Advice about coffee and eyes tends to come in two contradictory forms. One says caffeine raises the pressure inside your eye and anyone worried about glaucoma should cut back. The other says coffee drinkers have slightly lower eye pressure and a lower risk of some eye disease. Unusually, this is a case where the contradiction resolves cleanly, and the resolution depends on something you probably do not know about yourself: your genetic predisposition to high eye pressure. That makes this one of the few coffee questions with a genuinely personal answer.
What eye pressure is and why it matters
The eye is filled with fluid that is continuously produced and drained. Intraocular pressure, measured in millimetres of mercury, is the pressure that fluid exerts, and a typical value is in the mid teens. Glaucoma is damage to the optic nerve that causes irreversible loss of vision, usually starting at the edges, and raised eye pressure is its single biggest modifiable risk factor. Not everyone with high pressure develops glaucoma, and some people develop it at normal pressure, so pressure is a strong risk marker rather than the disease itself. Almost all glaucoma treatment works by lowering it.
This is why a drink that measurably raises eye pressure, even briefly, gets studied so carefully.
Caffeine does raise eye pressure, briefly
The acute effect has been measured directly with a built-in control. Ajayi and Ukwade[1] gave healthy Nigerian volunteers aged 20 to 27 either caffeinated or decaffeinated coffee and measured eye pressure and blood pressure at 0, 30, 60 and 90 minutes. Eye pressure rose in the caffeinated group, increasing steadily with time after ingestion, with a mean rise of almost 4 mmHg, alongside rises in systolic and diastolic blood pressure. The decaffeinated group showed essentially no change in either. The difference between groups was statistically significant.
A more recent experiment found the same pattern. Yan and colleagues[2], reviewing caffeine's effects across the eye, describe caffeine blocking adenosine receptors and affecting tear production, intraocular pressure, macular perfusion and choroidal thickness, and note the conflicting findings on whether any of that translates into disease.
So the acute effect is real. Roughly 4 mmHg is not trivial in a system where treatment targets are set in single-digit reductions. The question is what a repeated, temporary rise does over years, and whether habitual drinkers even show it.
Habitual coffee drinkers do not have higher eye pressure
Here the picture reverses. Nakano and colleagues[3] examined 9,850 participants in the Japanese Nagahama cohort, 9,418 of whom had no history of glaucoma, with a mean eye pressure of 14.7 mmHg. Habitual coffee consumption was significantly associated with eye pressure, but in the direction few people expect: the more coffee, the lower the pressure. People drinking coffee 3 or more times daily had pressure 0.4 mmHg lower than those drinking less than once daily. Coffee consumption was not associated with a history of glaucoma.
The UK Biobank analysis by Kim and colleagues[4] found the same direction across 121,374 people with corneal-compensated pressure measurements. Comparing the highest caffeine intake, at 232 mg a day or more, with the lowest, below 87 mg, the difference was 0.10 mmHg lower pressure. Mendelian randomisation, using 8 genetic variants associated with coffee intake, did not support a causal effect of coffee drinking on eye pressure at all, and the overall relationship between caffeine and glaucoma was null.
Both findings are tiny, a few tenths of a millimetre of mercury, and almost certainly not clinically meaningful in either direction. Taken together with the acute experiments, the sensible reading is that caffeine produces a short-lived rise that habitual drinkers largely adapt to, leaving no lasting elevation.
The exception that changes the advice
The UK Biobank study did not stop at the average. Kim and colleagues[4] built a polygenic risk score from 111 genetic variants associated with eye pressure and looked for interaction. Among people in the highest quarter of that genetic risk score, consuming more than 480 mg of caffeine a day compared with less than 80 mg was associated with 0.35 mmHg higher pressure, with a significant interaction. For glaucoma itself, people in the highest genetic risk quarter consuming 321 mg of caffeine a day or more had a 3.90-fold higher prevalence of glaucoma compared with those in the lowest genetic risk quarter consuming no caffeine.
That is a large number and it deserves careful reading. It compares two groups differing in both genetics and caffeine, so it is not the effect of caffeine alone. The interaction test, which is the part that isolates caffeine's contribution within genetic strata, was significant. The authors' conclusion was that habitual caffeine consumption was weakly associated with lower eye pressure and no glaucoma association overall, but that among those with the strongest genetic predisposition to elevated pressure, greater caffeine consumption was associated with higher pressure and higher glaucoma prevalence.
What the long cohorts found before the genetics
Two long prospective studies had already hinted at a subgroup effect without being able to measure it genetically. Kang and colleagues[5] followed 79,120 women from 1980 and 42,052 men from 1986, all aged 40 or over and reporting eye examinations, through 2004, confirming 1,011 cases of primary open-angle glaucoma by medical record review. Overall caffeine intake was not associated with risk: the rate ratios across rising intake categories were 1.05, 1.19, 1.13 and 1.17, with no significant trend. But drinking 5 or more cups of caffeinated coffee daily carried a rate ratio of 1.61, and tea and caffeinated cola showed no association. The adverse association was stronger among people reporting a family history of glaucoma, particularly for glaucoma with elevated pressure.
