Almost everything sold for eye health follows the same arc: a plausible mechanism, some laboratory work, an observational association, and then a product, with the trial that would settle it never arriving.
Nicotinamide is different, which is why it is worth an article. The trial arrived. Two of them. And the answer they produced is genuinely awkward rather than convenient.
Where the idea came from
Glaucoma is usually described as a pressure disease, and pressure is the only thing current treatment lowers. But some people keep losing vision with pressure well controlled, which means pressure is not the whole story.
The alternative account is metabolic. Retinal ganglion cells send long axons out of the eye, they are unmyelinated for the whole intraocular stretch, and sustaining that costs energy continuously. Energy metabolism runs on NAD, and NAD falls with age.
In 2017 a group published in Science that vitamin B3 modulates mitochondrial vulnerability and prevents glaucoma in aged mice. Not slowed. Prevented. Nicotinamide is a precursor the body converts to NAD, and it turns out to be low in the serum of people with glaucoma.
That is a strong opening hand, and it is the same shape as the one omega-3 had before AREDS2 tested it and found nothing at all. Mechanism is where you start, not where you finish.
What the two trials measured
Hui, 2020: fifty-seven people, crossover
A double-masked crossover trial across two tertiary centres. Fifty-seven participants, all already diagnosed and on glaucoma treatment. Six weeks at 1.5 grams a day, then six weeks at 3.0 grams, then crossover.
The main measurement was the photopic negative response, an electroretinography signal that reflects how the inner retina is functioning. Saturated PhNR amplitude rose 14.8%, with a confidence interval of 2.8 to 26.9 and p=0.02.
There was also a trend on visual fields, with 27% improving by at least 1 dB and fewer deteriorating. And one design limitation worth naming: the crossover was performed without a washout period, so any carry-over from the first arm sits inside the second.
De Moraes, 2022: phase 2, thirty-two completing
A single-centre phase 2 trial of nicotinamide at 1000 to 3000 mg combined with pyruvate at 1500 to 3000 mg, against placebo, randomised two to one. Of 197 people assessed and 42 randomised, 32 completed and were analysed. Median follow-up 2.2 months.
More visual field locations improved beyond normal variability on treatment than on placebo: a median of 15 against 7, p=0.005. No serious adverse events were reported.
| Hui 2020 | De Moraes 2022 | |
|---|---|---|
| Design | Crossover, double-masked | Phase 2, placebo-controlled |
| Analysed | 57 | 32 of 42 randomised |
| Duration | 12 weeks | median 2.2 months |
| Intervention | Nicotinamide 1.5 then 3.0 g/day | Nicotinamide + pyruvate |
| Measured | Electroretinography (PhNR) | Visual field locations |
| Already on IOP treatment | yes | yes |
Neither trial asked whether nicotinamide could replace pressure-lowering treatment. Every participant stayed on theirs. This was only ever an add-on question, and any account that drops that detail has changed what was tested.
And then the people who write the guidance read the same papers
In 2025 the American Glaucoma Society and the American Academy of Ophthalmology published a joint position statement on nicotinamide for glaucoma neuroprotection. It is short, and two sentences from it do most of the work.
These clinical studies' endpoints include functional findings detectable in a research setting, but do not consistently translate to visual recovery that is impactful to the patient.
That is the gap between a number moving and a person seeing better, stated by the people best placed to judge it. And then the practical half:
The use of nicotinamide at high doses (≥ 3 grams/day) is not recommended outside the confines of clinical trials.
They add that nicotinamide is not approved for glaucoma, that its safety at these doses is unknown, and that trial use is accompanied by monitoring of liver function.
What it adds up to
Two things that are both true
The case for taking it seriously
- Prevention of glaucoma in aged mice, published in Science
- Nicotinamide measurably low in the serum of glaucoma patients
- Two randomised trials, both finding a signal rather than nothing
- A coherent mechanism that explains progression at controlled pressure
The case for waiting
- 57 and 32 participants, twelve weeks and 2.2 months
- Endpoints measured in a lab, not vision a person notices
- One crossover run without a washout period
- Not approved, safety unknown at dose, not recommended outside trials
Both columns are the honest state of the field. Anyone selling you the left one without the right one is not informing you, and anyone giving you the right one without the left is not either, because there is genuinely something here worth following.
If you have glaucoma, the useful move is not a purchase. It is to take the position statement to the person managing your pressure and ask what they make of it, because trials are recruiting and being in one is the version of this with monitoring attached. We keep a longer technical note on the NAD literature for anyone who wants the primary sources.