Live fast, die young, leave some pretty research

There is a serious message at the heart of the study by Mark Bellis of the Centre for Public Health at Liverpool John Moores University. This is the report which seems to claim being a pop star is bad for your health, and upon the discoveries Bellis hangs a plea for everyone to take more care of our pop stars:

"Pop stars can suffer high levels of stress in environments where alcohol and drugs are widely available, leading to health-damaging risk behaviour.

"Their behaviour can also influence would-be stars and devoted fans.

"Collaboration between health and music industries should focus on improving both pop star health and their image as role models. Public health consideration needs to be given to preventing music icons promoting health-damaging behaviour among their emulators and fans."

Which, you know, it's hard to argue against when Pete Doherty's due in court and William Hill have been taking bets on whether or not Winehouse will turn up for the Mercury Awards.

But we're not sure that either the methodology of the study, nor its conclusion, are right.

The contention that rock and pop stars are more likely to die than non-pop stars is based on

1,064 US and European rock musicians and singers who featured in an all-time top 1,000 albums chart compiled in 2000.

The conclusion from looking at this subset is that:

Overall, 100 pop stars included in the study had died, with chronic drug or alcohol-related problems or overdose accounting for a quarter of deaths. The average age of death was 35 for European stars and 42 for those from the US.

Between three and 25 years after becoming famous - defined as appearing in the charts - pop stars were found to be 1.7 times more likely to die. Twenty years after first achieving chart success, 10 per cent of US stars and four per cent of those in Europe were dead - twice the normal rate.

But are the people who made albums in an 'all-time best-of' chart a fair cross-section of rock and pop stars? Isn't there an extent to which the best-known stars, the one whose records are remembered later on, who continue to sell, are the ones whose early death makes them notorious?

In other words, how can Bellis be sure these people are dead because they're on the chart, and not on the chart because they're dead?

Then there's the question of if Kurt Cobain and Ian Curtis, and the other rock and roll suicides, would not have killed themselves had they not been in the music industry? To paraphrase Nick Hornby, are they sad because they're rock stars, or are they rock stars because they're sad? Could 'being a musician' be a symptom of a tendency to suicide, rather than a cause?

And is it fair to compare 'rock stars' with the general population? If you're going to blame the fame and success of selling records as being a cause of an early death, would it not be more appropriate to compare the death rates with other jobs where young people suddenly become well-known and/or incredibly rich in a profession high on dangerous air-travel in private jets and long periods of not having much to do?

Not to mention how, the "twice as likely to die" for European rock stars seems to be quite a large claim based on a small sample and a comparison between 2 per cent and 4 per cent - how much of that would disappear in a margin of error?

Still, it's always nice to see the social sciences being treated seriously.