Many hopeful dieters find themselves frustrated when the diet that let their friend drop 50 pounds leaves them at the same weight.
A prominent diet researcher takes aim at this frustration with a new study, which he says is a major step towards personalized dieting.
Others say he does not have the numbers to back up his idea.
David Ludwig, director of the Obesity Program at Children's Hospital in Boston, and his colleagues monitored the weight loss of 73 young adults over the course of 18 months to see if they would do better on a low-fat diet or what is known as a low-glycemic load diet.
The results are published in the current issue of the Journal of the American Medical Association.
While dieters did about equally well on both diets, those who produce a lot of insulin after eating did markedly better on the low-glycemic load diet, losing, on average, 5 percent of their body weight versus the 1 percent lost by those on the low-fat diet.
Ludwig says these findings challenge the one-diet-fits-all concept.
"The key question in the field of obesity is why some can do well on conventional weight loss diets and others do poorly on the same diets," he said.
Ludwig said that potential candidates for the low-glycemic diet will usually be "apple-shaped," with fat around their midsections, rather than "pear-shaped" people who tend to collect fat around their hips and thighs. A blood test for insulin levels can tell for sure.
By taking in fewer heavily processed carbohydrates, apple-shaped dieters can avoid their bodies' buildup of insulin to digest those foods -- a buildup that seems to drive weight gain, Ludwig said.
A low-glycemic load diet includes fruits, vegetables, beans, nuts and whole grains, and discourages processed sugars and starches like white breads, white rice and potato products.
At the same time, it allows for dairy products, even full-fat ones, and urges people to look at the types of fats they consume (avoiding saturated fats) rather than restrict calorie intake.
"[The low-glycemic diet] is very similar to many traditional diets," said Ludwig, who endorses the low-glycemic load diet as a compromise between high-fat, low carbohydrate diets like Atkins and diets which promote carbohydrates while eliminating fats. "We believe that a low-glycemic diet serves as a perfect compromise," he said.
Despite the tests for insulin, Ludwig emphasized that the weight-loss research was unrelated to diabetes, and diabetics were excluded from the study.
"This isn't a very potent diet program," said George Blackburn, director of the Center for the Study of Nutrition Medicine at Harvard Medical School.
"Any program that stops you from gaining weight is a healthy program, but to justify the effort you're putting in, as in this program, you'd like to get a 5 to 10 percent weight loss."
Blackburn cites Ludwig's past research in pointing out the importance of fruits, vegetables, legumes and grains that aren't heavily processed. But he also thinks that the success of the low-glycemic and low-fat diets relied heavily on the increased intake of fiber and protein, and recommends cutting fat as well as processed carbohydrates in order to lower cholesterol -- a benefit the low-fat diet had that the low-glycemic one did not.
"There was good in both of those diets," Blackburn said.
Ludwig defended the weight loss, saying that, unlike most weight-loss studies, his numbers included people who could not comply with the diet and left the study.
"Among people who complied with the diet and stayed in the study, the differences were even greater," said Ludwig, citing an average weight loss close to 10 percent for people who stuck with the low-glycemic diet for 18 months.
Ludwig also emphasized that people on the diet were able to maintain weight loss after 18 months -- a time when most dieters have returned to their original weight.
"It may be less fast, but it's more sustainable," he said.
"David Ludwig is an international authority in this area and does high quality work. His bias, however, creeps into this report," said James W. Anderson, a professor of medicine and clinical nutrition at the University of Kentucky.
Anderson said he feels the study was not properly designed to back up Ludwig's current claims.
Specifically, Anderson believes the number of dieters in the study to be too small to determine that we have reached a time for personalized dieting.
Of the 73 original participants, only 35 completed the study and had blood insulin levels available for analysis (51 completed the study overall). Those 35 were further spread into four separate diet groups.
"You don't design a study with nine subjects in one group," Anderson said.
"More needs to be done, but this is a step in the right direction to identify what dietary approaches will work with one patient instead of another," said Lona Sandon, an assistant professor of clinical nutrition at the University of Texas Southwestern Medical Center.
Sandon feels Ludwig's study represents a valuable trend in dieting toward finding different diets to fit different patients.
Despite the small size of the study, Sandon felt enough past research had shown similar enough results to justify Ludwig's conclusions, and she is considering advising certain patients to undergo blood insulin tests.
Ludwig was equally sure of his results.
"[There's] no reason to believe the findings won't be representative," he said. "We have a great deal of confidence that these findings are real."