Here's what almost nobody tells you about hydrogen water: the science got there first. Long before the podcasts, before the $12 pouches, before any of it — there was a paper in Nature Medicine describing something genuinely new about how an antioxidant could work.

That was 2007. This "trend" is old enough to drive.

Since then: 81 registered clinical trials. A randomized trial showing reduced inflammatory markers in healthy adults. A national regulator in Japan approving hydrogen inhalation for hospital use. The evidence base for molecular hydrogen is real, it's growing, and it's legitimately exciting — and it sits inside a consumer market where a lot of what's for sale can't prove it contains anything at all.

Both of those things are true at once. That's the whole story, and it's more interesting than either the hype merchants or the hedge-everything institutions will tell you.

So here's what we're going to do. Every health claim on this page carries an evidence grade. Every study is named, linked, and reported with its actual sample size — including the one where the result didn't hold. We sell hydrogen products, and we'd rather you buy one understanding exactly what it does than buy one on a promise we can't back. That's how we work across everything else as water optimization specialists, and hydrogen doesn't get an exception.

The short version, before the long one. The mechanism is elegant — hydrogen appears to go after the free radicals that damage you while leaving the ones your cells use as signals. The human evidence is strongest for inflammation and antioxidant capacity, promising for metabolic markers, and genuinely thin for the anti-aging claims the category leans on hardest. Safety looks clean. And how you take it matters enormously: inhalation has the deepest clinical literature, drinking has the most consumer-scale research, and a bottle that shipped three months ago is water with a markup.

Hydrogen water is ordinary drinking water with additional molecular hydrogen (H₂) gas dissolved into it, measured in parts per billion (PPB) or milligrams per litre. The water's chemical formula does not change — the hydrogen is dissolved in the water, not bonded to it, the same relationship carbon dioxide has with sparkling water.

This is where most of the confusion starts, and it's a fair confusion. Water is already two-thirds hydrogen by atom count, so "hydrogen water" sounds either redundant or like a scam on first hearing. The hydrogen inside H₂O is locked into the molecule and biologically unavailable. Dissolved H₂ gas is a separate, free molecule moving between the water molecules, and it is the only kind the research is about. A hydrogen generator infuses that gas into water — it is an addition to your water, not a treatment of it.

At room temperature and atmospheric pressure, water reaches hydrogen saturation at approximately 1.6 milligrams per litre — about 1,600 PPB. Any figure above that is supersaturation, achieved under pressure inside a sealed generation chamber, and it begins declining the moment the seal opens.

Carry that number into every product comparison you do. A device advertising 4,000-plus PPB is claiming supersaturation — which is real, achievable, and lab-verifiable at the moment of generation. What it isn't is a shelf-life figure. Any brand presenting it as one is describing physics that doesn't exist.

Hydrogen can be delivered three ways — dissolved in water and drunk, inhaled as a gas, or absorbed through the skin during immersion — and inhalation carries the deepest clinical literature of the three, including regulated medical use in Japan.

This surprises almost everyone, and it's the most under-reported fact in consumer hydrogen coverage. Ohsawa's founding paper didn't study drinking water at all — it studied inhaled hydrogen gas. Clinical research has largely followed that route ever since, for a simple reason: inhalation delivers a far higher dose than any sane volume of water can.

In Japan, hydrogen inhalation is approved under the Advanced Medical Care B system for adult patients in continuous coma following out-of-hospital cardiac arrest. That is a national regulator permitting hydrogen gas in a hospital protocol — a meaningfully different level of scrutiny than a wellness product clears.

Randomised controlled trials of hydrogen inhalation have reported outcomes in liver disease, sudden hearing loss and post-surgical recovery, in disease populations under clinical supervision rather than in healthy consumers.

A 13-week randomised, placebo-controlled trial of hydrogen/oxygen inhalation in 43 subjects with non-alcoholic fatty liver disease reported improved serum lipids and liver enzymes, with improvement in liver fat content on imaging in moderate-to-severe cases (Tao G, et al. Hepatol Commun. 2022). A double-blinded randomised controlled trial across six hospitals in 65 patients with idiopathic sudden sensorineural hearing loss found significantly greater improvement in hearing threshold change in the hydrogen group (Okada M, et al. Front Neurosci. 2022). Further randomised trials have examined post-operative brain edema and knee osteoarthritis.

The honest answer is that these are clinical investigations in diagnosed patients, not evidence that a wellness user gets a comparable result. What they establish is real and not small: molecular hydrogen does something measurable in humans at clinical doses. What they don't do is convert into a promise about your morning session at home — and we're not going to make one.

GoodFor carries four Lumati hydrogen systems across the three delivery routes: a portable drinking bottle, two inhalation generators at different output grades, and a bath immersion generator for transdermal delivery.

Two specifications are worth understanding before any of the numbers below make sense. The inhalation machines are built around a DuPont Nafion™ N117 proton exchange membrane — the same solid-polymer technology used in fuel cells and laboratory electrolyzers — running on distilled water with no lye or added electrolyte, which is what produces 99.996% pure hydrogen with the gas separated at the source. And both the bottle and the bath generator carry third-party concentration testing by H2 Analytics, which is the criterion this page has been arguing for throughout. Lumati is a member of the American Hydrogen Health Association.

GoodFor is a reseller of Lumati products and Lumati's warranty terms govern. If you're weighing an inhalation or immersion system, that's a conversation worth having with a person rather than a checkout button — book a consultation and we'll talk through the routes, the protocols and whether the spend makes sense for what you're actually trying to do.

Three formats dominate the drinking side of the consumer market: point-of-use electrolysis using a PEM or SPE membrane, magnesium tablets that generate hydrogen through chemical reaction, and pre-bottled hydrogen water — and they differ sharply in achievable concentration, byproduct handling, and how much hydrogen is left by the time you drink it.

Electrolysis splits water into hydrogen and oxygen, and in the presence of chloride or under a poorly separated electrode design it can also generate ozone and chlorine byproducts at the anode. A proton exchange membrane physically separates the two sides so those byproducts can be vented rather than dissolved into what you drink.

This is the mechanical difference between a well-built generator and a cheap one, and it's completely invisible in a product photo. An open-electrode design with no membrane and no vent makes hydrogen and everything else at once, into the same glass you're about to drink from.

Three open questions sit underneath everything above, and they're the reason no claim on this page grades Strong: dosing isn't standardised, durations are short, and the delivery routes don't transfer cleanly to one another.

There is no established effective dose. Human trials have used roughly 0.5 to 5.5 millimoles of H₂ per day across very different methods and durations. Any brand quoting a specific therapeutic dose is quoting itself, not the literature.

The trials are short. Most run four to twelve weeks; the longest sit around twenty-four. That's enough to detect marker changes and enough to be reassuring on safety. It isn't enough to tell you what daily use over a decade does, in either direction.

Routes don't transfer. An inhalation result in a hospital population doesn't establish what a bottle does for a healthy adult, and an intravenous result establishes less still. We've graded each route on its own evidence for exactly this reason.

None of this makes hydrogen a bad bet. It makes it an early one — with a clean safety profile, a real and growing evidence base, and a low cost of entry. That's a genuinely reasonable place for a curious person to land. It's a terrible place to stand if you're promising outcomes, which is why we don't.