Ben Goldacre
The Guardian,
Saturday July 26 2008
The newspapers are so profoundly overrun with pseudoscience about food that there’s no point in documenting it any longer. They will continue with their Sisyphean task of dividing all the inanimate objects in the world into the ones that either cause or cure cancer, and I will sit at the sidelines, making that joke over and over again.
This week, however, the Telegraph, which has lost its science editor and its science correspondent in two months, deserves special attention, because two of its food stories went beyond stupid, and managed to give actively harmful information.
We’ll start off with “Red wine could help prevent breast cancer“. In fact it’s even better than that. “Laboratory tests have shown that a chemical found in the skin of grapes could halt the development of most cases of the disease.”
The story follows a standard template which they clearly now teach as valid in all journalism schools: a food contains a chemical, the chemical does something in a dish on a lab bench, therefore the food kills cancer in people. Or rather, red wine contains resveratrol: this chemical has been found to increase the activity of an enzyme called quinone reductase, which converts a derivative of oestrogen back to oestrogen, and that derivative can damage DNA, and damaging DNA causes mutations, and mutations cause cancer, so therefore, in the world of journalists, red wine prevents breast cancer in people.
This is a phenomena we might call “data mist”: where someone gets one piece of research information lodged in their imagination and suddenly, for them, it explains the entirety of medicine.
Back in the real world there is the wine, and a whole human body. Red wine has a lot of stuff in it, including alcohol which, notably, is broken down into a chemical called acetaldehyde, and that itself causes DNA damage. Is there reason to think this might cause cancer, or would I be over-extrapolating myself?
Well. A huge number of observational studies have now been performed in real living humans, finding that people who drink more have more breast cancer. They have included careful analyses, in which they try to account perhaps imperfectly for other possible explanations for this relationship. These studies have even been collected together in a systematic review, and a meta-analysis, where all the figures are run onto one big spreadsheet, and they estimate that overall, half a glass of red wine a day increases your risk of breast cancer by 10%. If their figures are correct, alcohol causes about 6% of all breast cancer in the UK, meaning 2,500 cases a year.
So you don’t muck about with breast cancer, and red wine, despite whatever it says on that press release you are holding in your hand, Kate Devlin, medical correspondent, does not prevent breast cancer.
Meanwhile they had “nutritional therapist Julia Fitzgerald” cheerfully informing Telegraph readers that “during the first four weeks of pregnancy taking a folic acid supplement (400 micrograms per day) can help prevent neural tube defects in the foetus such as spina bifida“. You will remember that “nutritional therapist” is a term we can all legally use to describe ourselves, alongside “dilettante” “handsome” and “clown”. Perhaps Fitzgerald’s training in the field of “nutritional therapy” might have caused some confusion here, since folic acid for pregnant mothers is a rarity in their domain, being an intervention with an evidence base.
Sadly for Fitzgerald the evidence is not in her favour. And as all of the random women in their 30s I have asked today already knew, both Department of Health and NHS guidelines are identical, recommending very clearly that potential mothers should take a 400-microgram supplement from the time that they start trying for a baby, or from when contraception is stopped, up to and including the 12th week of pregnancy.
Perhaps Fitzgerald knows better. Perhaps she can also tell you the day you get fertilised. Perhaps this is the first time a nutritional therapist has ever recommended too few pills. And perhaps she will want to clarify her advice to 800,000 readers and their families which will increase their risk of delivering a baby with severe and disabling neural tube defects.
I will now go back to ignoring the nutritionists.
· Please send your bad science to ben@badscience.net
References:
The very astute will note that going to pubmed and typing the phrase alcohol breast cancer meta-analysis will immediately give you this reference:
Meta-analysis of studies of alcohol and breast cancer with consideration of the methodological issues. Cancer Causes Control. 2006 Aug;17(6):759-70. Key J et al.
