The Supplement Nobody Questioned
What a 2023 Harvard study found about vitamin A and RP -- and why nobody told us
By Mark G. Hubers -- Just an Engineer with USH
Researched and written: February 22, 2026 | Rev B — March 12, 2026
If you have retinitis pigmentosa, you’ve probably heard this advice: take 15,000 IU of vitamin A palmitate — a specific supplemental form of vitamin A — every day. Your doctor may have told you. A website may have told you. The National Eye Institute still tells you right now, today, on their website.
I still see it in Facebook groups and community chats. Someone asks what they can do, and the replies come: “Take vitamin A.” “My doctor put me on it years ago.” “It’s the one thing that might help.” People share this advice because they are trying to help. Because for thirty years, it was the one thing you could DO.
That recommendation is based on a single study from 1993.
In 2023, researchers at Harvard went back to that same data and did something nobody had done before. They broke it down by gene. Not “RP patients” as one group — but which RP gene each person actually had. 587 patients. 53 different genes identified.
What they found should have changed everything.
What the 1993 Study Said
In 1993, Dr. Eliot Berson at Harvard published a landmark trial. 601 people with RP. Four to six years of follow-up. His conclusion: 15,000 IU of vitamin A palmitate per day slowed vision loss. Avoid vitamin E — it made things worse.
That became the standard advice. For thirty years, retinal specialists prescribed vitamin A to RP patients. The National Eye Institute put it on their website. Patient organizations repeated it. It became one of the few things you could actually DO.
And it made sense at the time. In 1993, genetic testing for RP barely existed. You could not easily tell one form of RP from another just by looking at the retina. So they treated “RP” as one disease and measured whether the whole group did better on vitamin A. The average said yes — a small but measurable slowing of decline on the electroretinogram (ERG), which measures how much electrical signal your retina produces.
The problem is that average was hiding something. When you mix hundreds of people with dozens of different genetic causes into one group, the numbers blur together. One gene might be getting worse. Another might be staying flat. Another might be declining slowly for reasons that have nothing to do with the vitamin. Average them all and the math can say “small positive effect” even if no individual group is truly benefiting. That is what happened here.
That is exactly what happened. When the 2023 team broke the same data apart by gene, the benefit vanished. The only genotype that even hinted at a positive response was one specific mutation in the rhodopsin gene (RHO P23H) — and even that was not statistically significant. Every other gene showed either no effect or harm.
But by then the advice had been baked into medical guidelines, government websites, and doctor visits for three decades. Most of us never got a handout. Most of us had to ask. You sit in the chair, you hear the news, and you say “is there anything I can do?” And they say “well, there is vitamin A.” And you trust it because what else are you going to do. Or your family goes home and searches for answers — and the first thing they find is that same advice on every website. Nobody went back to check because nobody had the genetic tools to ask the right question until now.
Nobody asked: does it work the same for every gene? Including me. I took it too. I will get to that.
What the 2023 Study Found
Thirty years later, a team at the same institution — Mass Eye and Ear, Harvard — went back to the original trial data with modern genetic tools. They sequenced the DNA of 765 patients from those original trials and identified the specific gene causing each person’s RP.
Two findings. Both bad. One worse.
First: vitamin A does not help any form of RP. Not Usher syndrome type 1. Not type 2. Not type 3. Not non-syndromic RP. None of them. When they broke the data down by gene, there was no genotype that showed a clear, lasting benefit from vitamin A. Thirty years of advice, and it did not hold up.
Second: for one group — the biggest one — vitamin A did not just fail to help. It made things worse.
That group is USH2A. 65 patients in the progression analysis — the largest single genetic subgroup. The most common genetic cause of both Usher syndrome and non-syndromic RP.
For USH2A patients, vitamin A showed a statistically significant adverse effect — the group taking it declined faster than the group that did not.
A few things to understand about this finding. USH2A is not one thing. There are over 1,200 known pathogenic variants in the USH2A gene. This study grouped them all together. It was also a subgroup analysis — meaning the researchers went back into the original trial and looked at one piece of the larger group. That kind of analysis does not carry the same weight as a dedicated trial.
The researchers themselves acknowledged this. In their own words: “We speculate that if a large study was conducted in any specific subgroup, then the potential adverse effects would be unlikely to be replicated.” They knew the limitations.
But they also said this: “Notwithstanding the many limitations of any statistical test, the final statistical calculation was well powered to detect an effect of vitamin A in the USH2A subgroup, and the observed effect was adverse.”
