Animal & lab study
2007
Antioxidant mechanism
A landmark 2007 paper in Nature Medicine proposed that molecular hydrogen (H₂) can selectively neutralise the most damaging reactive oxygen species inside cells, while leaving the useful ones alone.
What they found In cell cultures and a rat stroke model, hydrogen lowered markers of oxidative brain injury.
Our honest take This is laboratory and animal work — not a study of people drinking Kangen water — and the exact mechanism is still debated. It's the spark, not the proof.
Ohsawa I, et al. Nature Medicine, 2007. PMID 17486089.
Literature overview
2023
The evidence base
Interest in "hydrogen medicine" has grown quickly. One 2023 overview counted more than 1,000 scientific publications on molecular hydrogen — 1,126 indexed on PubMed as of mid-2021.
What they found The literature spans antioxidant, anti-inflammatory and metabolic effects across many laboratory and clinical models.
Our honest take A large volume of papers is not the same as proof. Most of this work is still early-stage, and large, long human trials remain rare.
Li S, et al. Medical Gas Research, 2023. PMID 37077121.
Small human pilot · n=10
2012
Exercise & recovery
A pilot study gave ten male footballers hydrogen-rich water or placebo before a bout of intense exercise, then measured blood lactate and muscle function.
What they found The hydrogen-water sessions showed lower peak blood lactate and a smaller drop in muscle function than placebo.
Our honest take Just ten athletes across single sessions. It's an interesting signal that deserves a bigger trial — nothing more, yet.
Aoki K, et al. Medical Gas Research, 2012. PMID 22520831.
Randomised trial
2008
Metabolic markers
In a randomised, placebo-controlled crossover trial, people with type-2 diabetes or impaired glucose tolerance drank hydrogen-rich water for eight weeks.
What they found Several blood-lipid and oxidative-stress markers improved — but the key long-term blood-sugar marker, HbA1c, did not change significantly.
Our honest take We show the mixed results on purpose. This is precisely why no one can honestly promise a disease benefit.
Kajiyama S, et al. Nutrition Research, 2008. PMID 19083400.
Human questionnaire study
2018
Digestive comfort
Alkaline electrolysed water is recognised in Japan only for easing mild gastrointestinal symptoms. A 2018 study followed self-reported digestive comfort in everyday drinkers.
What they found Participants reported a subjective improvement in mild gut symptoms over the study period.
Our honest take Self-rated symptoms are a weak form of evidence, and this is a narrow, specific use — never a claim to treat disease.
Tanaka Y, et al. Medical Gas Research, 2018. PMID 30713669.
Meta-analysis · 76 studies
2024
Antioxidant capacity
A 2024 meta-analysis combined the results of 76 studies of hydrogen-rich water to look for consistent effects.
What they found Pooled data suggested H₂ water can raise the body's measured antioxidant capacity — but did not clearly reduce oxidative-damage markers or exercise lactate.
Our honest take A better lab number is not a proven health benefit. The authors themselves call for larger, higher-quality trials.
Li X, et al. Frontiers in Nutrition, 2024. PMID 38590828.
Randomised controlled trial
2018
Neurological disease
A double-blind, randomised, placebo-controlled trial tested hydrogen water against placebo in people with Parkinson's disease.
What they found The trial did not meet its primary endpoint — hydrogen water did not significantly outperform placebo.
Our honest take We include the failures too. Hydrogen water is not a treatment for Parkinson's, or for any other disease.
Yoritaka A, et al. Movement Disorders, 2018. PMID 30207619.
Peer-reviewed review
2024
The bottom line
A 2024 review in the International Journal of Molecular Sciences surveyed the human evidence for hydrogen-rich water across a range of health outcomes.
What they found The authors summed up the overall picture as "encouraging" — but explicitly "preliminary."
Our honest take That is our position, word for word: genuinely interesting science worth following, not a medical product, and no promises.
Dhillon G, et al. Int. J. Molecular Sciences, 2024. PMID 38256045.