For years authorities such as the American Heart Association and the US Department of Agriculture have recommended a high carbohydrate, low fat diet for cardiovascular health and weight loss, while Americans have grown ever fatter. Meanwhile Dr Robert Atkins' low carbohydrate "steak and eggs" diet books have sold 10 million copies and led to anecdotal reports of successful weight loss.
No prospective, randomised trials have shown the Atkins diet to be safe and effective—and it might actually be harmful, according to some doctors. But two studies published last week indicate that Dr Atkins, who died in April (17 May, p 1090), was right (New England Journal of Medicine 2003;348:2074-81 and 348:2082-90). The studies found that a high protein, low carbohydrate diet led to greater weight loss and improved blood lipid concentrations than the recommended high carbohydrate, low fat diet in healthy obese people as well as in obese patients with complicating conditions.
In a related article American Heart Association president Dr Robert Bonow and nutrition expert Dr Robert Eckel urged caution (348:2507-8). Eat a balanced diet of fruit, vegetables, whole grains, legumes, and fish, they said.
In one of the studies, a one year, multicentre, randomised controlled trial, 63 healthy obese men and women (mean body mass index 34) were randomly assigned to follow either Dr Atkins' diet or a high carbohydrate diet based on the Department of Agriculture's "food pyramid" (60% of calories from carbohydrates, 25% from fat, and 15% from protein).
Participants following the Atkins diet lost significantly more weight than the other group at three months (P=0.001) and at six months (P=0.02), although at one year the difference was not significant. Decreases in serum triglyceride concentrations and increases in HDL cholesterol were greater in the Atkins diet group. LDL cholesterol concentrations did not differ significantly.
In the other study, a six month trial, 132 severely obese patients (mean body mass index 43) were randomised either to a diet with only 30 grams of carbohydrate a day, similar to the Atkins diet, or to a high carbohydrate, low fat diet, with frequent counselling. Thirty nine per cent of patients had diabetes and 43% had the metabolic syndrome. Mean weight loss in the low carbohydrate diet group was 5.8 (SD 8.6) kg, compared with 1.9 (4.2) kg in the low fat diet group (P for difference=0.002).
Serum triglyceride concentrations decreased by 20% (SD 43%) in the low carbohydrate group, whether or not the patients were taking lipid lowering drugs, compared with 4% (31%) in the low fat group (P=0.001). Total, HDL, and LDL cholesterol concentrations did not differ significantly between the groups.
The dropout rate in both studies was about 40%, not unusual for diet studies.
Dr Gary Foster of the University of Pennsylvania School of Medicine, lead author of the first trial, was surprised at the results. "You would expect an increase in LDL, but we didn't see it. We didn't expect the improvement in HDL and triglycerides to the degree we saw," he said.
Dr Frederick Samaha, cardiologist at the Philadelphia Veterans Affairs Medical Center and lead author of the second study, said he was surprised by the metabolic benefits, especially the decrease in triglycerides.
Co-author Dr Linda Stern said patients probably restricted intake of simple carbohydrates from sweet drinks and snacks but did not reduce intake of complex carbohydrates.