Free Vitamin D with Purchase of $30+, Learn More Here >>

Home / Blog


Why H₂ Concentration Matters

Contributor Bio

Alex Tarnava is the CEO of Drink HRW, and the primary inventor of the open-cup hydrogen tablets. Alex runs the clinical outreach program for our company, working with over a dozen universities coordinating research. Alex has also published research of his own. You can find it on his ResearchGate. Additionally, he has been interviewed for many prominent publications, such as Entrepreneur and Forbes, and on many popular Podcasts. You can find all of his interviews and articles on his media page.

Why H₂ Concentration Matters

As hydrogen water popularity is spreading, more and more companies are cropping up claiming to deliver “therapeutic hydrogen,” with associated benefits. Most of these companies have not supported any research, and the dosage as well as concentration provided is often below or hovering around the minimum observed therapeutic threshold. Many of these products do not accurately, or even at all, state the hydrogen concentration and dosage consumers will ingest. Some of these products contain or deliver no hydrogen gas due to packaging or processes that break down quickly. As discussed in my article “Comparing H2 Technologies”, this fraudulent marketing led to the Japanese consumer affairs department to put out a statement that 17 of 19 tested hydrogen water products on the Japanese market contained either no dissolved hydrogen or were below the minimum therapeutic threshold.

Also concerning is that many proponents of technologies that deliver low concentrations of dissolved hydrogen will accurately, but misleadingly, claim the majority of research has been done using technologies producing around 1 ppm and as low as 0.5 ppm hydrogen water. What is misleading in this assertion is that the majority of research has been conducted in mice. Hydrogen therapy alters cell signalling when cellular concentrations of hydrogen are sufficiently raised in intermittent exposures. Mice tend to consume over 10x more water as that given to humans in the majority of the clinical trials. We can come to this rough estimation by considering a 25 g mouse will consume roughly 4 mL of water p/day. The average person in North America weighs just over 80 kg. Therefore, the average person is 3200x the size of a mouse. However, the recommended water consumption for the average person is only 3.2L per day (2.7L for women, 3.7L for men). The mouse equivalent when correcting for weight would be 12.8 L, meaning mice drink 4x as much as humans. When considering that human trials have not used ad libitum dosing of hydrogen water, meaning as much as desired, but a set amount- typically between 500 mL and 1 L, the numbers skew even more dramatically. A mouse receiving the same concentration of H2 as a human ad libitum will receive roughly 12.8x the therapeutic dosage as a human instructed to drink 1 L p/day.

A further potential variable is that it typically takes larger dosages of molecules to work in mice than in humans, with a standard conversion being roughly 12x more needed in mice. That said, since molecular hydrogen does not seem to work in the same pharmacological sense as most molecules, with cellular concentrations likely being key, this conversion may be irrelevant or inaccurate. Further research is needed to address this issue. By ignoring this and converting to a 1:1 ratio, the dosages are skewed much higher towards what mice have received than humans.

Let’s Look at Some Examples:

Non-Alcoholic Fatty Liver Disease (NAFLD)

In this study on a mouse model of non-alcoholic fatty liver disease (NAFLD), 0.8 ppm was effective where 0.3 ppm was not. This study was done after an initial study using electrolyzed alkaline water with a high pH, -495 ORP, and a concentration of 0.2ppm was found to be ineffective. Many marketers would have you believe that 0.8 ppm of hydrogen water will be effective for humans in this model, but no current evidence supports this assertion. For instance, the mouse model would suggest that a human drinking 1 L of water p/day, as is typically studied in hydrogen water clinical research, would gain no therapeutic benefit at 3.84 mg of H2; or 3.84 ppm hydrogen water at a liter consumption. The observed benefits were only observed at comparative dosages of 10.24 mg of hydrogen dissolved in water for a human.

In the pilot study on NAFLD using our tablets, a strong benefit was observed in a randomized double-blind placebo-controlled crossover design study. The dosage? While the water was tested at 6 mg/L with a 1L dosage, the methods to conduct gas chromatography are quite prone to losses, particularly with quasi-dissolved nano-bubbles. Our own data suggests the tablets were around 10 ppm, or a 10 mg dosage, and recent gas chromatography reports we have had conducted by a third party are much closer to this number. With this data in mind, if we are to extrapolate from the animal research, virtually no other technology is able to deliver even the levels we know to not be effective in mice in 1 L (3.84 mg when adjusted for humans), and only our technology is capable of delivering the comparative concentration found to be effective. Furthermore, most technologies cannot even reach this “shown not to be effective” threshold when dosed at 3.2 L/day, a 1.2 ppm dose of hydrogen water for all daily water consumption. This has not stopped many company reps from pushing technologies, often below even 0.5 ppm, claiming the benefits of this study on NAFLD.

Mild Cognitive Impairment and Alzheimer’s

Early animal research on rodent Alzheimer’s models have used a hydrogen-rich saline, which has at this time, less well-established dosing protocols in hydrogen research, and no known comparative summaries of dosages required to reach similar increases in cellular hydrogen compared to inhalation or consumption of hydrogen water. One publication simultaneously studied the effects in mice with those in humans, although quite frustratingly, the concentrations and dosages administered were dramatically different; with the mice potentially receiving a far higher concentration of hydrogen dissolved in the water.