I think it’s worth noting that the exact same medical correspondent, Kate Devlin, bafflingly wrote about this exact issue only three months ago. You would think two and two might fall together, but I’m told things are a bit rough in the Telegraph at the moment.
www.telegraph.co.uk/news/uknews/1585095/A-little-alcohol-%27increases-breast-cancer-risk%27.html
Before we get distracted, Cancer Research UK do a very good page called “How do we know?” which would always be, as you can well imagine, my preferred subtitle for any page describing a bunch of risks.
info.cancerresearchuk.org/healthyliving/alcohol/howdoweknow/#Breast
And it’s also worth mentioning that the Cancer Research UK blog also covered this issue. I didn’t copy them, and made a point of not reading them first, but they say almost exactly the same thing, only longer, and arguably better.
scienceblog.cancerresearchuk.org/2008/07/08/no-red-wine-doesn%E2%80%99t-prevent-breast-cancer/
I think they could turn out to be quite a good thing, and it’s great to see this kind of organisation standing up to nonsense for a change.
CDavis said,
July 19, 2008 at 11:57 am
Oh, well – if it’s the alcohol that’s a problem, clearly it’s grape juice that cures cancer, silly.
Stands to reason…
CD
evidencebasedeating said,
July 19, 2008 at 8:50 pm
UK RD agrees.
With such a neatly rationalised argument you appear to have silenced your critics…
used to be jdc said,
July 19, 2008 at 9:54 pm
Sisyphean task? I’ve been calling it the oncological/ontological project (I think I nicked that from one of the commenters on a blog post you wrote about the Daily Mail though).
Incidentally – having read the print edition first, I think the pictures you’ve used here help to drive your point home. I hope that someone from the Telegraph see this post and it helps them to understand why it is important not to send out the wrong message to their readers on health matters.
Robert Carnegie said,
July 20, 2008 at 11:07 am
Sentenced eternally to wander the newspapers, fruitlessly mocking nutriwoo? Who did that to you then? Gillianus, goddess of scatomancy? The ghosts of John and Will Kellogg?
I think it was not Sisyphus of course but Tantalus whom I’ve seen depicted tormented by grapes, the torment being that he wasn’t allowed to eat them or to drink from the water he was standing in. Wikipedia doesn’t specify the fruit, but does remind us that a “tantalus” is a drinks decanter that locks, a bit like the mini-bar in your hotel room, so you can’t get at it (or, in the case of the mini-bar, you can’t afford it).
PO8 said,
July 21, 2008 at 1:23 am
I know you’re a physician and thus a bit desensitized, but were the graphic medical photos really necessary? I’m all for them in their proper place, but here they seemed to me to be a bit of a sideshow…
johnsmith said,
July 21, 2008 at 11:46 am
One problem here is in the interpretation of these type of headlines, with phrases like “could help”,”may help” or “can help.”
“Red wine could help prevent breast cancer” is not really incorrect, as its only suggesting it as a possibilty, and surely we can’t rule it out as a possibilty. But it is read as “Red wine will help prevent breast cancer.” There is a huge difference between the two headlines that many fail to see.
As important as it is for journalists to send out the right message, its just as important for the readers to realise what they are actually reading.
Tom Bird said,
July 21, 2008 at 2:03 pm
Another neat article, but would it be possible in future to have a version of the article without the photos of tumors and horribly deformed babies or at least some type of warning? Maybe you could hide the images and behind a button that says “Click here to see a cancerous breast”?
I understand that you’re trying to make a rather graphic point about how bad these illnesses are and the consequences of disinformation and all, but still, the images actually make me feel sick to look at. I almost abandoned reading the article as soon as I saw them…
Ben Goldacre said,
July 21, 2008 at 3:33 pm
interesting, fair enough. you’re absolutely right about the reasoning, they were absolutely there to make the point that in medicine, you’re not just talking about data, you’re talking about real distress. i did think fairly carefully about the risks and benefits, and thought i’d made the right call, but fair enough, point heard. reckon i should take em off?
seems a bit wrong to change a blog post, and poss better to have a “yeah maybe you’re right”.
pv said,
July 21, 2008 at 3:53 pm
What’s the saying? A picture paints a thousand words.
I think the photos are fine. They reinforce the point appropriately and, more to the point, perhaps people should forced to look at that kind of thing every now and again. I reckon if people are complaining about the images then they’ve served their purpose well. The subject matter isn’t interior decoration or haut cuisine. Life is not Disney. It’s unfair sometimes – brutal and undiscriminating – and we should occasionally be reminded.
