Monday, 3 June 2013

"Microlives": I'm pretty certain this is utter tripe...

****

...from the Guardian (where else) with some people's wonderful new 'live forever' theory:

The Norm Chronicles, a new book by journalist Michael Blastland and Professor David Spiegelhalter that has a neat idea which turns all these abstract dangers into a concrete figure.

It centres on this idea: once you hit adulthood (or, being more precise, 22 for a man and 26 for a woman) you can expect to live for around 500,000 more hours – or a million half-hours. Each of those 30 minutes of life is a "microlife".

By working out the average effect of, say, smoking or eating red meat, we can figure out a cost in microlives for different habits. A portion of red meat, for example, costs you a microlife – in the words of Blastland, it's "a 30-minute chip off your stock of adult life".

This seems to me to be ignorance squared - taking averages (I guess 'norms') and using them as a predictor of individual life expectancy is not either good maths or good science. Maybe that's not what the book says - the good professor is, after all, a statistician. But it is what the Guardian says the book says - essentially that we can quantify the effect on our individual health of actions where the effect is based on estimates of how much the action adds or subtracts from our lifespan.

The problem is - and if we thought about it for a second we'd know this - that the estimates are open to question. We really haven't much of a clue about the impact of eating red meat on life expectancy even if we do have a general (if challenged) idea that a diet of red meat isn't ever so healthy. For sure, where there's a known dose-response effect (e.g. with smoking - note the word is smoking not tobacco - or alcohol) there's perhaps a bit of a case. But for things such as exercise there is little evidence that getting sweaty on the treadmill extends life - the the adding of microlives on the rowing machine is probably nonsense.

What we have here is extending the general to the specific (from the whole population average to little old me or you) combined with evidence that, to put it mildly, is open to question and perhaps not epidemiologically sound.

But I guess that the gullible Guardianistas are looking for a 'Spirit Level' for personal health and these authors have delivered! However, such tripe is best served with onions and accompanied by a good claret.

....


5 comments:

Anonymous said...

It is rather disappointing that Michael Blastland and David Speigelhalter have come out with this. Because you can only calculate attributed risk if you know ALL risks and to all intents and purposes there are an infinite number that could be found.

And they should be well aware of that!

Tony

Leg-iron said...

It is indeed arrant nonsense.

Some years ago, there were idiots making money by pushing the 'no mistakes ever' consultancy scam.

They would turn up and give a talk at enormous expense to gullible companies. The gist of it was -

If you can go a minute without making a mistake, you can go two minutes without making a mistake... and so on.

It was based on exactly the same principle. So this 'microlives' is just a rehash of an old scam.

johnpopham said...

I think you are probably right Simon.

Whatever, I maintain that stories are much more powerful in changing people's behaviour than statistics

michael blastland said...

You could always read it, and find out if it says any of the things that you assume that it says - whether the authors are entirely ignorant of the strengths of weaknesses of an average, whether they think everyone should strive to live forever... etc.

Anonymous said...

Michael,

I have always held David Speigelhalter and yourself in high regard so I hope Simon and I are wrong. I was commenting on the basis of the short extract in the Guardian and you're right, perhaps I should hold fire until I've read the book in full. But the extract did seem somewhat dubious to me.

Tony