"Increasingly trendy low-carbohydrate diets are no more effective than traditional low-fat diets," reports The Daily Telegraph.
Researchers in California found people lost an average 5 to 6kg (11 to 13lbs) over 12 months, whether they were assigned to a low-fat or low-carb diet.
The researchers found no evidence that some people are genetically adapted to respond better to one type of diet than another. Previous studies suggested that certain gene variations were linked to diet response, meaning some people lost more weight with a low-fat diet, while others lost more weight with a low-carb diet.
People in this study were tested for a range of genetic variations previously identified with diet response, but were no more likely to lose weight if assigned the "right" diet for their genotype. Researchers also looked at insulin response, also previously linked to diet response, but found this didn't affect which diet worked best either.
Within the diet groups, some lost more weight than others, with a range of weight loss from losing 30kg (4 stone 10lbs) to gaining 10kg (1 stone 8lbs). However, this did not seem linked to genetic variation or type of diet followed. The researchers conclude that other factors must be responsible for the difference in weight loss seen.
The study was carried out by researchers from Stanford University School of Medicine in the US. It was funded by the US National Institutes of Health and the Stanford Clinical and Translational Science Award and published in the peer-reviewed Journal of the American Medical Association.
The study was covered reasonably accurately in the UK media. Most reports focused on the low-fat versus low-carb aspect of the study, paying less attention to the findings regarding genetic type or insulin.
The Guardian reported that: "Participants who ate the most vegetables and consumed the fewest processed foods, sugary drinks and unhealthy fats lost the most weight." While that may be true, that information is not presented in the study and the source of the claim is unclear.
This was a randomised clinical trial, which is usually the best type of study to see which of two interventions (in this case diets) works best.
Researchers recruited 609 adults, aged 18 to 50 with a body mass index (BMI) of between 28 and 40, from around San Francisco. The participants underwent a variety of measurements and tests, including their weight, BMI, insulin response to glucose, and tests for genetic variants linked to diet response. Researchers then randomly assigned them to either a low-fat or low-carb diet.
Over the course of a year, participants were invited to 22 group sessions to help them stick to their diet. The sessions were run by registered dietitians. All participants were encouraged to eat healthily, with plenty of vegetables and fibre, while avoiding sugar and refined grains.
The low-fat group was initially encouraged to cut down to 20g a day of fat, and the low-carb group to 20g a day of carbohydrates. They did this for 8 weeks, then gradually increased the amounts to a level they thought they could maintain.
People were weighed and measured after 3 months, 6 months and 12 months, and filled out sporadic food questionnaires to see how closely they were following the diet.
Other factors measured included:
The last 2 measurements were not taken for the first 78 people in the study, because funding only became available for this later in the study.
Researchers looked at:
Average weight loss for the 2 diets was very similar:
People with genetic variations linked to low-fat diet response were no more likely to lose weight on the low-fat diet than the low-carb diet. The same was true in reverse – people with genetic variations linked to low-carb diet response were no more likely to lose weight on the low-carb diet than the low-fat diet.
Poor insulin response had previously been thought to indicate that people would benefit from a low-carb diet, but again, in this study, people with poor insulin response were no more likely to lose weight on a low-carb diet than a low-fat diet.
The dietary questionnaires showed that people stuck to their diet types, with big differences in the proportion of carbohydrates and fat consumed between the groups. Although they had not been instructed to reduce calories, both groups cut the calories they ate by around 500 to 600 a day.
The only difference between the groups was in lipid levels in the blood. Those on the low-fat diet improved their "bad" LDL cholesterol more, while the low-carb group improved their "good" HDL cholesterol and reduced their triglyceride levels more.
The researchers say their results showed "no significant difference in weight change between a healthy low-fat diet vs a healthy low-carbohydrate diet" and that "neither of the two hypothesised predisposing factors was helpful in identifying which diet was better for whom".
They say the differences from previous study findings may be because this study stressed the importance of eating healthy whole foods, rather than eating any food so long as it was either low-fat or low-carb. "Both diet groups in the current study were instructed to minimise or eliminate refined grains and added sugars and maximise intake of vegetables", they said.
Arguments over whether reducing fat or carbohydrate is more important for weight loss have raged for years. This well-conducted study suggests that both can work well, so long as people stick to them, eat less overall, and eat a healthy diet with plenty of vegetables and little sugar or refined grain.
The theory that some diets work better for some people may still hold true – but not for the reasons previously suggested. It may be that some people find low-fat or low-carb diets easier to stick to, because of personal preference. Or there may be genetic variations at work – just not the ones that have been identified as potential explanations so far.
The study was large and well run, but has a few limitations:
The results may be more relevant to populations with relatively high education levels and the resources to buy good quality food, as in this study.
The factors considered in the study, such as the insulin test used (INS-30) and the genetic variations identified, may not be the right ones to use – although they were considered the best at the time of the study.
Not everyone in the study had the full range of measurements taken, although this is unlikely to have affected the main results.
The study only had limited power to show whether insulin or genetic variation directly affected the results. To give more reliable results, the study would need to randomise people according to their genetic or insulin status.
Find out more about how to lose weight.