T said,
July 21, 2008 at 4:27 pm
Interested to know if people generally agree that flour should be fortified with folic acid in this country as it is in America?
Re the red wine, my sister is an alcholic and has switched to a bottle of red wine (sometime a bottle and a half)rather than a bottle of white a night, as she believes its better for her…..people who drink too much will find an excuse, reason, whatever to continue drinking. They dont listen to health advice as they are addicts and are in denial.
johnsmith said,
July 21, 2008 at 4:41 pm
slight defence of journalists;
“so therefore, in the world of journalists, red wine prevents breast cancer in people.”
– no, the journalists suggest red wine COULD prevent breast cancer in people.
Having read the article a bit more closely, it seems that it is guilty of regularly dropping “coulds” and “mays.” This seems to be the case in many articles.
Surely claiming that something “could be” or “may be” a cure for cancer is different from claiming that it IS a cure. But you attack the journalists as if they are making actual claims of certainty, when they are only suggesting possibilities. It’s just that they are not read as mere possibilities.
If I see a headline such as “red wine COULD help prevent cancer,” then I just ignore it as it essentially meaningless. Saying something “could” happen is allowing for the possibility that it “could not” happen. It’s just like saying: “We don’t know whether red wine will prevent cancer.”
used to be jdc said,
July 21, 2008 at 4:58 pm
Re johnsmith @ comment 6:
–
The Key et al meta analysis linked to in the refs section above notes that “each additional 10g ethanol/day was associated with risk higher by 10%”, “the association between alcohol and breast cancer may be causal” and that “risk did not differ significantly by beverage type”. I read this as all types of alcohol increase breast cancer, so it seems to me that ‘red wine may prevent breast cancer’ should be ruled out as a possibility for all practical purposes. I don’t see how it is possible for all types of alcohol (including red wine) to increase breast cancer, but for red wine to help prevent it.
Re POB08 @#5 – the pics didn’t seem like a sideshow to me. As I wrote in my first comment, I thought they drove the point home and I hope the Telegraph journalist has seen this blog and those pics. I’m with PV on this one – a picture paints a thousand words. I don’t thing anything else would have been as effective.
johnsmith again @11 –
“the journalists suggest red wine COULD prevent breast cancer in people.”
I would suggest, given the meta analysis results, that red wine can’t prevent breast cancer in people. Therefore the journalists are wrong – they are wrong (in fact, dangerously wrong) to even suggest it as a possibility. The fact that Kate Devlin herself wrote about the meta analysis in the Telegraph three months ago only makes it worse.
bumpkin said,
July 21, 2008 at 10:29 pm
It’s “a phenomenon” not “a phenomena”. Otherwise, good.
mrstrellis said,
July 22, 2008 at 12:12 am
I’m glad you included the photos. I got bored with taking folic acid supplements during interminable infertility treatments and gave up, but I’m back on them now.
johnsmith said,
July 22, 2008 at 10:20 am
used to be jdc, yes, of course the journalists are now proved wrong by the analysis.
Yes, of course it is very dangerous for journalists to suggest these possibilities. But the readers are at fault aswell as the journalists.
If after reading such a story you decide that you should drink more red wine as it will help prevent cancer, then there is a mistake by you, the reader, aswell as by the journalist for suggesting it as a possibility.
Whatever the evidence and analysis shows, if articles such as this just leave out the “coulds” and the “mays” whenever it suits them, then they are twisting the journalists’ claims.
outeast said,
July 22, 2008 at 10:42 am
I hated the pictures, on the level of ‘I just don’t want to look at that’, but their inclusion is undoubtedly effective: this is probably the angriest BS column I’ve read, and the pictures hammer home why.
Marcus said,
July 22, 2008 at 11:04 am
I think all the “could” and “may” is fine and necessary. The actual problem is a lack of context.
The reasons for this has to do with underqualified journalists. I think naming and shaming is certainly a step in the right direction.
But that will only take us so far. The underlying problem has to do with how the media functions as a business – i.e. things like staff cuts, high turn-around times, low salaries, a readership that tens to expect sensation.
Pro-reason said,
July 22, 2008 at 11:18 am
Well done, Bumpkin. I can’t believe I read the article without noticing “a phenomena”. It’s one of my pet peeves!
