The evidence,
in the order medicine ranks it.
70 peer-reviewed studies on molecular hydrogen, organized by the strength of their scientific design. Not every study carries the same weight. A meta-analysis pooling thousands of patients is not the same as an isolated case report — but both matter, in different parts of the reasoning.
01Why the levels matter
Evidence-based medicine doesn’t treat all studies as equal. There is a hierarchy: the more controlled the design and the larger the number of patients, the more confidence you can place in the result.
The directory respects that hierarchy. It starts at the top — meta-analyses that synthesize dozens of prior trials — and descends gradually to individual case reports: valuable clinical anecdotes, but not generalizable.
— Stronger evidence above · Greater specificity below —
No one should decide a treatment based on a meta-analysis alone, nor dismiss it just because the evidence is a case report. The levels tell you how much weight the finding carries, not whether it’s worth reading.
02The 4 levels
-
L1 6 studiesMeta-analyses · PRISMA systematic reviews
The ceiling of evidence
Studies that synthesize dozens of prior trials into a single statistical number. It’s the closest answer to "what does the science say, as a whole?". These are the findings that best resist time and individual bias.
-
L2 54 studiesRandomized controlled trials (RCT)
Applied evidence
Randomized, double-blind, placebo-controlled studies. The heart of the directory: 54 trials where H₂ was compared against placebo in humans. This is where you see what hydrogen does in real conditions: diabetes, hypertension, athletic recovery, Long-COVID, autoimmune disease.
-
L3 8 studiesObservational studies · Safety
Real-world data, without randomization
Cohorts followed over time, pilot studies, accumulated safety data. Less control than an RCT, but closer to everyday clinical use. They are the basis on which we know H₂ is safe at the doses studied.
-
L4 2 studiesExtraordinary case reports
Frontier medicine
Individual patients, exceptional situations, responses that don’t fit the average. A case proves nothing on its own — but it opens questions that RCTs haven’t yet answered. Read them as hypotheses, not conclusions.
03How to read each level
If you want certainty: start with Level 1. What a meta-analysis shows has already passed the filter of dozens of independent studies.
If you want concrete clinical application: Level 2 is where the directory lives. 54 RCTs with doses, durations, populations and measurable outcomes. This is where you find "what happens when a person with X takes Y for Z?".
If you want safety: Level 3 is your reference. Before accepting an intervention, medicine demands knowing it does no harm — and observational studies accumulate years of real-world data.
If you want to understand the limits: Level 4. Extraordinary case reports show what H₂ can do in specific circumstances, and what we still don’t know at the population scale.