Extrapolation in extremis, or, How the Daily Mail went all Homer Simpson and claimed that 'purple is a fruit.'

A big thank you to @Evidencematters on Twitter for pointing out this story...

'Remember to eat your purples: Fruit can 'ward off Alzheimer's, heart problems and cancer,' or so the People's Medical Journal (aka the Daily FMail) tells us.

So, I won't bother writing the rest of this post - instead I'll just head off to buy shares in growers of purple fruits...

Oh, wait...

The Mail story is an object lesson in how to take seemingly sound science and extrapolate its findings to the logical (or illogical) extreme; a picture-perfect example of the art of insinuation, conjecture and wishful thinking. Oh, and it's also an addition to the honourable Oncological Ontological Project (the mammoth classification of which has been completed by Paul Battley) whereby, according to Bad Science maestro Ben Goldacre, 

The Daily Mail, as you know, is engaged in a philosophical project of mythic proportions: for many years now it has diligently been sifting through all the inanimate objects in the world, soberly dividing them into the ones which either cause – or cure – cancer

This time, it's purple fruits which may 'cure' cancer, and help 'ward off' diseases such as multiple sclerosis and Parkinson's - and according to Mail reporter Fiona Macrae it's all there in black and white, written by a scientist no less. No ordinary scientist either - Prof, Kell is a prolific scientist interested in the biochemistry of metabolism, and serves as the Chief Executive of the Biotechnology and Biological Sciences Research Council. The article cites a paper by Prof. Kell from the University of Manchester, and says that the polyphenols contained within purple fruits (and green tea, curcumin and chocolate, although sadly not the Mail's favourite anti-cancer compound, red wine) may help

fight the harmful effects of iron, which can damage cells if it makes its way through the digestive system in the wrong form.

We've been here before on this blog, with a Mail story that linked perfumes to male infertility - again based in some way on studies by Professor Richard Sharpe but that bear no resemblance to the science contained therein.

So I did a little digging - I emphasise the little, as it doesn't take too long to dig for the absurdities in Ms. Macrae's article.

Prof. Kell recently published a review article in the journal Archives of Toxicology, to which I assume (given the Mail's insistence on not linking to original sources...) Ms. Macrae's article refers. If you're interested you can read the full article here, or just the abstract via PubMed here.

Prof. Kell draws on extensive evidence (his review is one of the most thorough and heavily-referenced I've ever seen with over 1,700 references...) regarding the role of iron - specifically, iron that is 'poorly liganded' - in the progression of many seemingly disparate diseases. Poorly liganded iron? Qu'est-ce que c'est?

Iron, as Mrs. Teekblog the chemistry teacher had to remind me, is a transition metal that can exist in several valencies - in other words, it can choose to make bonds with any number of ligands (other atoms/ions/molecules) up to six. Iron is said to be fully liganded when it's bound to its full complement of six ligands - according to Prof. Kell's review, iron within cells that is bound to fewer than six ligands can cause the production of highly reactive hydroxyl radicals (free radicals) through the Fenton reaction, which are well known to cause damage to DNA and proteins and play a significant role in cell death during disease.

Prof. Kell goes on to take a 'systems biology' approach to the role of poorly-liganded iron in disease, postulating that the build-up of the incompletely-bound form of iron may be the 'nexus' that links diseases with apparently unrelated causative factors with seemingly disparate outcomes - that poorly-liganded iron and the free radicals it creates may be the gatekeepers channel various insults to cells - whether genetic or environmental - to causing cell death by myriad mechanisms.

As ever, pictures say much more than words ever could, and this elegant diagram may help you in understanding Prof. Kell's view of the importance of iron's valency in disease:



The review article meticulously trawls through data regarding poorly-liganded iron and its potential contribution various disease processes - and yes, at one point Prof. Kell does indeed mention that

the only way to stop the damaging activity of free or partially liganded ‘iron’ is to ensure that all of its six possible liganding sites are satisfied, whether by endogenous chelators or those added from the diet or as pharmaceuticals

Prof. Kell then lists a sizeable chunk of literature that points to the positive effect that many polyphenols (rings of carbon, hydrogen and oxygen of varying properties found in plants) may have on the chelation of iron - largely these studies are in cells in culture or in animal models that recreate aspects of human disease, using purified dosage of the polyphenol under consideration. The theory is that dietary intake of such compounds may boost the body's anti-oxidant capacity, negating the creation of excess free radicals by poorly-liganded iron - but it remains just that, a theory, and there's already evidence that anti-oxidant supplementation is of little or no benefit when taken in the forms cited in the Mail.

At no stage that I could tell does Prof. Kell mention studies looking at the effect that purple fruits, or green tea, or curcumin (all of which do indeed contain polyphenols) have on the pathology of any disease that may have poorly liganded iron as a contributory factor - presumably because no such studies exist.

To go from Prof. Kell's position to the suggestion in the Mail that purple fruit may be protective is, err, somewhat ambitious. What's clear from the studies Prof. Kell reviews is that when iron is poorly liganded it creates high levels of free radicals, which may be a link between lots of different diseases - but that's about it. I'd say that the evidence to date regarding the use of polyphenols as iron chelators suggests it might be worthwhile conducting prospective trials to determine the effect of fruit-derived anti-oxidant on disease progression - but then how's a sub-editor ever going to boil that down to a catchy populist headline?