The myth of the low-fat diet

Looking for something healthy and non-fattening for your evening meal? How about a nice porterhouse steak, which is 50-50 fat and protein? It's a suggestion that comes about as close to heresy as we get these days, but there is increasing evidence that a low-fat diet is not the panacea we have been promised. For the past 30 years such a diet has been officially promoted, on both sides of the Atlantic, as the route to plaque-free arteries and a slim figure. A message that has propelled 15,000 low-fat products on to American supermarket shelves.

Looking for something healthy and non-fattening for your evening meal? How about a nice porterhouse steak, which is 50-50 fat and protein? It's a suggestion that comes about as close to heresy as we get these days, but there is increasing evidence that a low-fat diet is not the panacea we have been promised. For the past 30 years such a diet has been officially promoted, on both sides of the Atlantic, as the route to plaque-free arteries and a slim figure. A message that has propelled 15,000 low-fat products on to American supermarket shelves.

However, the campaign has had no obvious effect on the incidence of heart disease, nor have the pounds been falling off the national waists and hips. In fact, according to a recent report, we are getting fatter. Not only is a low-fat diet largely irrelevant to reducing heart disease but it may be responsible for the worrying rise of diabetes.

Praise for the fat-laden porterhouse steak came in an award-winning investigative article, published last year, on just how little evidence there is supporting the low-fat dogma ( Science, 30 March 2001). Virtually ignored in the UK at the time, it should be required reading for anyone interested in diet. The problem with the low-fat message is that it is far too simple.

For instance, we've all been told to avoid animal fats because they are saturated and that saturated fat raises cholesterol levels in the blood. But half the fat in a steak is actually "monounsaturated" – the same type as found in "good for the heart" olive oil. The other half is, indeed, saturated but about a third of it is a type called stearic acid, which, like olive oil, raises the "good" HDL cholesterol in the blood. So just 30 per cent of the fat in a steak is the sort of saturated fat that can raise "bad" LDL cholesterol. However, even this demonised fat will simultaneously raise the "good" HDL. "All of this suggests," writes science journalist Gary Taubes, author of the Science article, "that eating a porterhouse steak rather than carbohydrates might actually improve heart disease risk".

The recommendation that dietary fat be reduced to 30 per cent of the total calorie intake is contained in a 1976 Senate report. Written by a journalist, who had only previously reported on labour relations, it drew on just two days of testimony, most from an eccentric Harvard nutritionist Mark Hegstead, who regarded dietary fat as the nutritional equivalent of cigarettes.

That would not have mattered had the evidence come in to support his recommendation – but, beyond a certain point, it hasn't. Undoubtedly if you are at high risk of having a heart attack – overweight, high blood pressure, no exercise, etcetera – and you have very high levels of cholesterol, then reducing them with diet or drugs can significantly reduce your chance of a heart attack. What has not been shown convincingly, however, is that someone who is not at risk will have their life cut short as a result of regularly eating more than the recommended level of dietary fat. As Taubes reports in his article, at least four large trials between 1980 and 1984 comparing disease rates and diet "showed no evidence that men who ate less fat lived longer or had fewer heart attacks".

Since the early Seventies Americans' fat consumption has dropped from an average of 40 per cent of the diet to 34 per cent, but the incidence of heart disease hasn't fallen too. In fact, between 1979 and 1996, largely reflecting the range of new developments, the number of medical procedures for heart disease increased from 1.2 million to 5.4 million. At the same time the proportion of obese Americans has soared from 14 per cent to 22 per cent.

A low-fat diet may be actively harmful. In the late Eighties, David Jacobs, from the University of Minnesota, did a study in Japan on the effects of cholesterol and, interestingly, found a link between low blood cholesterol levels and an increase in non-heart disease related deaths. He reported to the National Heart, Lung and Blood Institute, which hosted the American Heart Association conference in 1990. At that conference the results of 19 studies from around the world on the links between cholesterol levels and disease were pooled. Taubes writes: "The data were consistent. When investigators tracked all deaths instead of just heart disease, the cholesterol curves were U-shaped for men (both high and low increased the risk) and flat for women." He adds: "As for women, if anything, the higher their cholesterol the longer they lived."

Meanwhile, the link between low-fat diets and weight loss hasn't fared well either. The ongoing Women's Health Initiative – a $100m study on women's health – enrolled 50,000 women in a randomised trial, putting half of them on a draconian diet that provided only 20 per cent of their calories from fat. After three years they had lost, on average, just one kilogram.

