Introduction

Before secondary plant metabolites were gathered together with the vitamins and some of the minerals and classified as antioxidants, they in particular, had other less therapeutic roles in the plant kingdom.
The coloured, scented and aromatic dietary flavanoids are used as attractants for insect, bird and animal polination and seed dispersal in plants. The bitter proanthocyanins and the tannins in general are important in maintaining the balance between the amount of foliage browsed by animals and the fertility of the surrounding soil. Their bitterness limited the animal browsing activity, preventing serious defoliation while not repelling the animals totally, and by so doing, depriving the plant of natural fertiliser. This symbiosis is maintained by secondary plant metabolites.
At the time that these notions were being debated, the tannins, known to bind to proteins, were thought of as useless antinutritional by-products of biosynthesis that were disposed of by chemically employing them to build the lignin skeleton forming the structural elements of the plant.
No doubt there are many more functions performed by so-called antioxidants that continue unaffected by their newer role of plant, animal and human antioxidants.

Alternative theories

A. Variations on antioxidant mechanisms.
If we keep an open mind about whether or not food-based natural antioxidants are themselves active therapeutic agents in the process of reducing oxidative stress in humans, then what are the alternative mechanisms that need to be considered?
1. They may not work alone. The known in vitro antioxidants may require synergistic combinations, e.g. vitamin C and lycopene.
2. They may be preliminary signalling compounds that simply trigger secondary processes such as the activation of cellular antioxidant enzymes near or at the target site.
3. While some of them have been shown to raise the antioxidant capacity of the blood plasma soon after ingestion, relatively little evidence exists to link them directly to the in vivo reduction of oxidative stress and the subsequent therapy which epidemiological trials would seem to be suggesting for them.

B. Occlusion of noxious compounds. Recent alternative processes have been recognised for food-based natural compounds which are not associated with the neutralisation of free radicals, and therefore are thought of as being potentially therapeutic through an alternative mechanism - occlusion of noxious compounds. An example reported this year is the role of a modified pectin molecule, a complex polymeric galactan, in removing from circulation the protein galectin-3, associated with the occurrence of cancer. However, diets rich in pectin are also rich in vitamin and flavanoid antioxidants. Hence the mechanistic dilemma: Do flavanoid antioxidants neutralise free radicals, reduce oxidative stress, and reduce the spread of cancer, or do modified pectins bind to and inhibit proteins such as galectin-3 with the same end effect? A bit of both, no doubt.
Prof. Vic Morris, the senior author (A.P. Gunning et al., FASEB, doi: 10.1096/fj.08-106617.) advises us to hedge our bets by continuing with the five-a-day fruit and vegetable, high fibre, high pectin, high antioxidant food regime.

[It would be interesting to examine the behaviour of proanthocyanins as potential anticancer agents, by determining which proteins are preferentially occluded by them. Whatever the outcome, it is certain that a balance must be struck between their therapeutic and antinutritional roles.]

C. Hormesis Hypothesis of Caloric Restriction (HHCR)
Hormesis - which defines a substance having a beneficial effect at low doses even though it inhibits performance in medium doses and is toxic at high doses - is not a new concept. In the 16th century, Paracelsus said, “All things are poison and nothing is without poison, only the dose permits something not to be poisonous”. The work of Schulz (H. Schulz, Pflugers Arch., 42, 517-541, 1888) and Arndt was combined to create the Arndt-Schulz rule which briefly states that "low doses stimulate while high doses inhibit" was sidelined by mainstream scientific opinion because of Arndt's involvement in homeopathy. Unfortunately, the concentrations used in homeopathy are orders of magnitude lower, but the principle is the same. Although the concept of dose related response was the province of toxicology, and the linear non-threshold model was generally adopted for carcinogens, the notion of a "J" shaped hormetic response curve is of interest in both toxicology and nutritional biochemistry.
Caloric restriction is thought to improve health and slow the aging process by putting the organism into a defensive state of marginally increased oxidative stress in response to food-based energy restriction, thus conditioning the organism against further insult. For an explanation, the work on a worm (Caenorhabditis elegans) by Michael Ristow's group describes HHCR as the low intensity induction of oxidative stress which systematically increases resistance to further oxidative stress, thus improving body fitness, health and lifespan.

D. Foreign body hypothesis Prof. Balz Frei, Director of Linus Pauling Institute
Facts:
1.Flavonoids are strong in vitro antioxidants.
2. Flavonoids are poorly absorbed (<5%).
3. What is absorbed is quickly converted for excretion.
4. The high TAC values observed after eating high flavonoid meals is thought to be increased levels of uric acid in the blood.
Hypothesis:
That the very small amount of flavonoid absorbed is "toxic", causing a "vaccine-like" response which induces a larger immune response e.g. phase II enzymes that act to eliminate carcinogens which would otherwise cause cancer for example.

Deleterious effects of so-called therapeutic antioxidants

Bearing in mind the above provenance for the polyphenols, things only get worse when the facts of human nutrition are added, such as the fact that the non-essential polyphenol proanthocyanin (tannin) nutrients occlude essential minerals, vitamins and other important nutrients, rendering them inaccessible for digestion/absorption. So, why are these astringent polyphenols so popular as food-based antioxidants? Truth is we do not know, exactly. A recent study of certain condensed tannin-containing food extracts found a positive relationship between the sensation of astringency index (SAI), the human salivary protein precipitation capacity (HSP) and their total antioxidant activity (TAA).
Reference: R. Amarowicz et al., J. Food Lipids. 15 , 28-41, 2008.
These observations set the scene for a possible new area for debate. Is it acceptable to tolerate the negative effect of astringent antioxidants actually occluding useful, possibly non-astringent antioxidants e.g. vitamins, selenium?

Biochemical footnote:

It may be essential to differentiate between the anthocyanins and proanthocyanins, the polymeric complex polysaccharides of the anthocyanidins and proanthocyanidins which are the monomeric building blocks which are less likely to occlude essential nutrients and more likely to scavenge free radicals.