Good article, although the baby images seemed gratuitous.
mikewhit said,
July 22, 2008 at 11:52 am
Print article – good.
Online article – please use a hint from The Register and annotate item header on main page with “[NSFW]” (not suitable for work) – due to pix which actually IMHO will distract people from reading the text properly.
Delster said,
July 22, 2008 at 12:16 pm
@ Marcus – The could and may are inexcusable.
Something has to be taken into account when protesting that the media only said “could” or “may”.
This is that many people don’t have the reading skills in terms of analysis of the terminology used within the story.
A lot of people i know would read a story written that way and come out with the conclusion that the “insert item here” is helpful for preventing “insert condition here”.
I would say from reading the comments, over the last year or two, that a lot of the people who read this column are either above average in intelligence or have a better than average education in the sciences. This leads to better skills in critical analysis of what is written in the media… or to put it another way we’re a bunch of pedants at heart.
This doesn’t excuse the media from writing in this style or from the sloppy research into articals.
The media is the way that many people get their information on science developments etc. The media therefore have a responsibility to get it right and write about it in a clear, understandable fashion for the general public. Think lowest common denominator.
Re the pictures. Ben was right to include them. Many people (including myself) will not be familiar with neural tube defects. The pictures illustrate exactly what’s being talked about and exactly the kind of damage that can be caused. This is a science column folks….science can get messy and ugly!
perspix said,
July 22, 2008 at 5:31 pm
Re photos. They make one feel icky and that’s exactly as it should be imo. They are real and pertinent/informative, not gratuitous/entertaining.
Agree with above comments about Journos needing to be more responsible. The public do not read newspaper articles in the same way that academics read journals. They do not parse each sentance accurately but rather get impressions.
“Red wine may be helpful in curing some Cancers” becomes “Cancer-Wine-Cure” and the Journalists and Editors of newspapers and comics know this very well.
used to be jdc said,
July 22, 2008 at 6:20 pm
Sorry if I’m going on, but…
johnsmith, I think it is human nature to justify the choices we make [in fact, Oliver Burkeman had a column related to this recently*] and I think if we see an article in the press that discusses the possibility that our favourite drink, or food, may be beneficial there is a fair chance we will use that article to justify our choice of beverage or meal. This does not in any way excuse the writer of the article if they are not only wrong but *should* have a pretty good idea themselves that they are wrong [particularly if they themselves have written about the wrongness of their idea a couple of months previously]. If the casual reader knows no better and assumes that an article in their favourite daily newspaper will accurately reflect the science then how are they to blame if the article is, in fact, completely inaccurate? Do you not think in this case the journalist should take responsibility for what they have written? This is not a case of readers assuming that ‘red wine could prevent cancer’ means ‘red wine will prevent cancer’ – it is a case of a journalist misleading their readership by writing a story they should damn well know is wrong, because they wrote about a study that showed their story to be wrong.
The journalist in this case has written about a compound found in red wine that may have some effect on cancer prevention. The story is based on a lab bench study that looked at a compound that is found in red wine. Not only was this not a study looking at the effects of red wine in living, breathing humans it was not even a study that looked at the effects of red wine – so how did the headline come about? I think the Cancer Research UK blog that Ben linked to covers the study and the report better than I could, but basically the journalist has extrapolated from a lab bench study looking into the effects of a compound (that is one of many found in red wine) on cancer cells and written a headline that relates to the effects red wine ‘could have’ on humans. Except it couldn’t – and we already know it doesn’t. Even the journalist who wrote the article knows red wine does not – and could not prevent breast cancer… because a few months ago she wrote a story on the meta analysis that shows red wine causes more cases of breast cancer.
*Burkeman on Cialdini [re: emotional investment in decisions etc]
Cheers,
jdc.
Delster said,
July 22, 2008 at 6:51 pm
If a new study or paper comes out and a journo wants to do an artical on it but doesn’t have the background to do so why don’t they contact a group or person who knows, somebody like a cancer charity for instance?
I’m sure these groups would rather see articals intelligently written and passing along the correct info than the made up rubbish we currently get.
daikonsensei said,
July 23, 2008 at 2:51 am
Tobacco has big leaves, and is dead easy to grow. Also there’s decades of detailed research on Tobacco Mosaic Virus [wonder why the BBC used quotation marks “tobacco mosaic virus” instead of tobacco mosaic virus. That’s really the virus name! and also it really does infect. so why “infect” instead of infect.] So if you want to make proteins in plants using viruses, and in relatively large amounts, tobacco would be a good choice.
Chapeau! to whoever wrote this one for not starting a scare story about genetically modified viruses..
Judging by my experience in searching for well paid jobs in plant science, Big Tobacco does its research in switzerland (presumably to avoid eu regulation) and no doubt in the USA (but i didn’t see any posts advertised). The work seemed to be focussed on breeding, aiming to produce tobaccos they could market as low in particular compounds. “healthier” tobacco.I suspect they’re actually not trying all that hard and these days just busy aggressively targeting emerging markets with conventional tobacco instead..
johnsmith said,
July 23, 2008 at 12:37 pm
jdc – in response to a few of your points:
“If the casual reader knows no better and assumes that an article in their favourite daily newspaper will accurately reflect the science then how are they to blame if the article is, in fact, completely inaccurate?”
Isn’t it partly the responsibility of the casual reader to make sure that they know better? Should they expect the newspaper to accurately reflect the science? I think that if the reader makes a potentially harmful decision based on reading a news story then there is some reponsibility in both camps(and in many other camps aswell).
“Do you not think in this case the journalist should take responsibility for what they have written?” – Yes, but obviously what they have written is interpreted in different ways by different people. If it is interpreted wrongly, then is there not a wider responsibility to make sure more people make better judgements about what they are reading?
“This is not a case of readers assuming that ‘red wine could prevent cancer’ means ‘red wine will prevent cancer’ – it is a case of a journalist misleading their readership by writing a story they should damn well know is wrong, because they wrote about a study that showed their story to be wrong.”
Its both – poor journalism and misinterpretation of what is actually being claimed.
Journalism is not about truth and accuracy, maybe it should be, but the fact that is isn’t isn’t the just the fault of the journalists. This BS article is close to implying that bad journalism can cause cancer – this would be spectacularly unfair – what causes bad journalism? why do they get it wrong? Surely there are many, many reasons. We can all find much horror and amusement from journalist errors every week, but does it really get to the heart of the problem?
I guess this could go on, maybe we’ll have to agree to disagree on some points.
used to be jdc said,
July 23, 2008 at 3:55 pm
I think you’ve made some interesting points johnsmith – and I at least partially agree with some (e.g., that journalism is not about truth and accuracy – though I wish it were).
Apart from anything else, I think you’re right about agreeing to disagree. I’d be happy to continue the discussion, but I’m not sure how thrilling it would be for other readers (a commenter on my blog has today used the phrases “particularly ramble-some” and “unnecessarily wordy” in describing my most recent post and I don’t disagree).
Cheers,
jdc.
pv said,
July 24, 2008 at 12:08 am
johnsmith wrote:
If I’ve understood this correctly it looks remarkably like the “just following orders” line of defence. And I disagree with it. Journalists are responsible for what they write, even if what was written has been sub-edited out of all recognition.
Newspapers set themselves up as arbiters of truth and morals, therefore no special knowledge should be necessary for the public to decode any news journalist’s output. The fact that special knowledge is necessary is not the fault of the public but entirely the fault of journalists who choose to lie and choose to work for organisations whose policy is dishonesty.
Robert Carnegie said,
July 24, 2008 at 1:46 am
Hmm, one moment in the [Dispatches] which I just got around to watching about HPV vaccines was where The Guardian got the press release about the vaccine from a cancer charity – a majority (?) of mothers of daughters would choose to have them vaccinated – but the survey was paid for by the vaccine maker company. So maybe you can’t trust the charities either.
sean.salvador said,
July 24, 2008 at 10:12 am
This just in ha ha. The beb website is now talking about tobacco curing cancer. Now i didnt check the reasearch myself so i have no idea whether ther is any truth in it but i found it really funny anyway. Here’s the link.
news.bbc.co.uk/1/hi/health/7517799.stm
sean.salvador said,
July 24, 2008 at 10:12 am
make that beeb
misterjohn said,
July 24, 2008 at 1:13 pm
More good stuff in The Guardian. There may be hope for us after all.