Critics of the low-fat hypothesis, such as Peter Ahrens of Rockerfeller University in New York City, have always been concerned that simply lowering fat intakes could have a range of unforeseen effects. Fat is a major component of cell membranes, the brain is 70 per cent fat, and changing fat ratios could affect all sorts of processes, from immune responses to hormone levels.

Just how much else is involved in determining the effect of fat levels in the diet was illustrated by the Lyons Diet Heart Study (16 February 1999). This involved two groups of heart attack survivors, one getting a typical low-fat diet and the other a Mediterranean diet with more bread, cereals, beans, vegetables, olive oil, fruit and fish. The total amount of fat and the type of fat type that each group ate were very different. Intriguingly, however, since high fat is supposed directly to affect cholesterol, the cholesterol levels in the blood of the two groups were very similar. After four years the Mediterranean group had had 14 heart attacks, compared with 44 for those on the "low-fat" diet. This suggests that reducing blood cholesterol is not simply a matter of reducing dietary fat. What is crucial, it turns out, is the type of fat and what you eat along with that fat.

A danger of the low-fat advice may be that it is encouraging us to eat too much of the wrong sort of food. Given the chance, people tend to eat about the same amount of calories, however varied their composition, and those who eat lots of meat and dairy products, like the Finns or Americans, tend not to eat lots of vegetables and fruits. So if you reduce fat, it is likely to be replaced with refined carbohydrates, and that seems to be the problem.

Troublingly, the evidence has been growing that diets high in carbohydrate can increase the blood level of dangerous fats called triglycerides and reduce the "good" or HDL cholesterol. Diets high in sugar and other carbohydrates may also lead to a condition called insulin resistance – the extra carbohydrates are turned into extra glucose, which makes the body produce extra insulin and after a while the body becomes less sensitive to insulin. This combination produces something that Stanford endocrinologist Gerald Reaven has called "syndrome X" ( New Scientist, 1 September 2001).

In the United States an estimated 30 per cent of males and 10 per cent to 15 per cent of post-menopausal women have insulin resistance, which commonly leads to diabetes and is linked with a raised risk of heart disease. High-energy snacks are one way to expose the liver to damaging levels of insulin, although exercise can keep the level of harmful triglycerides down. Another element of the high-carbohydrate diet that has been linked with syndrome X is a high consumption of a type of sugar known as fructose. Fructose makes up half of ordinary sugar but corn syrup, now used to sweeten a vast range of foods – breakfast cereals, many low-fat snacks and fizzy drinks – is almost pure fructose. Rather than a high-fat diet, a major contributor to our creeping obesity epidemic could be increased consumption of carbohydrates, especially those coming from sugars.

What's very interesting is that researchers who are concerned about syndrome X come up with the same sort of dietary advice to avoid it as those who are studying fats and heart disease. One again, olive oil, fish oils, plenty of fruits and fresh vegetables and slow release carbohydrates like lentils, beans, brown rice and oats are recommended as a way of avoiding insulin resistance, as well as rendering saturated fats safe.

One reason for the survival of low fat as a recommended diet, which really only seems relevant to people at risk of a heart attack due to high cholesterol, is the difficulty of giving general dietary advice. Not only do fats and cholesterol levels interact in a variety of complicated ways but so also do the ways we lay down fat. The billions that a successful anti-fat pill would generate, means huge amounts of research effort are being put into research to uncover the multiple pathways controlling appetite and fat storage. The complexity of this emerging system suggests why the simple low-fat mantra has proved so unsuccessful as a weight loss programme, too.

A glimpse of just how complicated this system is came from a study by researchers at Rockerfeller University, using an extraordinary technique to trace the brain areas involved in appetite. A pseudo-rabies virus, which infects linked brain cells, genetically modified to produce green glowing jellyfish protein, was injected into rats' brains. Most hunger research concentrates on the appetite centre in the hypothalamus but tracing the green lines left by the virus revealed that brain centres that control the emotions, smell and the higher centres had all been infected. Our appetite seems very much part of who we are.

A similar programme to begin to tease out what is involved in the relationship between dietary fats and cholesterol should eventually yield more sophisticated and useful advice than the low-fat diet.

A longer version of this article appeared in the monthly newsletter 'Medicine Today'. www.medicine-today.co.uk