A study is now under way at America's National Institute of Aging, in Washington. Two groups are being fed identical diets of the ordinary foods - burgers, ice cream and so on - eaten by the average American. One group is required to eat all they need between 4pm and 7pm, and then nothing until the same time the following day.
The aim is to find out whether consuming all one's daily calories in a couple of hours is better for health. Dr Mark Mattson, chief of the Laboratory of Neurosciences at the Institute and the man leading the research, says it is only recently, in evolutionary terms, that humans have eaten three meals a day. Our ancestors were fortunate to eat once a day and often went for days without food.
From an evolutionary point of view, our bodies are accustomed to feasting and fasting rather than grazing. Three or more meals a day may be good for young people, but not for adults.
This idea - that we need to pay attention to the frequency of our eating as well as the amount consumed - was popularised in America in The Warrior Diet by Ori Hofmekler, in 2002. It is based on the idea that we should eat like early humans, consuming all daily calories in the early evening.
The Warrior diet is now a major brand in the US, used to market books, products and seminars. But it has received no attention in the UK. Its website (www. warriordiet.com) proclaims: "Humans are primarily destined to follow certain feeding cycles and physical activities. In accordance with the 'Thrifty Genes theory', scientists speculate that humans have adapted to survive better during cycles of feast and famine, action and rest. The current epidemic of obesity, diabetes and impotence bears testimony to the fact that humans today have betrayed their biological destiny."
Even if you think this is nonsense, it is surprising that something as obvious as meal frequency and its impact on health has never been systematically tested. Many religions practise fasting. During Ramadan, which lasts one month, Muslims do not eat between the hours of sunrise and sunset, consuming most of their food in the evening.
Blood samples taken before, during and after Ramadan show that skipping meals has a beneficial effect on heart-disease risk by reducing cholesterol and the stickiness of the blood caused by platelet aggregation that leads to blood clots.
Dr Mattson is carrying out the first controlled study of what is, in effect, the Warrior diet. Two groups of adults of normal weight, aged between 40 and 50, get identical diets based on three meals that reflect what the average healthy American eats. One group, however, is required to eat all three meals at a sitting.
Both groups attend the National Institute of Aging every afternoon. The first group are served a three-course evening meal, which might consist of soup and bread, roast chicken and vegetables followed by fruit. They can eat all they want.
They are then given their breakfast and lunch to take home for the next day. This might consist of cereal and bread, sometimes accompanied by sausages or eggs, and for lunch a sandwich, soup or a salad.
The second group are given the same menu but are required to eat dinner, breakfast and lunch all within a couple of hours. Many found the new regime difficult for a few weeks, but after the rules were relaxed, compliance improved.
UK experts are sceptical. Professor Andrew Prentice of the Medical Research Council's international nutrition group at the London School of Hygiene and Tropical Medicine says that interest in the difference between nibblers - who consume their calories through the day - and gorgers, who take them all at once, is long-standing. "Metabolically, it looks as though nibbling is a good way to do things. It reduces the insulin required to dispose of the glucose, which imposes a lesser load on the pancreas, and it improves the lipid profile [fat content] of the blood," he says.
The weakness of the US study, he says, is that by forcing the volunteers to eat all their calories between 4pm and 7pm, crucial psychological factors are affected. "What does it do to appetite?" he asks.
Dr Mattson's results will be available by the end of the year, but anecdotal evidence suggests that it is an effective way to lose weight. "I have had 40 or 50 e-mails from people who eat one meal a day or fast every other day. Most say it was uncomfortable for the first one or two months, but then they have felt great. Some lost a lot of weight. Some have seen changes in their blood glucose and blood pressure that suggest better health."
Dr Mattson is continuing with his own regime. "For 20 years I have not eaten breakfast. Studies have shown it is good for your health to eat breakfast, but it has never appealed. I eat a fairly light lunch and eat most of my food in the early evening."
Need to know
Under-eat during the day
Eat your main meal at night
Provide your body with all essential nutrients (ie vitamins, minerals, EFAs, Amino Acids, Probiotics)
Introduce all tastes, aromas, textures and colours possible in your meals
Detox by minimising animal food intake for a few days
Rotate between days of high protein, high fat and high carbs
Avoid foods containing hormones, pesticides, additives, sugar alcohols, and artificial sweeteners
Do not eat carbs alone