Gary Taubes

THE DIET DELUSION

Unfortunately, as every clinician knows. few people who are obese or overweight manage to lose much weight, and those who succeed do so at the cost of perpetual self-denial and mild hunger, which explains why most of them regain the lost weight within a few years. So there seems to be something wrong with the basic assumptions on which current advice is based. Perhaps obesity is not really a behavioural problem or due to lack of "will power".

Gary Taubes is a respected science journalist who has written a comprehensive challenge to the currently accepted wisdom. He didn't expect this when he started his investigations, he tells us, but he has become convinced that most of what we think we know about diet is wrong.

What I particularly like about his approach is that he takes the long view and goes back in time to show how ideas about obesity changed radically during the twentieth century, often as the result of vociferous advocacy by certain influential people. In fact, he begins even further back, with the publication in 1862 of William Banting's description of his experience of reducing his carbohydrate intake to lose weight.

Banting's decision was based on medical advice, and carbohydrate restriction remained the essential advice to dieters throughout the first half of the twentieth century. But by the 1970s a complete revolution in thinking had taken place. Now we were being told that it was a mistake to limit carbohydrates and the real enemy was fat, especially saturated fat.

We are quite often exhorted to adopt a diet of the currently fashionable kind— low in saturated fats, high in carbohydrates— because this is supposed to be what our ancestors in the Palaeolithic were eating: nuts. fruits, vegetables and small game. The ability to hunt large animals, on this hypothesis, would only have been acquired in the last 25,000 years or so and the eating of large quantities of fat is therefore "unnatural." But it is difficult to know what our remote forebears were eating, and what evidence there is, based on studies of modern hunter-gatherers, points to a strong preference for fat. Only since the introduction of agriculture in the last 10,000 years have carbohydrates become such a prominent feature of human diet. If anything, therefore, it is our modern diet that is "unnatural".

There are villains and heroes in the story that Taubes tells here. The ghost of Robert Atkins hovers over much of the discussion, and although Taubes does not specifically endorse the Atkins diet or any of the other current low-carbohydrate approaches it is clear that he thinks Atkins was on the right lines. One of the principal villains is Ancel Keys, whose advocacy of the role of dietary fat in heart disease, initially greeted with scepticism, shaped conventional thinking decisively after the 1950s. Taubes discusses his research at length and concludes that it was fatally flawed.

The book has three parts. Part 1, "The Fat-Cholesterol Hypothesis", shows that the advice to avoid saturated fat was based on an unjustified extrapolation from the effects of taking cholesterol-lowering drugs. In fact, diet has little influence on serum cholesterol levels, nor is there good evidence that adopting a low-fat diet reduces the incidence of cardiovascular disease.

Part 2, "The Carbohydrate Hypothesis", puts forward the case for believing that carbohydrates, especially refined carbohydrates, cause an excessive rise in insulin secretion which in turn leads to deposition of fat. Part 3, "Obesity and the Regulation of Weight", is the longest section, and offers a more detailed explanation of how and why an excessive carbohydrate intake leads to obesity.

At the end of his book Taubes provides a useful summary of the conclusions that his five years' research has led him to (p.454). In essence, these are that dietary fat is harmless and it is carbohydrates that cause obesity through their effect on insulin secretion. When insulin levels are high, either chronically or after a meal, we deposit fat in our fat tissue. When insulin levels fall, fat is released and metabolised. "The fewer carbohydrates we consume, the leaner we will be."

Although Taubes is not himself a clinician or a laboratory researcher, his book is an impressive example of critical science at its best. In it he shows convincingly that most of the received wisdom about diet is insecurely based in terms of evidence. The carbohydrate hypothesis emerges as at least a plausible alternative, which deserves to be taken seriously. Taubes concludes with a plea for better-quality research to test the validity of this idea.

This is an important book. Although it is written clearly and accessibly, it is long, with no shirking of physiology, and the casual reader may be overwhelmed by the sheer weight of evidence that Taubes cites. The effort is well worth while, however; he is a sceptic in the true sense of that often misused term.

13 April 2010