For those who want the numbers: the measure of how much worse they did (regression coefficient) was -0.04 — negative means the vitamin A group declined faster. The chance this happened by coincidence (P value) was 2%. In research, if that number is under 5%, the result counts as real. Two percent is real. The effect was small. But it was there.
Knowing all of that, the researchers did not hedge their recommendation. Their exact words:
“For patients with RP associated with USH2A mutations who are on vitamin A supplements, we make a recommendation to stop supplementation.”
Let me be clear about who this affects:
If you have USH2A (Usher syndrome type 2A or USH2A-related RP): The data showed a statistically significant adverse effect. The researchers say stop. Bring the study to your doctor and have the conversation.
If you have a different type of RP or Usher syndrome (USH1, USH3, other genes): The data says vitamin A does not help. It showed no overall benefit for any genotype. The researchers stopped recommending it for all new RP diagnoses and for children with RP.
If you do not know your genotype: That is part of the problem. You might be taking a supplement that is hurting you and not even know it because nobody tested your DNA.
Their current position applies to everyone: “We currently do not recommend vitamin A or E supplementation for patients with RP.”
Where the Information Stopped
Here is where it gets bad.
That paper was published in August 2023. It has been over two years. Here is what happened to the USH2A finding as it traveled from the researchers to the patients:
The researchers said it clearly. Stop vitamin A if you have USH2A.
The Foundation Fighting Blindness covered the study. They reported “no overall benefit from vitamin A.” They did not mention the USH2A-specific harm.
The Usher Syndrome Coalition covered the study. They said vitamin A was “not effective enough to recommend.” They did not warn their USH2A community that it may be actively harmful.
Retina UK covered the study. They said patients already on vitamin A could continue under supervision. No USH2A warning.
The National Eye Institute page still says vitamin A “may help slow vision loss.” When I first found it, that page had not been updated since 2008. It has since been updated — August 2025 — two full years after the Harvard study was published. Someone touched that page and still did not add the finding. No mention of the 2023 study. No mention of genotype-specific effects. No warning for USH2A. It is the first result many patients find when they search for help.
GeneReviews — the clinical reference doctors actually use to look up Usher syndrome — still says vitamin A “may limit progression” for USH2. That entry was last updated in March 2023, five months before the paper was published.
I looked for patient-facing resources that specifically say “if you have USH2A, stop taking vitamin A.” I searched patient forums. Facebook groups. Reddit. Blue Book. Organization websites. Doctor Q&A pages.
I found nothing. Not one.
The researchers said stop. The information never reached the people who need it.
I am not blaming these organizations. I respect what they do and I want to find ways to help. They are small teams doing important work with limited resources. Monitoring every study and pushing genotype-specific updates to every patient who needs them is a massive job. Nobody has figured out how to do it well yet.
But the gap is real. A finding this important should not sit in a journal for two years without reaching the people it is about. Someone needs to read these studies, put them in plain language, and get them in front of the community. Someone needs to check back and make sure the information stays current. That is all I am trying to do here.
“It Can Only Help or Do Nothing”
That is what most people think about supplements. Vitamins are natural. They support your body. Worst case, you pee out what you do not need. Right?
That is the assumption. And for most healthy people taking a daily multivitamin, it is probably close enough to true.
But for RP patients taking high-dose vitamin A — 15,000 IU per day, ten times the recommended daily amount for a healthy adult — it was never that simple. This is not a multivitamin. This is a therapeutic dose based on a specific clinical trial. And when the researchers went back and checked who it was actually helping, the answer was: nobody. And for USH2A, the largest group, it was doing the opposite of helping.
I still see this assumption in Facebook groups and community chats. Someone asks about supplements, and the replies come in: “I take vitamin A, my doctor said it might slow things down.” “Can’t hurt to try.” “I’ve been on it for years.”
It can hurt. The data says so. And the reason nobody knows is that the people who found it did not get the word out far enough.
Why This Matters Right Now
This is not a minor footnote about a rare subtype.
USH2A is the most common cause of Usher syndrome — about half of all cases. It is also one of the most common causes of non-syndromic RP. When the researchers broke their data down by gene, USH2A was the single largest group. Nearly one in four genotyped participants.
For that group — the biggest one — vitamin A is harmful.
For everyone else with RP, vitamin A does not help.
Either way, the thirty-year-old advice was wrong. Some people are taking a supplement that does nothing. Some are taking one that is making things worse. And both groups are doing it because nobody updated the recommendation.
