Why obesity predictions are propaganda

    22 June 2015

    Two weeks ago I wrote an article for the Spectator in which I argued that obesity predictions are not worth the paper they’re written on and that the healthcare costs of obese people tend to be lower than those of people who are of ‘normal’ weight. Both arguments can be easily backed up with evidence, but since they undermine the narrative of an obesity time bomb destroying the NHS, they are rarely acknowledged and often contradicted.

    An article in the Huffington Post offers such contradiction. Written by Aseem Malhotra (Action on Sugar), Mike Lean (University of Glasgow) and Mahiben Maruthappu (senior fellow to the CEO of the NHS), it begins with a lengthy discussion of advertising which is not directly relevant to my original article so I will resist commenting on it here. When it comes to the obesity predictions, the authors say that ‘Snowdon critiques the “wild predictions” of “campaigners” that 75% of UK adults will be overweight or obese by 2030. It might actually be sooner: the figure is already 68%’.

    As a point of fact, the wild prediction said that 75 per cent of men would be overweight or obese by 2030. In England, the current figure is 67 per cent, which is slightly lower than it was in 2001. If we take both sexes together, the current figure is 62 per cent, which is exactly the same as it was in 2001.

    And this is my point. Since the numbers haven’t risen in the last decade and a half there is no reason to think that they will suddenly surge upwards in the next decade and a half (though they might, of course). Predicting such things is a fool’s errand, but if you had to make a forecast based on current trends, you would start by acknowledging that the current trend is flat. Given that previous predictions of spiralling obesity have been shown to be hopelessly wrong, similar predictions made today should be treated with scepticism. A recent episode of Radio 4’s More or Less is worth listening to on this topic.

    Rather than asking how plausible the predictions are, the authors simply remind the reader that lots of people are obese, with the implication that obesity must therefore be spiralling. ‘By the age of 65,’ they write, ‘some 80% of UK adults currently become overweight, and almost 40% are obese.’

    These figures are not correct. Looking at the Health Survey for England (the figures are similar for the rest of the UK), obesity is nowhere near 40 per cent for any age group. The obesity rate in 2013 was 24.9 per cent overall, ranging from 11 per cent amongst 16-24 year olds to 32 per cent amongst 65-74 year olds. No reasonable person would consider 24.9 per cent, or even 32 per cent, to be ‘almost 40%’.

    In any case, the real question is about the trend. The 24.9 per cent figure recorded in 2013 is higher than the 22.4 per cent figure from 2001 but lower than the 26.1 per cent figure recorded in 2010. In other words, obesity rates are increasing very gradually if at all. And if they are rising, they are rising much slower than had been predicted.

    Having failed to address my first point, Malhotra and co move on to the thorny issue of the respective healthcare costs of fat and thin people. They say that the claim that ‘obesity saves money, because people die younger, is absurd’. Why so? Because it is ‘insulting to the intelligence of those who work to treat chronic diseases, and insulting to the obese people who struggle for years to try to control their problem’.

    Regardless of who may or may not feel insulted, it is a claim rooted in economic evidence (the reference to hurt feelings and the appeal to incredulity suggest that the authors’ problem with it may be more emotional than rational). They say that ‘Health economic analysis worldwide, carefully checked by peer-review, have never produced any evidence that obesity is not a major burden on healthcare.’ This is simply not true. I explicitly cited a peer-reviewed study in my article which concluded that ‘total lifetime health spending was greatest for the healthy-living people, lowest for the smokers, and intermediate for the obese people.’ This was thanks to ‘differences in life expectancy’ – ie. the obese people died younger, thereby saving the state money. If the analysis were extended to other state budgets, notably pensions, the savings would likely be even higher.

    This, of course, is not an argument for more obesity. It would be crass and – yes – insulting to advocate obesity as a way to cut costs, but since ‘public health’ campaigners insist on claiming that obesity is going to bankrupt the NHS, some cold economic facts are in order. The basic conclusion of the study mentioned above was confirmed in another study last year, among others. These studies could be wrong, but it is ridiculous to deny that they exist.

    Rather than point to a flaw in the economic analysis, Malhotra and friends simply state that ‘obesity is directly costing the NHS close to £6billion per year’. They also mention ‘additional indirect costs to the economy’ of which they claim that ‘Type 2 diabetes alone adds £20 billion’. As indirect costs are things like lost productivity which are not borne by the NHS, they are not relevant to the argument, but it is still worth looking at this £20 billion figure to appreciate the irony of their article being headlined ‘The truth about obesity’.

    There are several things wrong with it. Firstly, although obesity is a major risk factor for Type 2 diabetes, most cases are not caused by it, and Type 1 diabetes is not caused by it at all; it is therefore quite wrong to imply that the full cost is obesity-related. Secondly, they are presumably referring to this standard reference study, which finds an indirect economic cost of £20 billion for both forms of diabetes, not just Type 2. Thirdly, the estimate is a projection for 2035, not today (the current estimate is much lower). Fourthly, £20 billion is the top end of an estimate that ranges from £13.8 billion to £20 billion.

    The £6 billion cost-of-obesity estimate includes the cost of obesity-related diabetes, so there is no need to throw extra figures about diabetes into the pot. The £6 billion is itself questionable – it is a projection made several years ago which assumed that rates of obesity would rise faster than they have, and other studies have arrived at much lower estimates) but that doesn’t really matter here. There are obviously costs associated with treating obesity-related diseases. The point is that they are gross costs, whereas the taxpayer is only affected by net costs. The net cost of obesity is the cost of treating obesity-related diseases minus the costs that would have been incurred had the patient not been obese. Non-obese people do not avoid healthcare costs, they just have different diseases to treat and these will tend to be more expensive over the course of a lifetime.

    As van Baal et al explain in the study mentioned above: ‘Obesity increases the risk of diseases such as diabetes and coronary heart disease, thereby increasing health-care utilization but decreasing life expectancy. Successful prevention of obesity, in turn, increases life expectancy. Unfortunately, these life-years gained are not lived in full health and come at a price: people suffer from other diseases, which increases health-care costs. Obesity prevention, just like smoking prevention, will not stem the tide of increasing health-care expenditures. The underlying mechanism is that there is a substitution of inexpensive, lethal diseases toward less lethal, and therefore more costly, diseases.’

    If you want to know whether something is a burden on the public finances, it is not good enough to merely identify costs. You must also look at the costs of the alternative. In this instance, this means looking at the costs of ‘substitute diseases’. By ignoring these, Malhotra and friends spectacularly – and perhaps deliberately – miss the point.

    Finally, they mention that incidence of diabetes is at ‘an all-time high, affecting 3.9 million Brits’. This is true, and it is the single biggest problem associated with obesity. The rise of diabetes and the costs associated with it are very real but, once again, rates have risen much slower than predicted. In 2004, it was claimed that there would be six million people living with diabetes by 2010. Five years after that date passed, the actual number is still less than four million. That is not a reason to be complacent, but it is another example of public health projections being way off base.

    When The Times reported the diabetes forecast in 2004, it also predicted that: ‘On present trends, half of all children in England in 2020 will be obese’. In 2004, the childhood obesity rate was 19 per cent. It is now 15 per cent. This brings me back to the argument of my original article. Obesity predictions have been shown time and time again to be worthless. Along with the mangled and exaggerated estimates of the cost of obesity to the NHS, they have value only as propaganda.