Which diets are most effective in the long-term to help patients lose weight?

The low-carbohydrate diets, low-fat diets, and Mediterranean diet all help patients lose weight. Although they modestly improve lipid profiles and fasting insulin levels, there were few changes in the medical treatment of participants as a result of their dieting. If patients want to lose weight and lower their cholesterol, they should simply choose a diet they can stick with. (LOE = 1b)

Shai I, Schwarzfuchs D, Henkin Y, et al, for the Dietary Intervention Randomized Controlled Trial (DIRECT) Group. Weight loss with a low-carbohydrate, Mediterranean, or low fat diet. N Engl J Med 2008;359(3):229-241.  [PMID:18635428]

Randomized controlled trial (single-blinded)

Government

Uncertain

Population-based

Most organizations still recommend a low-fat, calorie-restricted diet for weight loss and control of lipids and blood sugar. Recently, the Atkins diet (low-carbohydrate, no calorie restriction) and the Mediterranean diet (moderate monounsaturated fats, restricted calories, poultry and fish instead of beef and lamb) have become popular. In this study, 322 obese workers (86% male; mean body mass index = 31 kg/m2; mean age = 52 years) at an Israeli research station were randomized to consume 1 of the 3 diets. The main meal of the day in Israel is lunch and was prepared for them at the company cafeteria; participants and their spouses also received education from a dietician on how to choose and prepare meals at home. Groups were similar at baseline, and patients were followed up for 2 years. Compliance was good, with 95.4% compliancy for 1 year, and 84.6% for the full 2 years. There were 10 withdrawals in the low-fat diet group, 16 in the Mediterranean diet group, and 20 in the low-carbohydrate diet group (P = .04). All groups lost weight: 2.9 kg (6.4 lbs) in the low-fat group, 4.4 kg (9.7 lbs) in the Mediterranean group, and 4.7 kg (10.3 lbs) in the low-carbohydrate group. For the small number of women in the study, the Mediterranean diet resulted in the most weight loss and the low-fat diet resulted in very little weight change (-6.4 kg vs -0.1 kg, respectively). For men, the mean weight loss for all 3 diets ranged from 3.4 kg to 4.9 kg. All groups reduced their blood pressure and modestly improved their lipid profiles. The low-carbohydrate diet was more likely to lower fasting insulin levels, which is of uncertain clinical significance. The effect on blood sugar in pateitns with diabetes was variable. Perhaps most tellingly, there were few changes in the patients' medications for hyperlipidemia, blood pressure, or blood sugar over the 2-year study.