Go Figure: Life expectancy and the risk of dying

Three lifestyle factors - television
Image caption Stories about lifestyles and life expectancy are rife

We've been told "death risk" can be cut by 15 minutes exercise a day, but just how should we view life expectancy data?

"Ah yes, Mrs Badcrumble.

"Rich, non-smoker, eats her veg, watches three hours of TV daily, does 15 minutes on the treadmill, and, oh dear yes, wears a green hat.

"Despite the hat, I'm pleased to tell you that you've a 98.2% chance of not being dead until 86."

Not quite realistic, not yet. But reports this week about a link between watching TV and dying early show how researchers are slowly matching lifestyle with death rates to work out the odds on your future.

These findings come in dribs and drabs. This week it was TV and exercise. What if we could put all the disparate research together?

Will they, one day, produce a personalised when-might-I-die calculation? In a limited way they already have. As with a dating site, you can - sort of - match your own profile to discover when your time might be up.

Image caption It's possible to make a life expectancy curve

Here's an illustrative example from the Understanding Uncertainty website where you enter a few of your significant behaviours - smoking, drinking, physical activity and diet - and see how they affect your chances of survival to any given age.

On the website, click on the tab labelled "behaviours". This survival curve - as it's called - shows all the people who've survived to age 45 and the steadily falling percentage who make it further. The fewer healthy boxes you tick, the steeper the curve. Try it.

So how do they do it? The basis of the calculation is what are known as UK interim life tables. These tell us, among other things, the average chance of dying during the next year depending how old you are now, based on what actually happened in the UK between 2006 and 2009. This is the starting point, a basic death rate.

Next, they multiply this average risk by the extra risk from drinking too much, exercising too little and so on. The size of this extra risk comes from a vast international study of half a million people known as EPIC, which counts the extra bodies at each age associated with each unhealthy behaviour.

Put the two together and you have a customisable survival curve.

Incidentally, the difference between most unhealthy behaviours identified by EPIC and the healthiest is about 14 years of life.

Of course, there are caveats. First, there are a lot of averages in here and no average is fully personalised. You might not even own a TV, be trim and active, and still have a heart attack at 40. This is only about behaviour and lifestyle. We've not mentioned biological factors like genetic susceptibility to disease, and there might be other things not yet spotted or understood.

It all depends on people's ability to recall reasonably well what they do with their time.

Next, the link between watching TV and shorter life might not be a result of sitting in front of a TV, but of inactivity generally. Although the researchers say they took exercise into account.

Then, when we are told that 15 minutes of daily exercise can lengthen life, we are not really told how hard. Quality matters as well as quantity. I once interviewed the UK's top 400m runner who described a training session of 5 x 200 metres at just below race pace with short rest intervals that sometimes left him throwing up at the track side. That's only about two minutes of exercise.

Finally, all this talk of reducing or raising the chance of dying - as some media coverage has lately - can't help reduce the risk below, well, one. Not in the end, not yet.

But someone soon is going to work out how to put it all together - the diet and the TV and all the rest, combined with factors like height and weight - in a user-friendly way that allows people to personalise the data and see what happens to the risks if they change their behaviour.

The next question is - who really wants to know?

Related Internet links

The BBC is not responsible for the content of external Internet sites