diet drinks

    Ignore the headlines. Diet drinks don’t make you fat

    5 January 2017

    The front pages of yesterday’s newspapers had some sensational news for the yo-yo dieters and temporary gym members of post-festive Britain. Readers of the Times were greeted with ‘Diet drinks no healthier than sugary versions, scientists warn’, while the Telegraph ran with ‘Diet drinks can pile on the pounds’. The Guardian was only slightly less excitable with the headline ‘No evidence sugar-free soft drinks aid weight loss – study’.

    There are a few problems with this. For example, there was no study and diet drinks do not make you fat. It was worthless as health advice, but as an example of how to get ‘public health’ misinformation into the media it was almost flawless.

    times I’ll say it again: there was no new research. All the media coverage was based on a commentary in a journal. Despite being essentially an opinion piece, readers could be forgiven for assuming it to be a fresh scientific study. Why else would the Guardian use the word ‘study’? Why would newspapers cover it at all, let alone on the front page, unless something new had been discovered? Those are questions for another day. Let’s just say that this is not the first time it has happened. Loyal readers will recall an opinion piece in the journal Addiction becoming front-page news on two national newspapers back in July.

    The opinions expressed in the commentary are highly controversial and perhaps deliberately inflammatory. It was criticised by nutritional experts within hours and its most eye-catching claim — that diet drinks could be a ‘potential risk factor for highly prevalent chronic diseases’ — is unevidenced in the article and almost entirely speculative.

    Insofar as it attempts to review the evidence on diet drinks and weight loss, the analysis is perfunctory, misleading and frequently inaccurate. The authors cite a number of randomised controlled trials to show that the evidence is ‘mixed’, with some studies finding no effect on weight and others finding a ‘modest reduction’, but several of the studies they reference do not even look at weight loss. They claim that this study found no evidence of weight loss (it did) and cite this study as evidence that diet drinks lead to both weight loss and no weight loss. In truth, the scientific base is not so ‘mixed’. The balance of evidence from randomised controlled trials strongly points to a positive effect.

    To make the package even more irresistible to the press, the commentary was press released with additional comments from one of its authors, Christopher Millett, that were even more sensational than those in the article itself. ‘Far from helping to solve the global obesity crisis,’ he said, ‘artificially sweetened beverages may be contributing to the problem.’ Quite how the substitution of zero-calorie drinks for sugary drinks could contribute to obesity is barely explored, let alone proven, in his article.

    A one-eyed view of the evidence, a bit of hyperbole, and a spot of exaggeration. That’s how you get front-page news coverage in the world of ‘public health’ without having to do any research of your own.

    So what’s the reality? The fact is that virtually all randomised controlled trials which substitute artificial sweeteners for sugar find evidence of weight loss. Much the same is true of artificially sweetened soft drinks (see here, here and here, for example). The amount of weight loss is usually not great — diet drinks are not magic wands — but it is real, and they certainly don’t make you fatter.

    To be fair, the evidence from observational studies is less impressive. It has often been noted that people who drink diet drinks are not always very thin. The next president of the United States once tweeted that he has ‘never seen a thin person drinking a Diet Coke’. This phenomenon can be seen in epidemiological studies which compare people’s dietary habits to their physical characteristics, but the authors of yesterday’s commentary correctly attribute this to ‘residual confounding’ (ie, people who drink fizzy drinks tend to eat more food than those who do not, regardless of how the drink is sweetened) and ‘reverse causality’ (ie, some people start drinking diet drinks when they get fat, but did not get fat by drinking diet drinks).

    It is sometimes argued that people feel that having a zero-calorie drink gives them permission to have another doughnut, thereby increasingly their calorie intake. It has also been suggested that artificial sweeteners make people hungry. These theories seem to be the only basis for Millett’s claim that diet drinks are contributing to the ‘global obesity crisis’, but the only evidence cited for them in his commentary is a letter to the Lancet in 1986 and an evidence review that gives short shrift to both theories.

    Don’t let them worry you. If you are trying to work off the mince pies, you would be well advised to switch from Coke to Diet Coke and start putting slimline tonic in your gin. It’s not the whole answer but it will do some good and it certainly won’t do you any harm.

    The idea that zero-calorie products are a cause of obesity is absurd on its face and is unsupported by the evidence. It is not merely junk science, it is anti-science in that it implies that obesity is not caused by a surplus of energy but by some magical process involving bubbles and tin cans. It is front-page news because it is sensational, and it is sensational because it is not true.