Pasquale and colleagues[6] then examined exfoliation glaucoma, a distinct form, in the same two cohorts through 2008, confirming 360 cases. People consuming 500 mg of caffeine a day or more showed a non-significant trend toward higher risk, at 1.43. Those drinking 3 or more cups of caffeinated coffee daily had a rate ratio of 1.66, which was significant. Again the associations were stronger among women with a family history of glaucoma.
A family history is a crude proxy for the polygenic score the UK Biobank study could measure directly, and the two lines of evidence agree. That agreement is what lifts this from an interesting subgroup finding to something worth acting on.
Does coffee dry your eyes out?
This is the other common question, and it points the opposite way from what people expect. Caffeine is mildly diuretic, covered in our article on that, and the intuition is that a diuretic should dry the eyes. Magno and colleagues[7] assessed dry eye disease in 85,302 participants in the Dutch LifeLines cohort, with a mean caffeine intake of 285 mg a day. Adjusting for demographics, body mass index, smoking and alcohol, higher caffeine intake was associated with slightly lower odds of dry eye, at 0.971 per 100 mg a day. Once 48 medical comorbidities were added to the model, the effect was no longer significant. Decaffeinated coffee intake was associated with slightly increased risk when adjusted for caffeinated coffee, which more likely reflects who chooses decaf than anything decaf does.
Jeong and colleagues[8] found the same in 9,752 Korean adults. Raw prevalence of dry eye syndrome fell as coffee consumption rose, from 9.2% in those drinking less than 1 cup to 6.3% in those drinking 3 or more, but after adjusting for other risk factors there was no significant relationship. Their conclusion was that there is no relationship between coffee frequency and dry eye risk.
The consistent answer from two large population studies is that coffee is not a cause of dry eye. If your eyes feel dry at your desk, screen time, air conditioning and contact lenses are better suspects than the flat white.
Macular degeneration and the rest
Age-related macular degeneration comes up often because caffeine has been proposed as protective. Electra and colleagues[9] systematically reviewed and meta-analysed 5 studies covering 9,318 patients. Caffeine intake from coffee or other beverages showed no significant association with the occurrence of early macular degeneration, at an adjusted odds ratio of 1.11, or of late disease. Vote-counting across studies suggested caffeine, mainly from tea, may slow progression. Their conclusion was that the protective effect on development remains uncertain, and the evidence base is limited.
The broader review by Yan and colleagues[2] reaches similar conclusions across cataract, diabetic retinopathy, myopia and central serous retinopathy: conflicting evidence, few strong findings, and a specific recommendation that people genetically predisposed to glaucoma be cautious because of the possibility of clinically significant pressure elevation. They also note a practical detail that has nothing to do with disease, that caffeine can cause tremor and may hinder performance in delicate procedures such as microsurgery.
What this means before an eye test
There is one immediately practical consequence of the acute finding. If eye pressure rises by up to around 4 mmHg within 90 minutes of a coffee, as it did in Ajayi and Ukwade[1], then a pressure reading taken shortly after your morning coffee is not measuring your baseline. For someone being monitored for glaucoma, where decisions are made on changes of a few millimetres, that is the difference between a reassuring reading and a worrying one.
No guideline we found tells patients to skip coffee before tonometry, and this is not advice to change what a clinic has asked of you. It is a reason to mention it: if you had a strong coffee on the way to the appointment, saying so gives the person reading the number the context to interpret it. The same logic applies to blood pressure readings, for the same reason, as described in our blood pressure article.
Reading this from the Philippines
Glaucoma is a leading cause of irreversible blindness worldwide, and because early open-angle glaucoma has no symptoms, it is usually found at an eye examination rather than noticed. That is the actionable part of this article for a coffee drinker anywhere, and more so for anyone with a parent or sibling who has it, given how consistently a family history showed up as the modifier in Kang and colleagues[5] and Pasquale and colleagues[6]. The studies above were run in the United States, the United Kingdom, Japan, the Netherlands, Korea and Nigeria, which is a broader spread of populations than most coffee research manages, and the acute pressure finding was consistent across them.
The limits of this evidence
The acute pressure experiments are small, short and mostly in young healthy volunteers, and they measured pressure rather than nerve damage. The large studies are observational, with caffeine intake self-reported, and the UK Biobank analysis was cross-sectional, so it captures glaucoma prevalence rather than new cases developing over time. The genetic interaction is a statistical finding that has not been tested by randomising anyone to drink coffee. And nobody has run a trial asking whether people with high genetic risk who cut caffeine go on to develop less glaucoma, which is the study that would actually settle the advice.
The practical answer
For most people, coffee is not bad for the eyes. Caffeine raises eye pressure for an hour or two, habitual drinkers show no lasting elevation and may sit fractionally lower, and coffee is not associated with dry eye or with developing macular degeneration. The group the evidence singles out is people with a strong genetic predisposition to raised eye pressure, or a family history of glaucoma, at high caffeine intakes, roughly 3 to 5 cups of coffee a day and up. If that is you, this is worth raising with your optometrist or ophthalmologist, who can measure your pressure rather than guess at it. See also our article on coffee and blood pressure, since the same acute rise shows up there, and our article on safe caffeine limits. Nothing here is medical advice.