Most people find most diets don't work for them. That's why the diet industry is so lucrative, with so many disenchanted dieters still willing to try the latest new plan in their pursuit of weight loss.

Yet a team of Israeli scientists may have found a "miracle" diet that really does work for everyone.

In a remarkable new study, they have shown that people's bodies react very differently to the same foods and this may be why some people stay stubbornly heavy however healthily they eat, while others can consume all sorts of junk food without getting fat.

For instance, it turns out that some people can eat croissants without a problem, and yet they needed to go very easy on apricots; others were fine - and even lost weight - eating ice cream as part of their diet, but tomatoes were bad news.


What this study suggests is that the standard dietary advice - that some foods are fattening and some aren't - is wrong. Instead, the foods that make someone else slimmer and healthier aren't necessarily the "good" foods for you.

These remarkable findings by researchers at the Weizmann Institute in Tel Aviv, and published in the prestigious journal Cell, are a real game-changer.

The key is blood sugar - specifically the different effect a food can have on the blood sugar levels of different people.

We all need blood sugar (or glucose, to give it its medical name) for energy. Most of it comes from carbohydrates found in foods such as bread, rice, apples and peanuts.

Our digestion breaks down the carbohydrate into sugar, which is released into the bloodstream. After eating, a slight rise in our blood sugar - known as the post-prandial glucose response (PPGR) - is normal, but large and regular increases raise your risk of weight gain and obesity, and disorders such as type 2 diabetes, high blood pressure and heart disease.

Thirty years ago, Canadian scientists developed a way of measuring a food's effect on blood sugar. Known as the Glycaemic Index (GI), this system rated foods according to how much they will push up blood sugar after eating.

This was later turned into a diet based on the idea that everyone's blood glucose response to a food will be much the same.

So high-GI foods, such as white bread or croissants, will produce a rapid spike, while low-GI ones, such as vegetables that release carbs slowly, will keep the post-prandial glucose response down.

Today, healthy eating advice often recommends low-GI foods. But the new research turns this on its head.

"The take-home message from this study is that the GI approach may not suit everybody," says Professor Eran Segal, a computational and systems biologist who led the study. "If a diet didn't work for you, it may not be your fault, just that you were on the wrong one."


His co-author Dr Eran Elinav, an immunologist, adds: "Clinicians believe diets fail because people don't follow instructions properly. But now it seems likely the problem is many people have been getting advice that was wrong for them."

For their study, known as the Personalised Nutrition Project, the researchers collected an impressive amount of data about the minute-by-minute effect of food on the blood sugar levels of 800 volunteers.

Each volunteer was monitored with an unobtrusive hi-tech device, about the size of a watch, that was taped to their stomachs to measure the changes in their blood sugar, round the clock, for a week.

By the end, the researchers had information on the response to 46,898 meals and recorded more than 1.5 million blood sugar measurements from a total of 4,435 days. Such a detailed analysis had never before been done.

The scientists compared this data against the detailed daily food and activity diaries all the volunteers kept to discover which foods pushed up their blood glucose level or had little or no effect.

"Many of the volunteers discovered they responded very differently to foods with the same GI," explains Dr Elinav. For instance, the response of four different volunteers to a daily slice of bread varied widely.

"And two participants had completely opposite responses to a banana and a cookie: the glucose response of one shot up with a cookie but stayed flat after a banana, and vice versa."

One of the volunteers was Neeta Zur, a 51-year-old specialist in medical and biological products.

She is also a WeightWatchers instructor with an extensive knowledge of diets, but when she got her results there were plenty of surprises.

"I was delighted to discover a hot caramel sundae - a treat WeightWatchers recommends - is healthy for me because it does nothing to my blood sugar providing it comes with nuts,' she says.

"Sadly, the yoghurt with granola that I often had as a healthy breakfast was a no-no for me because it made my blood sugar spike, but a vegan cake with no eggs was fine.

"I learned that although white rice pushed up my blood sugar, if I added cashew nuts, it had no effect. But nuts didn't always lower my glucose response; bananas, often excluded from diets because they have a high GI, were "good" foods for me, unless I mixed them with sharon fruit plus walnuts, when my blood sugar shot up."

A major problem with studies involving food and activity diaries is that people are notoriously bad at keeping them. But with this study, there was "a big incentive", says Neeta.

"We were told that at the end we would get a list of foods that were likely to keep us healthy and ones to avoid, and that the more accurate information we provided the better feedback we would get." And when Neeta was put on her individualised diet, she lost one kilogram after a  week.

It isn't just the GI that looks as if it needs a rethink. For years there has been a heated debate over official advice from authorities such as the NHS and Diabetes UK that healthy eating involves a low-fat, high-carb diet.

Academics and researchers point to evidence that a high-fat, low-carb diet is better for weight loss and improving diabetes. Now it looks as though both camps may have a point - it just depends on your individual response.

In the new study, heavy carb consumption certainly pushed up glucose levels in general, but there was plenty of individual variation.

Some of the volunteers, dubbed "carb-sensitive", had a higher blood sugar reaction the more carbs they ate, but there were also "carb-insensitives" whose blood sugar could shoot up even on a low-carb meal.

"Tomatoes in particular seemed to produce this counter-intuitive response," says Professor Segal. "They pushed up the glucose response, even though they contain relatively few carbs."

So if you're a "carb-insensitive" dieter, eating lots of low-calorie healthy tomatoes could actually drive up your blood sugar and weight gain.

The same thing happened with fat. In general, adding fat to meals lowered the blood sugar response - but not for everyone.

For example, one standardised daily meal all 800 volunteers ate included bread on its own or bread with butter. Three-quarters produced less of a blood sugar response to bread plus butter - however, the rest had a lower response only with the plain bread.

Professor Segal adds: "In the future, knowing your glucose responses should be as common as knowing your blood pressure or cholesterol levels."

If the results of this study are supported, there is likely to be a big shake-up in the way the glucose response is used to diagnose diabetes.

Currently it is tested using the glucose tolerance test, where the patient is given a glucose drink, then their blood sugar levels are measured shortly afterwards.

But as Professor Segal points out: "People don't usually eat glucose on its own, they eat real food, and we now know that individual foods can have very different effects on the glucose response of different people. So to test it properly and then give the right dietary advice, you need to how they respond to their normal diet."

In theory, you could work out your own body's response to certain foods using a DIY glucose monitor. Realistically, you could only do it for a fairly limited number of foods and 'you'd need to know how to analyse the results properly', explains Professor Segal.

That's why the team has developed a computer algorithm to help people work out their "good" and "bad" foods.

This was based on all the data from the seven-day monitoring, plus the detailed daily diaries on diet, exercise and sleep - and one more vital and surprising bit of information: what was going on in their guts.

It turned out that gut bacteria can have a direct effect on the size of our post-prandial glucose response to foods. What's more, certain foods and behaviours can encourage the helpful bacteria, while others increase the numbers of the harmful ones.

"For example, we found out that if you were over 50 and more than 30 per cent of your gut bacteria were a type called firmicutes (vital for the body's various systems to communicate), then eating meals with higher levels of fat would bring down your blood sugar response. But if you had a smaller firmicutes population, below 30 per cent, then more fat would push it up."

(Foods rich in fibre help to keep firmicutes in a healthy balance with the other gut bacteria.)

In the second part of the study, the Israeli researchers tested their algorithm by getting it to calculate two sets of personalised menus - a good and a bad one - for groups of volunteers.

The results were pretty surprising, says Professor Segal. "Some of the 'good' diets included alcohol, along with chocolate and ice cream - in moderation."


Here are examples of two menus, one "good" and one "bad", which were designed for a specific volunteer - we'll call him Dave. Can you guess which is which?

Menu 1: Muesli for breakfast; sushi for lunch; corn-on-the-cob and nuts for dinner; marzipan as a snack.

Menu 2: Eggs with bread and coffee for breakfast; hummus and pitta for lunch; vegetables with tofu for dinner; ice cream as a snack.

Menu 2 - with the ice cream - was the "good" one, surprisingly, for Dave, and Menu 1 - with the "healthy" sushi - was "bad".

"On the bad diets the blood glucose reached really abnormal levels, but on the good diets they normalised to healthy ranges," explains Professor Segal.

"The next step is a trial to see if "good" diets personalised for diabetics can improve their post-prandial glucose repsonse and reduce the need for drugs," says Professor Segal.

A private company now has the licence to make the computer algorithm available to the public, but that's a couple of years away.

In the meantime this research is more evidence that when yet another diet fails, it is not necessarily your fault. And when one does work, it is may be because it suits your biochemistry.

- Daily Mail