Superfood products

“Superfood” is a marketing term with no scientific or regulatory definition. That is not a criticism of the foods so labelled — many are genuinely nutrient-dense — but it does mean the word carries no information, and several jurisdictions restrict its use on packaging for exactly that reason unless an authorised health claim supports it.

What is worth examining is the mechanism the category was built on, because it changed.

Nutrient density is the part that is straightforwardly true

Some foods deliver a high quantity of essential nutrients per unit of energy. Leafy greens, liver, oily fish, legumes, seeds and berries score well on this by any measure, and nutrient density is calculable from composition tables rather than asserted. Where a food is promoted for its vitamin, mineral, fibre or long-chain omega-3 content, the claim is checkable and often correct.

The difficulty begins where the promoted property is not a nutrient.

The antioxidant story, and why it was withdrawn

Through the 1990s and 2000s the category was organised around antioxidant capacity, measured by assays such as ORAC, which quantify how well a food extract quenches free radicals in a test tube. Foods were ranked on it and marketed on it.

The metric was abandoned. The United States Department of Agriculture withdrew its ORAC database in 2012, stating that the values had been misused to imply health benefits and that the assay does not reflect what antioxidants do in the body. That is not a fringe position; it is the position of the body that produced the numbers.

The reasons are mechanistic. An in-vitro radical-scavenging assay measures direct chemical reactivity in a cuvette, at concentrations far above anything achievable in plasma. Dietary polyphenols are poorly absorbed — a large fraction never leaves the gut lumen intact — and what is absorbed is rapidly conjugated by phase II enzymes into glucuronides and sulfates that are much weaker radical scavengers than the parent compound. Plasma antioxidant capacity after a polyphenol-rich meal rises mostly because of urate and other endogenous molecules, not the polyphenols themselves.

There is also a conceptual problem: reactive oxygen species are signalling molecules, not merely damage. Large-scale trials of high-dose antioxidant supplements have generally failed to show benefit, and some — notably β-carotene in smokers — found harm.

The model that fits better

The current explanation for why polyphenol-rich foods associate with good outcomes runs the other way. Rather than neutralising radicals directly, these compounds appear to act as mild stressors that provoke an adaptive cellular response — activating pathways such as Nrf2, which upregulate the body’s own antioxidant and detoxification enzymes. This is hormesis: a small challenge producing a protective adaptation.

It fits the evidence better on several counts. It explains why low, food-level doses could matter while high supplement doses do not help and may harm. It explains why the conjugated metabolites can still be relevant, since signalling does not require radical-scavenging potency. And it is consistent with the colonic microbiota’s role, since microbial metabolites of polyphenols are absorbed in quantity and are plausible signalling agents.

It is a model with good support, not a settled account, and specific dose–response relationships in humans are not established.

Reading a product

The defensible questions are concrete: which nutrient or compound, at what measured content per realistic portion, with what human evidence at that intake. A food can be excellent without any of the framing above — and the framing does not make a food excellent.