Reactive oxygen species act as signals before they ever act as damage

A theory that was too tidy

For a long period the dominant account of ageing held that oxygen-derived molecules accumulate damage until function eventually fails. The theory was attractive because it connected metabolism, damage and ageing in a single causal chain that was easy to state and easy to test. Testing it produced results that did not cooperate, with several animal experiments failing to show the expected relationship between defences and lifespan.

In some cases increasing antioxidant defences produced no lifespan effect, and in others mild oxidative stress was associated with the opposite of harm. The theory was not abandoned so much as substantially narrowed, and what replaced it is a more complicated account involving signalling.

Oxidants as regulated messengers

One particular oxidant species is comparatively stable, diffuses across membranes and modifies specific amino acids reversibly, which are the properties a messenger needs. Those modifications alter the activity of target proteins in a way that can be undone by dedicated reducing enzymes, giving the system an off switch. Cells possess enzymes whose function is to produce this species deliberately in response to hormones and to mechanical stimulation.

A molecule the cell manufactures on purpose, in a controlled location, is difficult to describe as purely a by-product of damage. The current picture is of a signalling system that becomes damaging when its regulation fails rather than of an intrinsic poison.

Location and concentration do the discriminating

Production confined to a small region near a target protein can modify that protein without affecting the rest of the cell measurably. The same total quantity released indiscriminately would react with whatever it encountered first, which is chemistry rather than signalling. Cells maintain steep local gradients using the enzymes that consume these species, so the effective concentration varies enormously across a single cell.

This is why measurements of whole-cell or whole-blood oxidative markers correlate poorly with what is happening at any particular signalling site. It is also why the concept of total oxidative status, offered by some testing services, has limited interpretive value.

Why broad antioxidant intervention disappoints

If some oxidative signalling is required for normal adaptation, then suppressing it broadly should interfere with that adaptation. Work in exercising humans has examined exactly this, with several trials suggesting that high-dose antioxidant intake blunts some training adaptations. The findings are not uniform and the trials are small, so this is an area where the direction of evidence is suggestive rather than settled.

Large trials of antioxidant compounds for disease prevention have generally failed to show the benefits that observational data implied. In some cases those trials found signals of harm, which is a serious result and one reason blanket supplementation is not recommended in guidelines.

What remains solid

Oxidative damage to molecules is genuinely measurable and genuinely occurs, and nothing in the revised account denies that. What changed is the assumption that less of these species is always better, which turned out not to follow from the chemistry. The useful framing is regulation rather than quantity, with health corresponding to a system that controls production and removal appropriately.

That framing is harder to sell than a number on a supplement label, which is part of why the older account persists in marketing. Anyone considering high-dose antioxidant intake, particularly alongside medical treatment, should raise it with a clinician because interactions are documented.