Molecular hydrogen trials

The research record

This is the project's file for this landmark, as written and checked. Sources are listed at the end; anything not yet confirmed is marked "Not yet verified".

Who

Asako Yoritaka, Nobutaka Hattori and colleagues at Juntendo University and 14 Japanese hospitals; and Carine Maurer at Stony Brook University, New York.

What happened

Molecular hydrogen (H2) is a gas that, in cell and animal studies, mops up the most reactive oxygen molecules. Oxidative stress, damage from such molecules, is thought to hurt dopamine neurons; landmark 129 found the brain's main antioxidant, glutathione, depleted in the nigra.

• Pilot (2013). In Movement Disorders, 17 levodopa-treated patients drank 1,000 mL a day of hydrogen water or plain water for 48 weeks. Total UPDRS (a disease rating scale) improved by 5.7 points in the hydrogen group (n = 9) and worsened by 4.1 in the placebo group (n = 8), a difference the authors called significant (P < 0.05). The authors later acknowledged that a pilot of this size cannot tell symptom relief from slowing.
• Multicentre trial (published 2018). Same design in 178 patients at 14 hospitals (154 on levodopa, 24 not), 1,000 mL a day for 72 weeks, main measure change in total UPDRS. The Japanese grant report states no significant difference (p > 0.05); the water was safe but did not help. The Movement Disorders paper (2018;33:1505) is a short report; I could not read its abstract.
• Stony Brook (NCT03971617). Phase 2/3, hydrogen-producing tablets with 80 mg magnesium versus matching placebo, planned 70 early-stage patients for 52 weeks, primary outcome safety. It ran May 2019 to April 2021, was terminated for "insufficient enrollment" and had 2 participants (1 per arm), neither with adverse events.

Why it moved us toward a cure

It is a clean example of a small positive pilot being refuted at scale, in line with the antioxidant failures in the same branch (for example landmark 66, urate). It also shows a mechanism (reactive oxygen) can be real in cells while a hydrogen drink has no effect in people.

Outcome

Dead end. No further hydrogen trial for disease modification found.

Sources

• Yoritaka et al., Mov Disord 2013;28:836, DOI 10.1002/mds.25375
• Yoritaka et al., Mov Disord 2018;33:1505, DOI 10.1002/mds.27472
• Protocol, BMC Neurol 2016, DOI 10.1186/s12883-016-0589-0
• KAKEN final research report (source for the null 178-person result)
• ClinicalTrials.gov NCT03971617 (API, 19 Oct 2026)
• Not yet verified: the 2018 paper's own numbers (its abstract was unavailable); whether the multicentre trial was formally named a Phase 2; the 2013 page ranges (volume/page taken from the DOI record only).