What About Vitamin E?
The original 1993 study found vitamin E was harmful to RP patients. The 2023 re-analysis confirmed it across all genotypes. This matters because AREDS2 — the supplement formula commonly recommended for age-related macular degeneration — contains 400 IU of vitamin E.
If you have RP and someone recommends AREDS2 for your eyes, that formula was designed for a different disease. The vitamin E in it may accelerate your vision loss. Two eye conditions, two supplement recommendations, and they directly contradict each other. Nobody connects the dots because nobody is looking at both diseases in the same patient. This is the kind of thing that gets to me — people following advice that sounds right but nobody checked the facts behind it.
What Else Is Out There
Not everything is bad news. But after what you just read, you deserve honesty about what we know and what we do not:
Lutein (12 mg/day): One trial of 225 patients (Berson, 2010) showed modest preservation of mid-peripheral visual field. No harm found. I had not heard of this one until I started researching for this essay — and that is part of the problem. But here is what I ran into while checking it: that trial was done the same way as the 1993 vitamin A study. All RP patients in one group. No breakdown by gene. We just spent half this essay explaining why that method hid the truth about vitamin A. I cannot tell you lutein is safe for every genotype because nobody has checked. It is the supplement with the least concerning data right now, but “no harm found in the average” is exactly what they said about vitamin A for thirty years.
I almost let this slide myself. I wrote this section, reviewed it multiple times, and did not catch the contradiction until I read it again and thought: wait — did anyone check this by gene? They did not. The same gap I am writing about almost made it into my own essay. So I am not going to tell you lutein is safe for your gene. What I will say is that someone needs to do what the 2023 team did for vitamin A — go back and break the lutein data down by genotype. Until then, the people most likely to try it are the ones who feel they do not have much left to lose. That is how a lot of us learn what works. And that is a sad fact about where we are right now.
NAC (N-acetylcysteine): Targets the oxidative cascade that kills cone photoreceptors after rods die. A Phase 3 trial with 483 patients across 31 sites is actively recruiting. NAC is also available over the counter. Ask your retinal specialist.
Omega-3 from diet: Eating one to two servings of oily fish per week was associated with slower decline in the original studies. Diet-based, not mega-dose supplements. And unlike the others on this list, omega-3 from fish is good for your heart, your brain, and your joints too — so even if the retina benefit is small, you are not wasting your time. I just learned the brain part while writing this. Given how much brain power these essays take, I should probably eat more fish.
How I Found This
I did not set out to research this. I have USH2A. Two different mutations in the same gene, both broken — what genetics calls compound heterozygous. Those are the two they have found so far. Every time a doctor looks, they find something else. I have stopped asking.
I took vitamin A too. About a year. I stopped — not because I knew about this study, but because I never trusted it. Most supplements are not studied. I am not here to debate that. But I took it because my family said “do anything to help.” And when someone you love is watching you lose your vision, you do things for them more than for yourself. I got tired of taking something I did not believe in, so I stopped. Turns out that was the right call.
I mention this because I know the pressure. It is not just doctors recommending it. It is the people around you who need to feel like something is being done. That is real. And it makes it harder to stop even when the evidence says you should.
A friend emailed me about his 80-year-old friend who had been told he had macular degeneration for forty years. A new retinal specialist said it was actually RP. The doctors were contradicting each other and his friend did not know who to believe.
I started looking into whether RP patients can also get age-related macular degeneration as they age. That led me to supplements. That led me to the AREDS2 question. That led me to the 2023 Harvard study. That led me to the realization that nobody told us.
One email. One question from a friend about his friend. And here we are.
Here is the part I did not expect.
When I shared what I found, my friend wrote back. He had been taking vitamin A himself. His blood test in December showed elevated levels and his doctor told him to stop temporarily. He was planning to restart after his next blood test in May.
He is not restarting now.
His question about his 80-year-old friend led to the research that is now protecting him. He had no idea he was helping himself by asking about someone else.
This is what community does. We ask each other questions. And sometimes those questions lead somewhere nobody expected — including right back to ourselves.
We Are Getting Better at This
I do not want this essay to sound like everything is broken and nobody cares. That is not true.
The reason the 2023 study exists at all is because genetic tools caught up. In 1993, you could not sequence a patient’s DNA for a reasonable cost. Today you can spit in a tube and know your exact mutations in a few weeks. That is real progress. The science IS getting more precise. Treatments ARE becoming genotype-specific. The era of “all RP is the same” is ending.
