Do diets work?

Photo by petekraynak –

You probably think there are hundreds of different diets. There aren’t. There are two.

The first type, which I call the thermodynamic diet, is the sort of diet your friendly neighbourhood physicist might like. “Energy cannot be created or destroyed”, she says, “it’s basic thermodynamics.” Your body is using up energy for its various functions, like breathing, excreting waste, and walking up the stairs. So you need to take in this amount of energy from your food to break even. If you eat more, your body can save some of that energy in the form of molecular bonds by turning simple chemicals (sugars) into complex ones (fats) and storing them for later. If you eat too little, your body reverses the process. It takes fat, turns it back into simpler molecules and in doing so releases energy.

The second type, which I call the metabolic diet, takes a more biological approach. Food is made up of lots of different chemicals, for sure, but if we ignore the trace minerals, the vitamins, and the fibre which your body can’t get much energy from, we’re left with fat, carbohydrates, and protein —the macronutrients. The metabolic diet says that the relative balance of these macronutrients is all important. Fat makes you fat. Carbohydrates cause an insulin spike which enhances fat storage. Sugar is evil. You get the idea. These diets generally have in common an appeal to hormonal mechanism. They all talk about how the body deals with these various macronutrients and theorise some crucial (and usually evil) role for one of the three. (Sometimes the evildoer is a micronutrient but the general logic still applies; this is a substance that has a special effect on your body.)

Virtually every diet you’ve ever heard of is one of these types. Atkins? Metabolic. Red green? Metabolic? Cabbage soup? Thermodynamic. The Zone? Metabolic. Weight Watchers? Thermodynamic with a hint of metabolic.

In fact, there’s even less choice amongst dieting approaches than you think. Both types of diet talk almost exclusively about what you eat and talk very little (if at all) about how you eat it. The authors of diet books rarely remember that we’re living thinking beings for whom emotion is as important as logic. Yes it can be difficult to have to work out how many grammes of carbohydrates you’ve consumed in each meal for a week, but it’s even harder to resist the temptation to eat a piece of chocolate cake when everyone else in the office is having some.

Why am I telling you this? We need to realise that diets are much of a muchness because whenever anyone claims that restrictive diets don’t work, or don’t work in the long term, the immediate response is, “Ah! But you’ve clearly not tried the aubergine and cider diet! It’s amazing!” Sorry. But no. Notwithstanding the distinction above, diets are actually very similar. And so if we study the effects of a few of them, we can fairly safely begin to draw conclusions about the rest.

And that’s exactly what Traci Mann and colleagues at the University of California did. To answer the question of whether diets work, they searched through the records of hundreds of scientific studies, to find reports of controlled trials where participants were randomly put on a diet programme or put on the waiting list, and then followed up to see how their weight had changed. The first thing to note from their work is the astonishing lack of evidence. They found lots of studies, but most of them were quite poorly designed. For example, they found only 5 studies that followed people up long-term (years rather than months). This is crucial. No-one really wants to know whether a diet will help them loose weight for a month or two. We all want to know what’s going to help us keep the weight off long-term.

In the five studies that reported long-term follow-up of controlled trials, the dieters had lost an average of 1.1kg (2.4 lb). The differences were often statistically significant, but not clinically significant. The participants were still obese after years of dieting.

I’ve been searching for days now to find evidence to the contrary, and to be honest I can’t. There’s the odd trial that suggests that a diet works really well, but it’s usually flawed. Either there’s no control group, or the authors were paid by the makers of the food supplement on trial, or effects were only maintained for a couple of months.

Mann and colleagues reach the conclusion that dieting doesn’t work. And having read the scientific tests of lots of diets, I’m inclined to believe them.

It’s worth saying that this doesn’t necessarily mean there’s no truth in some of the basic science. It may in fact be true that eating lots of protein and reducing your carbohydrate intake will encourage your body to produce less insulin. Medicine has long recognised the difference between efficacy and effectiveness—the difference between whether or not something works in the lab and whether or not it works in the real world. A drug that works brilliantly in laboratory conditions (high efficacy) but which has lots of side effects and can’t be taken with alcohol will have low effectiveness because few people will continue to take it.  The theories behind metabolic and thermodynamic diets might well be correct (it’s very hard to argue with thermodynamics) but if they result in food cravings, having to mess about with kitchen scales, or always having to say ‘no’ to nice things at parties, few people are going to stick to them.

And thus, few people are going to loose a great deal of weight.