No, obesity isn’t ‘set to soar’ — I’d bet money on it

    11 October 2016

    Today is World Obesity Day and so the usual warnings of the coming Armageddon have been sent out to the newspapers. Back in 2007, the ironically entitled Foresight report predicted that 36 per cent of British men would be obese by 2015. But 2015 came and went with an obesity rate of 25 per cent, which is more or less what it was when the prediction was made.

    A slew of dodgy predictions have been made in the years since, so many that mere chance should have ensured that one of them would be right by now, but no. All the short-term forecasts have been proven to be wrong and all the long-term forecasts look deeply implausible. A number picked at random would have more chance of predicting future obesity rates than the computer models of the ‘public health’ industry.

    Graph1 Graph2

    In reality, the models are no more than crude extrapolations of old trends. The trouble is that recent trends are not very dramatic. Childhood obesity is particularly awkward for the doom-mongers since it remains at the same level as it was in 2002, as the graphs on the left show.

    Faced with childhood obesity rates that stubbornly refuse to rise, nanny statists resort to claiming that rates are ‘set to soar’ in the near future. The reasons for this imminent upsurge are never explained, but by the time the data disproves the prediction, everybody will have forgotten about it and the campaigners will have started afresh with another scary forecast.

    And so, to celebrate World Obesity Day, we get stories like this from the Guardian:

    Obesity set to soar for boys, but not girls, from poorer homes

    Childhood obesity is set to increase so sharply among boys from poorer homes in England that three in five of them will be dangerously overweight by 2020, research shows.

    … At the moment, 16% of boys in England aged between five and 11 from the most deprived backgrounds are overweight and another 34% are obese, with only 51% a normal weight. But on current trends those figures will rise to 17% (overweight) and 43% (obese) by 2020, leaving only a minority – two out of five (40%) – who have a healthy body mass index.

    You will have noticed immediately that the campaigners have narrowed their prediction down to a particular subset of one gender. What is less obvious is that their numbers are simply wrong. Current obesity rates are nowhere near 34 per cent among boys from the poorest homes. The National Child Measurement Programme reports a rate of 12 per cent for the poorest five-year-olds rising to 24 per cent for the poorest 11-year-olds. Since the average figure for five- to 11-year-olds is somewhere between these two numbers, we must conclude that the Obesity Health Alliance — for it is they who are responsible for this soothsaying — think that obesity rates among boys from the poorest families are going to double by 2020.

    How likely is this? The Health Survey for England uses figures for two- to 15-year-olds so let’s look at those. The graph below shows obesity prevalence among boys in the highest and lowest income quintiles.


    There is no real change since 2006, although there is a sharp uptick in the last year (2014) for both groups. Indeed, the rate of obesity appeared to double among boys from the wealthiest homes in 2014. It is unlikely, bordering on the impossible, that this doubling is real. As the authors of the survey note, the sample sizes are quite small once they are sliced into five income groups and so sharp year-to-year fluctuations are common.

    This can be seen by looking at the girls’ graph (below). At several points, the obesity rate among the wealthiest girls appears to double and then halve. Between 2008 and 2010, for example, it fell from 12 per cent to six per cent and then rose to 12 per cent again. In the latest year, it appeared to double from nine per cent to 18 per cent. A naive reading of this data would suggest that there are no longer any socio-economic differences in obesity rates among girls, but this goes against everything we know about obesity trends and the authors dismiss the finding as an aberration.


    In other words, the data is noisy and not terribly reliable except as a guide to long-term trends. A sensible interpretation of the data is that (a) childhood obesity rates are not rising, (b) children in low-income groups are more likely to be obese than those in high-income groups. Whether obesity rates rise in the future remains to be seen but there is no particular reason to assume they will. Nor is there any reason to assume that they will rise among boys but not girls, or among poorer children but not wealthier children. In fact, I would happily wager a reasonable sum of money on obesity rates being well below the predicted 43 per cent in 2020 for any group you choose to name.

    It’s all so much flim-flam, but the objective is not to make an accurate forecast. The objective is to scare people into accepting bans, taxes and restrictions. A sugar levy is already planned for 2018, and when you recall how loudly the ‘public health’ lobby campaigned for this regressive tax, you’d have thought their models would show a fall in childhood obesity between now and 2020. Instead they forecast an implausibly large increase. If the sugar tax is not going to have any impact — and of course it won’t — why are we going ahead with it?

    In practice, the sugar tax has already been largely forgotten and the health zealots now dismiss it as ‘no silver bullet’ (ie, it won’t work). Their sights are set on banning advertising, with a spokesman for the Health Equalities Group, a state-funded lobby group, making the risible claim that obesity rates are higher among boys because boys are ‘more brand loyal and therefore susceptible to the billions of pounds spent on marketing to children’.

    They’ve already succeeded in banning the advertising of tasty food during children’s programmes but that, too, proved to be ‘no silver bullet’, unless the bullet was intended for after-school kids’ programmes on ITV which came to an end as a result.

    My offer of a wager is genuine, by the way, but I know that nobody in ‘public health’ will take me up on it.