The problem is not the science. The problem is the lag between what researchers find and what patients hear. A study gets published. An organization summarizes it. A website does not get updated. A doctor reads the old guideline. A patient takes what they are told to take.
That lag is shrinking. Genetic testing is more accessible. Patient communities share information faster than journals ever could. People like the ones reading this are asking better questions and demanding better answers.
But we are not there yet. The NEI page was updated in 2025 and still says vitamin A may help — two years after the study said otherwise. GeneReviews still says vitamin A may help. And somewhere right now, someone with USH2A is taking a supplement they believe is protecting their vision.
We need to close that gap faster.
What You Should Do
I am not a doctor. I am not telling you to stop taking anything. I am telling you what a 2023 study from Harvard found, and what the researchers who ran it now recommend.
If you have USH2A and you are taking vitamin A: The data showed a statistically significant harmful effect. The researchers explicitly said to stop. Bring the study to your retinal specialist and have the conversation.
If you have a different form of RP and you are taking vitamin A: The data showed no benefit. The researchers no longer recommend it for any RP patient. Same advice — bring it to your doctor and decide together.
If you do not know your genotype: That is the first thing to fix. You might be taking something that is hurting you and not know it because nobody tested your DNA. A genetic test can tell you. GeneDx and Invitae both offer retinal disease panels. Insurance usually covers it. The Foundation Fighting Blindness has a program called My Retina Tracker that can help you get tested.
Know your gene. Then decide what to put in your body. I wish someone had told me that thirty years ago.
What This Does NOT Mean
If you took vitamin A for years and you have USH2A — do not panic. The adverse effect found in this study was small. This was one re-analysis of one trial, using a subgroup of 65 people. It is a signal, not a sentence.
It does not mean vitamin A poisoned you. It does not mean years of supplementation caused major damage. It means the data showed a small but real negative effect, and the researchers who found it said to stop. That is what you should know. That is what you should bring to your doctor.
If you gave vitamin A to your child because a doctor recommended it, you did what any good parent would do. You followed the best advice available at the time. The science changed. The advice should change with it. That is not your fault.
The point of this essay is not to scare anyone. The point is that the information should have reached you sooner.
A Note on Getting It Right
I care about accuracy more than being first. Everything in this essay is sourced from published, peer-reviewed research. The links are below so you can read the original studies yourself.
If I got something wrong, tell me. Send me the study. I will update this, note what changed, and credit you for the correction. I would rather be corrected than spread bad information.
This essay is dated and revisioned. That is not decoration. You just read about a government health page that went seventeen years without an update, got updated, and still did not include a major finding. Dates matter. Revisions matter. If the science changes, check my revision history — the essay changes with it.
Sources
Comander J, Weigel DiFranco C, Sanderson K, et al. Natural history of retinitis pigmentosa based on genotype, vitamin A/E supplementation, and an electroretinogram biomarker. JCI Insight. 2023;8(15):e167546.
Published August 8, 2023. DOI: 10.1172/jci.insight.167546
Berson EL, Rosner B, Sandberg MA, et al. A randomized trial of vitamin A and vitamin E supplementation for retinitis pigmentosa. Archives of Ophthalmology. 1993;111(6):761-772.
Berson EL, Rosner B, Sandberg MA, et al. Clinical trial of lutein in patients with retinitis pigmentosa receiving vitamin A. Archives of Ophthalmology. 2010;128(4):403-411.
Foundation Fighting Blindness. “New Report: Vitamin A Supplementation Provides No Vision Benefit to RP Patients.” November 2, 2023.
National Eye Institute. “Retinitis Pigmentosa.” Last modified August 6, 2025. Still states vitamin A “may help slow vision loss” without referencing the 2023 genotype-specific findings.
Acknowledgment: This essay exists because Brian Viens emailed me about his friend. One question led to all of this. Thank you, Brian, for asking the right question.
Mark G. Hubers — Just an Engineer with USH
Revision history:
Draft v1 — February 22, 2026
Rev B — March 12, 2026: Added researchers’ own caveats about subgroup analysis limitations. Noted USH2A heterogeneity (1,200+ variants). Softened language to match what the data actually showed. Added “What This Does NOT Mean” section addressing guilt and fear. Thanks to Dr. Paul Glover (USH Coalition Ambassador, NZ) for the feedback that made this revision better.
Have a correction? Reach me through Substack or the USH Blue Book.
Very impressive!! This is very helpful!
Great essay! Keep this up!