2009-09-20

TEA CUTS RISK OF DIABETES

From the Daily Express (aka "Tits By Christmas" - though this epithet was coined several Christmases ago and now, I suppose, has to be regarded as undeserved).

This all began when I tweeted on this headline. I had not, at that stage, actually read the article. I had simply walked past a news-stand with my son Max in tow. "Is that really true?" he asked as our eyes fell on the banner headline.

Now I don't think I've ever actually explicitly said to Max "Never believe anything you read in the Daily Express - especially if it's to do with science" (nor would I entirely go along with this bald statement); but I have obviously, unconsciously transmitted this sentiment to him over the years.

I remarked on the fact that Max seemed to be developing a healthy scepticism towards journalism in general and scientific journalism in particular in one of my tweets. Since then several people have kindly re-tweeted my original tweet, but the original context has been lost and I thought it was time I looked at the text below the headline - before the Express lawyers see one of my cropped tweets in a new context and begin libel proceedings against me.

The first thing to note is that the article concerns Type 2 diabetes (which is typically - though not exclusively - a problem for older people and is much more strongly associated with lifestyle factors than Type 1 diabetes). Although Type 2 is far more common that Type 1, it is Type 1 - and its associated treatment regime of insulin injections - which pops into many people's minds when the term "diabetes" is used.

Even if the research reported in the article is "kosher", there is no suggestion here that forcing your kids to drink tea will help protect from developing "insulin dependent diabetes" (as it used to be called) or that drinking tea will help your kids if they already have this condition. The same considerations apply to adults with Type 1 of course, but they can make their own decisions. In either case, I think the article - though it does say that the findings apply to Type 2 diabetes - ought (to be on the safe side) to say explicitly that the findings do not apply to Type 1.

To move on the research itself, there is (and this is, unfortunately, the rule rather than the exception in science articles in newspapers) no link the the original paper(s) or conference proceedings on which this article is based. Looking at what is reported, this was a "study of more than 40,000 people whose ­consumption was monitored for 10 years". This sounds impressive, but there is no indication of how the 40,000 people were recruited. Further down we read "They got participants to fill in a daily food questionnaire".

There are several potential problems here. Were the original recruits randomly chosen? Haw many of those approached agreed to take part? How many dropped out? How reliable were their recollections of what they had consumed? None of these considerations should be regarded as damning, but (depending on whether and, if so, how they have been addressed - something we are not told) we need to make (and above all report) inferences from research like this with caution.

Another more fundamental problem with research programmes of this kind concerns the direction of the causal relationship. If, as the reported research seems to suggest, there is a correlation between tea drinking and failure to develop Type 2 diabetes, this may be because drinking tea provides protection, or it may be that Type 2 diabetics (or those going on to develop the condition) drink fewer cups of tea. Such issues can, of course, often be settled by further research. But we are not told whether or how the second explanation was excluded. Again, this consideration does not necessarily damn the research in any way, but it does mean that we have to be circumspect in our conclusions.

A few lines further down comes an even more worrying bit: "Dr Carrie Ruxton, of Britain’s Tea Advisory Panel, said". Forget what he said (which was about black versus green tea), why is his opinion being reported here at all? Did the "Tea Advisory Panel" commission or fund this research? Did they bring it to the attention of the Daily Express? No suggestion of impropriety even if they did, but I think we should be told.

The article ends with some perfectly good stuff:


A spokeswoman for campaign group Diabetes UK gave the research a cautious welcome, but stressed there was much more that people could do to prevent developing type 2 diabetes.

She said: “This is interesting research, however it does not prove that coffee and tea protect against type 2 diabetes...It is impossible to know what other factors might affect a person’s risk of developing the condition.

“The best way to prevent it remains keeping active and eating a healthy, balanced diet that is low in fat, salt and sugar with plenty of fruit and vegetables.”
Earlier this week it emerged that drugs to treat type 2 diabetes caused by obesity were costing the NHS £600million a year, the biggest drug bill it faces.


And you can see why "The best way to prevent Type 2 diabetes remains keeping active and eating a healthy, balanced diet that is low in fat, salt and sugar with plenty of fruit and vegetables." would not make a great headline; but "TEA MIGHT HELP CUT RISK OF TYPE 2 DIABETES" is reasonably snappy and would actually reflect the information cited.

I think, on balance, the last word on Daily Express journalism should go to John Cooper Clarke.

1 comment:

  1. Hi,

    7.9% of the United States population is suffering from a form of Diabetes. That is over 23 million people! Now, more than ever, it is important for organizations such as yourself. We here, at Disease.com (a site dedicated towards disease and their treatments), believe in the work you do and would like to coincide for the fight against diabetes. If you could, please list us as a resource or host our social book mark button, it would be much appreciated. Separately, we can make advancements, but together we can find a cure.
    If you need more information please email me back with the subject line as your URL.

    Thank You,
    Sharon Vegoe
    Disease.com

    ReplyDelete

Comments are moderated, but you can leave them without registering.