It was initially suggested that the beneficial effect of hydrogen was due to an antioxidant as hydrogen selectively neutralized cytotoxic hydroxyl radicals in vitro. However, although H2 reduces *OH radicals, as has been shown in various systems, it may not occur via direct scavenging, and it also cannot fully explain all the benefits of hydrogen. For example, in a double-blinded placebo controlled trial in rheumatoid arthritis, hydrogen had a residual effect that continued improving the disease symptoms for four weeks after hydrogen administration was terminated. Many cell studies also show that pre-treatment with hydrogen has marked beneficial effects even when the assault (e.g. toxin, radiation, injury, etc.) is administered long after all the hydrogen has dissipated out of the system. Additionally, the rate constants of hydrogen against the hydroxyl radical are relatively slow (4.2 x 107 M-1 sec-1) [20], and the concentration of hydrogen at the cellular level is also quite low (micromolar levels), thus making it unlikely that H2 could effectively compete with the numerous other nucleophilic targets of the cell. Lastly, if the mechanism were primarily associated with scavenging of hydroxyl radicals, then we should see a greater effect from inhalation compared to drinking water or using tablets, but this is not always the case. In short, we consider it inaccurate or at least incomplete to claim that the benefits of hydrogen are due to its acting directly as a powerful antioxidant. Indeed, hydrogen is selective because it is a very weak antioxidant and thus does not neutralize important ROS or disturb important biological signaling molecules. Nevertheless, a metabolic tracer study using deuterium gas demonstrated that, under physiological conditions, deuterium gas is oxidized, and the oxidation rate of hydrogen increases with an increasing amount of oxidative stress, but the physicochemical mechanism for this may still not be direct radical scavenging. However, not all studies show that hydrogen is oxidized via mammalian tissues, and it has also been reported that deuterium gas did not exert a therapeutic effect in the model studied whereas 1H did (unpublished data).