www.guardian.co.uk/lifeandstyle/2008/jul/24/healthandwellbeing.radovankaradzic
Woo and genocide, an interesting combination.
coatgal said,
July 25, 2008 at 12:48 am
I don’t think I had ever seen a picture of a baby with spina bifida before. And the photos definitely hit home… I now get why that dose of folic acid is so important, without such a vivid image, it is just another pill you (well I) have to take. Which makes it easy to forget or to become a bit casual about it. I guess a bit like the MMR. Is their an argument for using imagery to push home the point the MMR won’t hurt your baby but M, M and R all can do quite nasty things? Or is this already being done? I did see a cartoon image of a kid with bright red spots in my surgery today with the tag line ‘Measles is Back!’ But it looked a bit Bash Street Kids and didn’t really stir a sense of urgency…
Vicky said,
July 25, 2008 at 3:08 pm
johnsmith said:
“If I see a headline such as “red wine COULD help prevent cancer,” then I just ignore it as it essentially meaningless.”
I consider myself fairly good at filtering out nonsense health advice but after years of being pelted with the same misinformation from magazines, adverts, colleagues, wild-eyed soothsayers on buses etc. it starts to break down the barriers.
Now, if I’m deciding between red or white wine, there’s an almost subconscious leaning towards the ‘healthier’ option, even though I know that’s bollocks. I’m sure I can’t be the only one whose decisions are being influenced like this.
used to be jdc said,
July 25, 2008 at 4:39 pm
Re Vicky’s comment #38 (“I’m sure I can’t be the only one whose decisions are being influenced like this”)… I think you are right Vicky – you are not alone in being influenced by newspaper articles.
From the Bad Science blog, 21st July:
“People read newspapers. Despite everything we think we know, their contents seep in, we believe them to be true, and we act upon them.”
[http://www.badscience.net/index.php?s=Ishango]
bob_calder said,
July 26, 2008 at 11:27 am
The image of the canerous lesion reminds me of what one of my friends described seeing on trips to small towns treating patients who seldom had normal medical treatment.
A home remedy (what we would call CAM today) was to stuff the opening with herbs to facilitate healing. He told me it didn’t work well.
Another treatment based on pull-it-out-of-your-ass witch doctor/home remedy-ville that caused problems was women douching with a dilute solution of water and Lestoil.
maryn said,
July 29, 2008 at 6:30 pm
I’m one of those non-readers emilypk referred to (#39) – I was unable to read beyond the first photo. Probably because I’m a squeamish humanities graduate 😉
Enjoyed the comments, though.
projektleiterin said,
July 31, 2008 at 11:28 am
I agree with Tom Bird, I’d rather have links to the pictures. I blocked them with a Firefox Add-on (Adblock Plus), because when reading the post I found them to be too distracting.
Kat Arney said,
July 31, 2008 at 2:06 pm
mikewhit said,
July 29, 2008 at 5:17 pm
>So who’s going to do a clinical trial on chemicals in red wine – or stuff like turmeric and broccoli for that matter ?
Funny you should mention that Mike – Cancer Research UK are funding a few studies – including clinical trials – of some of these “superfood” molecules, including resveratrol (red wine), di-indolylmethane (broccoli) and curcumin (turmeric). I’ve written a bit more about our work in this area on our blog here:
scienceblog.cancerresearchuk.org/2008/07/21/broccoli-%e2%80%93-the-latest-cancer-fighting-%e2%80%9csuperfood%e2%80%9d/
Enjoy!
Kat
Delster said,
August 2, 2008 at 12:51 pm
I think some people would actually read the artical because of the pictures where they would normally just read a few lines. It works both ways.
Doctor Spurt said,
August 4, 2008 at 10:37 am
I add my voice to the chorus saying that the pictures are salutory. They’re not easy to look at, but the stakes here are real people and real harm. Too often people respond to attacks on nutriwoo peddlers by claiming that they’re not so harmful.
wulftheo said,
September 10, 2008 at 2:43 am
Julia Fitzgerald seems to have issued a clarification:
No amendment to the original article. I suppose people who supplement iron, calcium etc. have the sort of backbone to maintain that this is a clarification rather than the rectification of an error.