The scientific case for the 10% inhalation-environment safety threshold and the non-recommendation of high-concentration devices has been built incrementally over eleven years across four peer-reviewed papers.
PMID: 26504515 — Kurokawa et al, *Medical Gas Research*.
A foundational paper comparing molecular-hydrogen delivery routes (inhalation, hydrogen-rich water, oral, intravenous). It establishes that inhalation has the highest systemic uptake efficiency, and raises the inhalation-specific safety issues (concentration, device design) that subsequent work addresses.
PMID: 31552881 — Kurokawa et al, *Medical Gas Research*.
A device-design synthesis. It distinguishes the closed-system LFL of 4% from the empirical inhalation-environment ceiling and proposes a safe-operating-concentration framework.
PMID: 36204781 — Ichikawa et al, *Medical Gas Research*.
Synthesizes the Japanese Consumer Affairs Agency accident-information database into an academic account of the safety problems with high-concentration devices, and derives consumer-facing selection criteria.
PMID: 41489350 — Ichikawa et al, *International Journal of Risk & Safety in Medicine*.
Analyzes the physiological mechanism of in-airway ignition and proposes the shift to low-concentration operation (≤ 10%) as an academic recommendation. This paper is the primary support for the database's safety guidance.
These four papers carry weight beyond each individual finding: for eleven years, an academic line of inquiry has continuously raised the safety concern. Accident-database accumulation and theoretical refinement have progressed in parallel. The current operational recommendations (10% empirical ceiling, non-recommendation of high-concentration devices) rest on that continuity.
https://h2-papers.org/en/safety-notes/lineage