Friday, 5 May 2017

When alcohol prophecies fail

 
In 2014, Ian Gilmore, chair of the Alcohol Health Alliance said that his aim was to get alcohol consumption back to where it was '30 or 40 years ago'.

'Of course, we do not want everybody to stop drinking. That is not the endgame. However, nonetheless, if we could just get the whole consumption curve shifted down by 10 or 20% - if we could get back to where we were in the UK, say, 30 to 40 years ago in our per capita consumption and we would see a huge benefit in health.'

Gilmore seemed unaware that alcohol consumption had already fallen so much by 2014 that it was already back to the levels of 30 or 40 years ago. His chums at the Institute of Alcohol Studies have this handy graph which I have annotated:





So Ian Gilmore must be pleased, right? Time to give up the crusade and move onto something more important?

Of course not. In response to the latest figures on alcohol consumption and harm published this week, he says:

'We know that over the long term, rates of binge drinking are falling, and more people are choosing to abstain from alcohol.

Worryingly, however, these trends do not appear big enough to stop alcohol harm from continuing to rise, and the sharp increase in alcohol-related hospital admissions over the last few years means hundreds of thousands more people each year are experiencing the misery associated with harmful alcohol consumption.

The data released today should be sobering reading for whoever wins the upcoming general election, and we would urge the next government to make tackling alcohol harm an immediate priority to save lives, reduce harm, and reduce the pressure on the NHS.'

It is slightly surprising that Ian Gilmore, who professes expertise on alcohol policy, did not know that alcohol consumption was at 1980s levels when he spoke out in 2014. It is less surprising to find a neo-prohibitionist pressure group demanding further action even achieving its stated goals.

Gilmore made a rookie mistake that most nanny state campaign groups are careful to avoid. He set an explicit target and promised real improvements in health if it was met. He made a further mistake by making a prediction about something that had already happened but which had not been accompanied by any notable improvements in health.

By the logic of the total consumption theory, health outcomes should have shown a marked improvement as a result of a 20 per cent reduction in alcohol consumption. Like most people in pretend 'public health', Gilmore is an advocate of this ridiculous theory, which is why he says 'if we could just get the whole consumption curve shifted down...'

It has taken a long time, but the decline in alcohol consumption is now a recognised fact. The BBC even published an article titled 'Are we falling out of love with booze?' yesterday. And yet estimates of the number of alcohol-related hospital admissions continue to rise.

There are two explanations for this. The first is that the hospital admissions estimates are bunkum. The second is that the total consumption theory is wrong. These two explanations are not mutually exclusive. In fact, I think both of them are correct.

Under the 'broad measure' of alcohol-related hospital admissions, numbers have risen by a third since 2004/05 despite a 20 per cent fall in alcohol consumption. Under the somewhat more realistic 'narrow measure', numbers have risen by 22 per cent. For those who believe in the total consumption theory, these figures are awkward, but even for those of us who are sceptical about the theory, it is difficult to take these figures seriously as a proxy for alcohol-related harm when there is so much evidence that 'binge-drinking' has declined.

The reality is that we could all stop drinking tomorrow and the number of alcohol-related hospital admissions would keep rising, not because of the 'lag effect' of chronic disease but because alcohol-related admissions are assumed to be a fixed proportion of each disease or injury. Alcohol-related admissions statistics are not compiled by people who work in hospitals. Statisticians simply take the aggregate number of admissions for each disease and assume that a certain percentage are alcohol-related.

As people live longer and develop more chronic diseases, the number of supposedly alcohol-related admissions is bound to rise, regardless of whether alcohol-related problems are getting better or worse. There is a large amount of garbage in, garbage out about alcohol-attributable fractions and it leads to counter-intuitive findings like this. They are counter-intuitive because they are probably wrong.

Nevertheless, if we look at diseases that are specific to alcohol or that result in the 'hard endpoint' of death (which is less susceptible to miscounting), there is no evidence of an improvement since the decline in consumption began twelve years ago. The rate of alcohol-related death has not risen but nor has it fallen.

That is because the total population theory is nonsense. It is a theory created to justify policy, rather than the other way round. It is perfectly obvious that reducing per capita consumption does not magically reduce heavy drinking or alcoholism (although a decline in heavy-drinking could certainly lead to a decline in per capita alcohol consumption). The overall decline in consumption seen in Britain since 2004 has not had any of the effects that would be predicted by the total consumption theory.

The 'public health' lobby are too thick and lazy to help people who are putting their health at serious risk from alcohol. It's much easier to go to conferences, talking drivel about 'the whole consumption curve' and lobbying the government for regressive, pig-headed, anti-market policies than it is to target the small minority of drinkers who are at serious risk of harm and who require help.

No comments: