"Vegetable diet will beat diabetes: Meat-free lifestyle cures killer disease," is the typically overblown headline in the Daily Express.
But researchers actually found a vegetarian diet led to a quite modest fall in only one measure of blood glucose called HbA1C, a measure of blood glucose control.
The paper reports on a systematic review which combined the results of six trials that involved 255 people with type 2 diabetes. They examined whether vegetarian or vegan diets improved blood glucose control compared with a control diet.
Overall, the pooled results of five of these trials found a vegetarian or vegan diet reduced HbA1c by 0.39%. There was no significant effect on fasting glucose levels, an assessment of how efficiently the body can process glucose in the short term.
This slight reduction in HbA1c is no cure. As the researchers themselves pointed out, the reduction is less than you would expect if a patient was being treated with the drug of choice for type 2 diabetes, metformin.
This review also has various important limitations, including the variable design and quality of the six trials included. So, it does not prove that a vegetarian or vegan diet is better for a person with type 2 diabetes, and any media claims of a "cure" for the condition are entirely baseless.
The study was carried out by researchers from Keio University in Japan and The George Washington University School of Medicine in the US.
Funding was provided by the Japan Society for the Promotion of Science and the Nestlé Nutrition Council, Japan.
One of the co-authors declared a non-financial conflict of interest. This author serves as president of the Physicians Committee for Responsible Medicine, without financial compensation.
This organisation is described in the publication as one that, "promotes the use of low-fat, plant-based diets and discourages the use of animal-derived, fatty, and sugary foods". This represents a potential conflict of interest in the interpretation of the results.
The Daily Express' coverage of the study is accurate and contains some useful background information, so it is frustrating that its headline is totally misleading, especially as it was on the front page.
In fact, this review of studies found vegetarian or vegan diets caused a slight reduction in HbA1c compared with non-vegetarian diets. This is not a cure in any sense of the word.
The current thinking is that there is no such thing as a cure for type 2 diabetes. The condition can be successfully managed, but not cured.
The study is also only applicable to type 2 diabetes, so the headlines do not apply to type 1 diabetes.
As the researchers say, previous research has suggested a link between a vegetarian diet and improved blood sugar control, but the relationship is not well established.
As an interesting aside, the researchers highlight how diabetes levels were found to be lower in Seventh-day Adventists, a Protestant Christian denomination whose followers are encouraged to adopt a vegetarian diet.
This review aimed to examine this grey area. A systematic review and meta-analysis of randomised controlled trials is the best way of examining the evidence to date that has assessed this question.
The researchers searched a number of literature databases (from their inception to 2013) to identify published clinical trials examining the effects of a vegetarian, vegan or omnivorous diet on blood sugar control in people with type 2 diabetes who were over the age of 20.
A vegetarian diet was defined as one excluding meat, poultry and fish, while a vegan diet excluded all animal products.
Eligible trials had an intervention duration of at least four weeks and examined the main outcome of changes in HbA1c.
This gives an indication of blood sugar control in the longer term, as it indicates the amount of sugar being carried by red blood cells, which have a lifespan of around three months. Change in fasting blood sugar measures was a secondary outcome.
In an added effort to find all relevant information for the review, the research team scoured the reference lists of all articles they found from the search of electronic databases, and also contacted research experts for additional material.
The researchers assessed the quality of the studies included, and pooled studies calculating the average difference in HbA1c and fasting blood sugar between vegetarian or vegan and comparison diets.
A total of six trials met the inclusion criteria, involving 255 people with type 2 diabetes with an average age of 52-and-a-half. The average trial duration was 23.7 weeks, or about six months.
Five of the studies examined vegan diets and one studied vegetarian diets. Four trials were conducted in the US, one in Brazil and one in the Czech Republic.
Of the six studies, three were randomised controlled trials, one was a cluster randomised controlled trial, and two were non-randomised controlled trials.
In the pooled analysis of five trials, the vegetarian or vegan diet was associated with a significant reduction in HbA1c (-0.39%, 95% confidence interval [CI] -0.62 to -0.15) compared with omnivorous control diets.
But the pooled analysis of four trials did not find a statistically significant reduction in fasting blood sugar: the average difference with the vegetarian or vegan diet compared with control was -0.36 mmol/L, 95% CI -1.04 to 0.32.
Compared with control, the vegetarian or vegan diets were also associated with significant reductions in the amount of total energy the diet provided, either through carbohydrate, protein, total fat, cholesterol and fibre.
The researchers concluded that, "Consumption of vegetarian diets is associated with improved [blood glucose] control in type 2 diabetes."
This systematic review has identified six trials assessing whether vegetarian or vegan diets improve blood sugar control in type 2 diabetes compared with control.
It found the vegetarian or vegan diet gave significant improvement in one measure of blood sugar control (HbA1c), but not in another (fasting blood glucose).
However, there are some important limitations to consider before we can categorically conclude that people with type 2 diabetes should switch to a meat and fish-free diet:
The pooled results of five trials found a vegetarian or vegan diet was associated with a 0.39% reduction in HbA1c, but we don't know that this would have made any meaningful clinical difference in diabetes control for the individual.
Overall, although any reduction is likely to be a good thing, the precise benefit would depend on what a person's HbA1c level was to start with.
The target HbA1c is usually set at a level below around 7%, so it may be more useful knowing whether a vegetarian or vegan diet improved the proportion of people achieving their target HbA1c level. The review also found no improvement in fasting blood glucose control.
Despite the publication tending to refer to the intervention diets as vegetarian, they were actually quite varied across the trials.
Four of the trials were described as low-fat vegan, one as lacto-vegetarian (a diet that includes dairy products but not eggs), and one lacto-ovo low-protein (similar to a lacto-vegetarian diet but, as the name suggests, with a focus on low-protein foods).
The researchers included diets described as omnivorous, low fat, "diabetic diet" and those that followed American Diabetic Association guidance.
Overall, this doesn't give a very clear picture of what diets were being compared, which makes it hard to conclude that a particular diet is associated with an improvement in blood sugar control compared with a particular control.
Only three of the six trials studied were true randomised controlled trials. They varied in the duration of the dietary intervention between four and 74 weeks.
Also, only one of the six trials (a controlled trial) is reported to have made any adjustment for potential confounders (sex, baseline HbA1c level and medication). The others report no adjustment.
We also don't know how the trials checked that the diets were being followed as assigned, or of any other intervention or advice that may have been given to the participants alongside the dietary intervention (such as advice about physical activity).
In their assessment of possible publication bias, the researchers observed that smaller trials that found reductions in HbA1c level were perhaps more likely to have been published and therefore included in this review.
Despite this being a systematic review of trials, the total number of participants was still quite small, at only 255. This is a very small number of patients, and it might be unwise to base any firm or generalisable conclusions on such small numbers.
A vegetarian or vegan diet can be a healthy lifestyle choice for a person with type 2 diabetes if it provides balanced nutrition. But such diets can still be high in fat, salt and sugar if this is not controlled carefully.
A healthy diet needs to be combined with regular exercise for people to be able to reap further health benefits, as well as avoiding smoking and only consuming alcohol at or below nationally recommended levels.
Overall, this review does not appear to conclusively prove that a vegetarian or vegan diet is better for a person with type 2 diabetes. It certainly provides no evidence that this diet cures diabetes, as one of the news headlines suggests.
Provided you do your homework, it is possible to eat healthily on a vegetarian or vegan diet. But if you do have type 2 diabetes, we recommend that you talk to the doctor in charge of your care before making any radical changes to your diet.