"The more favorable effects on lipids (with the low-carbohydrate diet) and on glycemic control (with the Mediterranean diet) suggest that personal preferences and metabolic considerations might inform individualized tailoring of dietary interventions," Dr. Iris Shai and colleagues write in the New England Journal of Medicine for July 17.

The Dietary Intervention Randomized Controlled Trial (DIRECT) was conducted with the assistance of a dietitian in a workplace setting in Israel between July 2005 and June 2007. The study cohort of 322 individuals was 86% male with mean age of 52 and mean BMI of 31, and included 46 subjects with diabetes and 118 with coronary heart disease.

The low-fat diet restricted energy intake to 1500 kcal/day for women and 1800 kcal/day for men, with 30% of calories from fat.

The Mediterranean diet was rich in vegetables and low in red meat, with calorie restrictions the same as in the low-fat diet and up to 35% of calories from fat, including olive oil and about half a dozen nuts (< 20 g) per day.

The low-carbohydrate, non-restricted-calorie diet allowed 20 g of carbohydrates per day during the first 2 months and after holidays, increasing to a maximum of 120 g/day.

"Although participants actually decreased their total daily calories consumed by a similar amount, net weight loss from the low-fat diet after 2 years was only 6.5 pounds (2.9 kg) compared to 10 pounds (4.4 kg) on the Mediterranean diet, and 10.3 pounds (4.7 kg) on the low-carbohydrate diet," Dr. Shai, at Ben-Gurion University of the Negev in Beer-Sheva, told Reuters Health. "These weight reduction rates are comparable to results from physician-prescribed weight loss medications."

Maximum weight reduction occurred within 6 months, followed by partial rebound and a plateau. Other health outcomes, including blood pressure, physical activity, inflammatory markers and liver enzyme levels, continued to improve throughout the 2-year trial. "This suggests that a healthy diet has beneficial effects beyond weight loss," Dr. Shai said.

The greatest improvements in lipid profiles were observed with the low-carbohydrate diet, whereas the most favorable changes in fasting plasma glucose and insulin levels among diabetics were associated with the Mediterranean diet.

"Clearly, one diet doesn't fit all," Dr. Shai noted.

To identify the diet that is likely to work best for individual patients, she recommends that, after describing the regimens, "physicians should ask questions regarding individual preferences (e.g., whether they have a hard time in counting calories and can try counting carbs), diet history (failures), and metabolic goals."

"Whatever the choice is, the patient should stick with his or her own diet strategy and continue to be followed" by their health care provider or dietitian.