glaucoma: A Review. J Curr Glaucoma Pract 2025. PMID 40417140
• Caldwell MD, et al. A GPR18-based signalling system regulates IOP in murine eye. Br J Pharmacol 2013. PMID 23461720
• Kumar A, Song ZH. Effects of palmitoylethanolamide on aqueous humor outflow. Invest Ophthalmol Vis Sci 2012. PMID 22589443
• Rapino C, et al. Neuroprotection by (endo)Cannabinoids in Glaucoma and Retinal Neurodegenerative Diseases. Curr Neuropharmacol 2018. PMID 28738764
• Gallo Afflitto G, et al. CB1R, CB2R and TRPV1 expression and modulation in in vivo, animal glaucoma models: A systematic review. Biomed Pharmacother 2022. PMID 35468582
• Somvanshi RK, et al. Cannabinol modulates neuroprotection and intraocular pressure: A potential multi-target therapeutic intervention for glaucoma. Biochim Biophys Acta Mol Basis Dis 2022. PMID 34921975
• Zhang Z, et al. Systemic administration with tetrahydrocannabinol causes retinal damage in BALB/c mice. Hum Exp Toxicol 2020. PMID 31680560
• Kokona D, et al. Endogenous and Synthetic Cannabinoids as Therapeutics in Retinal Disease. Neural Plast 2016. PMID 26881135
• Schwitzer T, et al. Association Between Regular Cannabis Use and Ganglion Cell Dysfunction. JAMA Ophthalmol 2017. PMID 27930757
• Schwitzer T, et al. Variations of retinal dysfunctions with the level of cannabis use in regular users. Front Psychiatry 2022. PMID 36465284
• Schwitzer T, et al. Delayed on- and off-retinal responses of cones pathways in regular cannabis users. J Psychiatr Res 2021. PMID 33636687
• Polli L, et al. Oscillatory potentials abnormalities in regular cannabis users: Amacrine cells dysfunction as a marker of central dopaminergic modulation. Prog Neuropsychopharmacol Biol Psychiatry 2021. PMID 32860840
• Ortiz-Peregrina S, et al. Effects of Smoking Cannabis on Visual Function and Driving Performance. Int J Environ Res Public Health 2020. PMID 33287427
• Ortiz-Peregrina S, et al. Effects of cannabis on visual function and self-perceived visual quality. Sci Rep 2021. PMID 33462319
• Ortiz-Peregrina S, et al. Comparison of the effects of alcohol and cannabis on visual function and driving performance. Drug Alcohol Depend 2022. PMID 35717788
• Dhingra D, et al. Illicit drugs: Effects on eye. Indian J Med Res 2019. PMID 31719293
• Hepler RS, Frank IR. Marihuana smoking and intraocular pressure. JAMA 1971. PMID 5109652
• Merritt JC, Crawford WJ, Alexander PC, et al. Effect of marihuana on intraocular and blood pressure in glaucoma. Ophthalmology 1980;87(3):222-228.
• Newell FW, Stark P, Jay WM, Schanzlin DJ. Nabilone: a pressure-reducing synthetic benzopyran in open-angle glaucoma. Ophthalmology 1979. PMID 530560
• Hommer N, Kallab M, Szegedi S, et al. The effect of orally administered dronabinol on optic nerve head blood flow in healthy subjects. Clin Pharmacol Ther 2020. PMID 31977076
• American Glaucoma Society. Position statement: marijuana and the treatment of glaucoma. 2009. PMID 20160576 — American Academy of Ophthalmology, Does marijuana help treat glaucoma? aao.org

This article is intended for information purposes and does not replace a medical consultation. Indications, expected benefits and risks depend on each individual situation. For any question concerning your vision, consult an ophthalmologist.