The human group saw modest results, in only those with an APOE4 allele. They were given 500 mL of water with 1.2 ppm dissolved to consume a day, and an average consumption estimated at just 300 mL p/participant. Going out on a leap and assuming that the 300 mL was consumed immediately before hydrogen dissipation lowered dissolved levels, the average participant consumed just 0.36 mg of hydrogen p/day. This is below the 0.5 mg with a max volume of 1 L p/day that the International Hydrogen Standards Association has declared the minimum observed therapeutic dose.

The mice, who experienced significant improvement in cognitive function and memory, decreased oxidative stress in the brain, DNA damage, recovery from neurodegeneration, and a significant increase in mean lifespan (but not maximal), began drinking hydrogen water at either 1month old, continuing until 18months old, or starting at a later stage of 8 months and continuing to 18 months, to simulate a mid-age protocol. The comparative dosage? The article suggests the mice were consuming super-saturated hydrogen water prepared at 0.4 mpa or 58 psi! If allowed to reach an equilibrium this would translate to roughly 6 ppm hydrogen water, given ad libitum, or a comparative dosage of 76.8 mg of H2 compared to 0.36 mg: A difference of 213x. Upon a closer look at another study using the same methods, the reported gas concentrations were at a minimum 1.2 ppm upwards of 1.6 ppm. With infrequent replacement of hydrogen water just twice a day (methods not specified), the scaled dosage with half-life would still be around 2.52 — 3.36 mg in comparative conversion, or over 6.75 — 9x the comparative dosage the human participants received.

Due to the “stronger” results from a human study using inhalation on Alzheimer’s, many peddlers of inhalation units declare gas to be more effective for neurodegenerative models than water consumption. Ignoring that the model in question has confounding variables, namely 1) A more advanced disease state and 2) Lithium carbonate was also administered alongside hydrogen, which in itself may have potential benefits in protection against Alzheimer’s (although the body of evidence does not support this yet, with mixed results). The data clearly demonstrates that lithium on its own had no effect, while hydrogen inhalation on its own did. However, hydrogen has shown to potentially be capable of lowering other drug dosages in the past. Further, there was no washout period, and the study was more in line with observational case series. Further complicating the matter was that the relative dosage was much higher with a smaller sample population (n=11 compared to n=73).

The inhalation study cites a concentration of 3% hydrogen gas for two one-hour sessions p/day, while not specifying a flow rate, but referencing another study using the same device. This study, likewise, does not specify a flow rate, but does state that it is delivered through a non-re-breathing facemask. The average adult breathes in 5-8 L p/minute at rest, so at 6.5 L p/minute, 3% hydrogen for 120 minutes p/day, participants were inhaling a total of 23.4 L of hydrogen gas. Reducing by a factor of 100* to consider impact between method, that is the equivalent of 234 mL of hydrogen dissolved in water a day, or 16.47 mg of H2 in a direct comparison to water- a comparative dose of 45.74x more than what was used in the mild cognitive impairment study.

*the only published comparison shows that H
2 when dissolved in water has similar and perhaps even more prominent effect than inhaled gas, at 1/100th the dose. This conversion is limited by a lack of data and a dearth of direct comparative studies and analysis. The exact conversion is likely different for different targets with comparative doses being affected by concentrations.

The Takeaway

We still need a lot more research to know if hydrogen works for the models discussed above (and many others) and what dose is effective. Direct comparisons between animal and human studies need to account for relative dosing, as do comparisons between various administration methods. For consumers not willing to wait until better trends emerge, the data so far suggests that higher dosages and concentrations come with greater benefits. When assessing animal models, conversions need to be considered. When considering claims from manufacturers and salesmen of inhalation units, conversions to effectiveness need to be considered.




  • Alex Tarnava

    Hi Charlie, if you are seeing great benefits in 1-2 per day no need to increase. I recommend dropping in up to 500ml (16oz) of room temp water, and gulping it down as fast as possible as soon as the tablet rises to the surface. If that is too much water to gulp down, try making as big of a gulp as you can immediately, and then quickly drink the remainder. Otherwise, try lowering to 12oz/335ml of water if you can gulp that down in one go. I hope this helps and glad you enjoy!

  • Charlie O'Shea

    I have been taking hrw for sometime. Usually 1-2 tablets per day. Should I be increasing the tablets? Is there a recommended water quantity that I should drink per day? Most of my water is filtered thru a Berkey which I find the best for the way the tablets dissolve. Thanks Charlie

  • Alex Tarnava

    hello, the brand used in the Japanese study is not available here, as far as I am aware. That said, the study seems to indicate a much higher dose, as given to the mice, would be far more effective. We have an article I wrote here comparing various technologies:

  • Daniel Aparicio

    Excellent article. I am looking for brand specific testing data completed on the Japanese study as I purchased a Buder portable hydrogen water generator and I need comparison data that shows which models met the “therapeutic” threshold in human studies. Thank you for this excellent article!

Leave a comment

Please note, comments must be approved before they are published