Observations Media watch

Behold the Christmas miracle of antioxidants

BMJ 2007; 335 doi: https://doi.org/10.1136/bmj.39413.403750.59 (Published 29 November 2007) Cite this as: BMJ 2007;335:1124
  1. Ben Goldacre, doctor and writer, London
  1. ben{at}badscience.net

    Apparently chocolate is good for you and red wine is better. But should we believe it?

    I'd like to make a sage prediction, seeing as it's early December. One of the joys of watching bad science coverage in the media—as I have done for four years now—is that you start to spot patterns: and this year, just like every Christmas, as regular as mince pies, I can confidently predict a specific rash of stories: they will explain solicitously that chocolate is good for you—“actually”—and red wine is even better.

    It's not much of a prediction, since in the world of public relations, Christmas has started already. “Choxi+” is milk chocolate with “extra antioxidants,” and the newspapers are fawning over it already: “too good to be true,” says the Daily Mirror; “chocolate that is good for you, as well as seductive,” says the DailyTelegraph. The company is said to “recommend” two pieces of its chocolate a day. “Guilt free,” says the Daily Mail: it's “the chocolate bar that's ‘healthier' than 5lb of apples.” Meanwhile, Sainsbury's is promoting Red Heart wine—with extra antioxidants—as if drinking the stuff was a duty to your grandchildren.

    These products represent triumphs of over-extrapolation from observational data, and laboratory hunches. A huge amount has been made of the j shaped curve in the relationship between alcohol or wine consumption and good health. Moderate drinkers, the media gushingly tell us, are less likely to be obese than both teetotallers and heavy drinkers. Nobody wants to ruin Christmas with dreary talk of confounding variables, pointing out that teetotallers themselves are an unusual breed: recovering alcoholics, the chronically ill, devout religious abstainers from varying ethnic groups, and of course people who lie in surveys, among others.

    Moderate red wine drinkers, we are specifically informed, come out better on all kinds of health measures, and nobody wants to ruin Christmas by mentioning confounding variables again (like how moderate red wine drinkers hang out at home with their friends eating salad and talking about their posh jobs and stable social support). A fairytale science story must be simple, reductionist, and mechanistic. Red wine is good for you because it contains lifegiving molecules, like antioxidants.

    And nobody wants to spoil Christmas—for the whole family—by mentioning that the antioxidants story is one of the great unspoken non-starters of 20th century medical research. People who eat fresh fruit and vegetables have lots of positive health outcomes, including reduced rates of heart disease and cancer; and fresh fruit and vegetables contain lots of antioxidants; and people who have high levels of antioxidants in their blood seem to be healthier.

    There's even a charming fable from the metabolic flow charts in biochemistry textbooks about what antioxidants do in the body. Sainsbury's loves that story, along with the others: “Exposure to UV rays, pollution and smoking produces free radicals,” it says. “Free radicals are compounds that cause cell damage, which in the long term can damage health.” It's such a gloriously simple tale of right and wrong you can almost picture it, in animated form, on ITV after the Queen's speech. “Antioxidants help counteract the harmful effects of free radicals. Red Heart has an antioxidant level which is 32% higher than the average level of other leading red wines.”

    Only a malevolent Scrooge-like figure, mumbling over his glass of tap water in the corner, would dare to point out that if you are going to pore over a biochemistry textbook, and pick pathways out at random, then you can prove anything you like. Phagocytic cells build a wall around invading pathogens and then use free radicals—among other things—to kill the bacteria off. Should we be selling free radical supplements to help people fight infections?

    The antioxidant story took a bit of a blow, of course, when people started to do placebo controlled randomised trials with antioxidant vitamin supplements, to see what happened: because overall they seem to do nothing, or at worst, reduce life expectancy. And that's when you might start to think, well now, perhaps people who eat fresh fruit and vegetables are, just like the people who drink red wine in decorous moderation, living healthily in all kinds of ways. Much like the people who buy vitamin pills. Lusty walks around country mansions. Cycling to work. That kind of thing.

    Of course there may yet be something valuable in the antioxidant story, although it's probably not going to be as simple as dishing them out by the spoonful. And of course observational studies aren't inherently evil or useless: they're frequently fascinating, as part of a puzzle. These are all interesting theoretical research findings, as we try to puzzle out the roots of cancer and heart disease.

    But they make a pretty thin excuse for flogging chocolate and alcohol. And somewhere out there—right now—a researcher is rubbing their hands with glee, poring over a press release, picturing themselves in the Today programme studios, planning some choice quotes for the Daily Telegraph: something racy about mince pies cutting heart disease because of the raisins, perhaps, or red wine helping you run faster. Well, it's Christmas. Have another.

    Nobody wants to spoil Christmas—for the whole family—by mentioning that the antioxidants story is one of the great unspoken non-starters of 20th century medical research

    View Abstract

    Sign in

    Log in through your institution

    Free trial

    Register for a free trial to thebmj.com to receive unlimited access to all content on thebmj.com for 14 days.
    Sign up for a free trial