Friday, 1 April 2016

Endless obesity babble

Here we go again. How many prophecies have to fail before we build up a resistance to them? I have written about obesity predictions before so I won’t bore you with the details again, but the key facts are these. In 2007, the government’s ironically-titled Foresight report predicted that 36 per cent of men would be obese by 2015. When this forecast was made, the male obesity rate in England was 24 per cent. By 2014 - the year for which we have the latest data - it had rocketed to, er, 24 per cent. 

In 2011, the Lancet predicted ‘a rise in obesity prevalence in men from 26% to 41–48% and in women from 26% to 35–43% [by 2030].’ The female obesity rate is currently 27 per cent whereas the male obesity rate has dropped slightly to 24 per cent. There is still time for the Lancet to be proven right, but the current trajectory is way off target.

Figures from the rest of the UK are similar to those from England. Adult obesity in Wales is currently at 22 per cent. In Scotland it is 28 per cent. In Northern Ireland it is 25 per cent.

If obesity rates are rising, they are doing so very slowly. I would not be foolish enough to make a prediction of my own, but the evidence from the last decade is more consistent with a levelling off than a surge. Since 2002, the male obesity rate in England has hovered between 22 and 26 per cent. For women, it has kept within the range of 23 and 27 per cent. At the most, adult obesity has risen by two percentage points since the Foresight report was published under Tony Blair. Childhood obesity peaked in 2004 at 19 per cent and has since fallen to 17 per cent.

The woeful track record of obesity predictions can be explained in four words: ‘If current trends continue…’ Between 1993 and 2002 there was a clear and steep increase in obesity. The adult rate in England rose from 15 to 23 per cent. If that trend had continued we would indeed be looking at half the population being obese by 2030. But it didn’t and there was never any reason to assume it would. It now seems clear that the 1993 to 2002 period was exceptional. It is absurd to assume that an exceptional trend is going to continue in a linear way indefinitely, particularly when the trend in question is no longer 'current'.

Extrapolating from current trends is a terrible way to make a prediction. At the time of the Foresight report, the bankers had looked at current trends and decided that house prices would keep on rising forever and, therefore, that there was no risk in lending money to people who had no chance of paying it back. Like the obesity forecasters, they were making a crude and naive error.

Given what has happened to obesity rates in the UK over the last ten years, you might expect obesity forecasters to be more circumspect, if not apologetic, but you would be wrong. Today saw the publication of yet another study from the Lancet which doubles down on its previous gamble by forecasting a 38 per cent obesity rate in Britain by 2025.

Actually, it doesn’t. The study contains no specific predictions for the UK. The 38 per cent figure only appears in a sexed-up press release sent to journalists. Readers can make up their own minds about what this says about standards at The Lancet these days, but it cannot be denied that they know how to play the media. The 38 per cent prediction is all the newspapers and, as always with these guesstimates, is being treated like the written word of God.

Contrast this media fanfare with the radio silence that greeted the publication of the official obesity figures in December. Those figures went almost entirely unnoticed, presumably because they once again failed to show any change. For the umpteenth year running, the Health Survey for England noted that around a quarter of adults are obese. Ho-hum.

Something has surely gone wrong when highly implausible, unpublished obesity predictions are given blanket media coverage while actual data is routinely ignored. No one has reflected on the dismal track record of all previous obesity forecasts, nor has anyone asked how the two percentage point rise in obesity seen in the last ten years translates to a twelve percentage point rise in the next ten years (as would have to take place for the Lancet’s latest piece of soothsaying to be correct). The increasing reliance on outrageous predictions about the obesity 'time bomb' seems designed to deflect attention from the stubborn refusal of the bomb to go off.

Cross-posted at Spectator Health

PS. Sarah Knapton, the alleged science editor of the Telegraph, confirmed that she is a nitwit with this gem in her article about the Lancet's claim...

Obesity currently costs the health service £47 billion a year,  just under half the entire NHS budget.

Half of the NHS budget spent on obesity?! It beggars belief that anyone could type these words and not think 'hang on, that can't possibly be right'.

It's not right, of course. The £47 billion is made up of societal costs (intangible costs, lost productivity, etc.). The costs to the NHS are a tenth of that. 

No comments: