"Dieting for just eight weeks can reverse your diabetes," the Daily Mail reports.
A small study of 30 people with type 2 diabetes found eight weeks on a very low calorie diet of around 600 to 700 calories a day, followed by a less radical six-month weight control diet, led to significant improvement in blood glucose levels in 12 people.
These findings are interesting: type 2 diabetes tends to be seen as a long-term condition that often gets worse over time or, at best, can be kept in check by medication, rather than one that could be reversed.
Tests found the 12 people had glucose levels below the usual cutoff for diabetes, measured after they switched to the weight control diet, which lasted over the next six months.
People who saw their glucose levels return to normal tended to be younger and have had diabetes for a shorter period.
Although the results are encouraging, the study did not compare a low-calorie diet with other treatments.
Another practical consideration is the issue of compliance. The study was self-selecting in that people responded to an advert, which suggests they were highly motivated to lose weight.
Whether the general population of people with type 2 diabetes would stick to a very low calorie diet is uncertain.
We now need bigger, longer-term studies to find out how feasible this is as a treatment approach for more people with type 2 diabetes.
The study was carried out by researchers from Newcastle University, the University of Glasgow, and Lagos University.
It was funded by the National Institute of Health Research and Novo Nordisk, a company that makes diabetes drugs.
The study was published in the peer-reviewed journal, Diabetes Care. Nestlé UK provided the nutritional drinks for the very low calorie diet, but it's reported they had no other input into the research.
The Daily Mail, the Daily Mirror, The Sun and the Daily Express all covered the study uncritically, saying the diet – described by some as a "crash" diet – completely cured diabetes.
While the papers state that only 12 of the 30 trial participants saw their glucose levels return to normal, the Daily Mirror and Daily Mail only said this several paragraphs down in the story.
The Daily Mail suggested the diet could "eradicate" type 2 diabetes, which is highly unlikely given it only worked for less than half the people (40%) in the study.
This was an uncontrolled, non-randomised clinical trial looking at changes in people's test results from the start to the end of the study.
This type of study is useful to show whether a treatment might work in ideal conditions, but doesn't give us a true picture of how it might perform in the general population.
Researchers recruited 30 volunteers with type 2 diabetes. After tests, they gave them an eight-week very low calorie diet, which mostly consisted of three ready-made drinks (diet shakes) a day and non-starchy vegetables. Total energy intake was between 624 and 700 calories.
People then gradually switched to a normal diet, albeit one that was strictly controlled, to make sure they didn't take in more calories than they expended.
People had further tests and remained on the weight control diet for another six months. At the end of that time, researchers looked at glucose levels to see if anyone had levels below the threshold used to diagnose type 2 diabetes.
The tests people underwent looked at the following:
People were also weighed and measured.
The measurements were taken at the start of the study, as soon as people had fully returned to a weight control diet after the very low calorie diet, and again at the end of the study, after six months.
Researchers wanted to know how the results correlated to see what was important in reducing participants' average insulin levels to normal.
Twelve of the 30 people involved in the study had normal average insulin levels – below an HbA1c of 7mmol/L – after the very low calorie diet. All of them still had normal insulin levels after six months.
The average weight across the whole group dropped from 98kg at the start to 84.7kg at the end of six months. Weight loss was similar between those whose glucose levels returned to normal and those who still had diabetes at the end of the study.
People who were able to return to normal glucose levels tended to be younger (average age 52 versus 60) and had diabetes for a shorter time (average 3.8 years compared with 9.8 years), although some people who'd had diabetes for more than 10 years were able to return to normal glucose levels.
The researchers said "responders" – people who got their glucose levels back to normal – had lower glucose levels at the start of the study and more insulin in their bloodstream, indicating that the ability of the beta cells to produce insulin was better.
The fatty content of both the liver and the pancreas decreased in both responders and people who continued to have diabetes.
The researchers said their results suggest weight loss through the very low calorie diet removed excess fat from the pancreas, and that in responders this allowed beta cells to return to producing normal levels of insulin in response to glucose.
The researchers said: "The present data confirm reversal of type two diabetes mellitus for at least six months in those who achieve non-diabetic plasma glucose after VLCD (very low calorie diet),". However, they question whether "truly long-term reversal" can be achieved by treating people in the community.
They concluded that type 2 diabetes "can now be understood to be a metabolic syndrome potentially reversible by substantial weight loss", although "not all people with type 2 diabetes will be willing to make the changes necessary".
The study points to the possibility that some people with type 2 diabetes may be able to be treated with diet alone if they are able to lose enough excess weight – and keep it off.
However, the results we have are from a small group of highly motivated volunteers, so we don't know how many people would be able to follow the diet and keep the weight off afterwards.
An intake of 700 calories a day is around a third of the recommended intake for a woman (2,000 calories) and around a quarter of the intake for a man (2,500 calories). Even the most committed dieter may find it hard to stick to these limits.
Even within this group, one participant was excluded from the study after week one of the very low calorie diet for not meeting the weight loss target of 3.8% body weight.
This treatment is not likely to work for many people with diabetes who have already tried and failed to lose weight.
We need to see properly controlled randomised studies of large groups of people, with follow-up for more than a year, to know whether this programme is a feasible treatment for many people with type 2 diabetes.
The science behind the study is interesting. The researchers say they may have discovered a "personal fat threshold" where fat can be stored around the body, but above a certain level it is deposited in the liver and pancreas, where it causes damage and can prevent the pancreas producing insulin properly.
If this finding is confirmed by further research, it might have implications for other conditions, such as fatty liver disease, as well as diabetes.
Very low calorie diets have been shown to be successful if people also get the right medical advice and it's followed by a strict weight control diet.
The diet used in this study was designed to ensure people got all the nutrients they needed, while drastically cutting down calories to around 700 calories a day.
Check with your GP or the doctor in charge of your diabetes care before trying such a drastic diet. There's no point losing a lot of weight on a crash diet if you put it straight back on afterwards – and the yo-yo effect on your weight could have health implications.
The NHS Choices weight loss plan offers a sustainable way to lose weight at a steady rate through a combination of diet and exercise.