The July 17, 2008 issue of the New England Journal of Medicine published the results of the Dietary Intervention Randomized Control Trial (DIRECT) study comparing the effects of low-carb, Mediterranean, and low-fat diets on weight loss (1). Predictably, this paper elicited a flurry of commentaries. You can find some of the headlines, and Tom Venuto′s take on the study, on his BURN THE FAT blog (2).

Trial details

A total of 322 participants (mean age 52 years, 86% men) enrolled in the two year trial. All were either moderately obese (BMI > 27), had type 2 diabetes or suffered from coronary heart disease. They were randomly assigned to follow one of three dietary protocols:

• A low-fat calorie-restricted diet. Daily target energy intakes were 1800 and 1500 kcal for men and women, respectively, with at most 30% of calories from fat and one third of that from saturated fat. The participants were advised to eat low-fat grains, vegetables, fruits and legumes.
• A Mediterranean diet, restricted to the same daily calorie limits, with no more than 35% of calories from fat. This diet was rich in vegetables, poultry and fish replaced red meat, and olive oil and nuts were the major sources of added fat.
• A low-carb diet. This group aimed for a daily carbohydrate intake of no more than 120g. There were no restrictions on total calories, protein or fat.

The two year study was about as well controlled as it is possible with a dietary intervention. It was carried out at an Israeli research centre with an on-site medical clinic. All participants were employees at that centre, and all ate their main meal of the day at the centre′s cafeteria. The cafeteria staff was advised by a dietitian, and suitable dishes were prepared for each of the three diet groups. In addition, a dietitian met with each group 18 times for 90 minute sessions. Total daily food and calorie intakes were estimated from food questionnaires.

Results

Of the 322 original participants, 272 completed the trial: 90.4% in the low-fat, 85.3% in the Mediterranean, and 78% in the low-carb group. The following weight-related changes were observed:

• Among all participants completing the trial, the mean weight losses were 3.3 kg in the low-fat, 4.6 kg in the Mediterranean and 5.5 kg in the low-carb group.
• Among the women, the mean weight losses were 0.1 kg for the low-fat group, 6.2 kg for the Mediterranean diet group, and 2.4 kg for the low-carb group.
• The decreases in BMI were 1.0 in the low-fat group, and 1.5 in the other two goups.
• Average waist sizes decreased by 2.8 cm in the low-fat group, 3.5 cm in the Mediterranean diet group, and 3.8 cm in the low-carb group.
• Levels of circulating leptin (a reflection of body fat) decreased significantly in all diet groups.

Changes in biomarkers:

• The greatest average drop in blood pressure, 5.5/2.2 mm Hg, was observed in the Mediterranean group.
• The greatest average increase in HDL levels, 8.4 mg/dL or 0.22 mmol/L, occurred in the low-carb group. LDL levels remained unchanged in all three groups.
• The level of high-sensitivity C-reactive protein decreased significantly only in the Mediterranean (21%) and low-carb (29%) groups.
• Among diabetic subjects, only the Mediterranean diet group had a decrease in fasting glucose levels. There was no corresponding change in non-diabetic subjects.
• Fasting insulin levels dropped in all three groups, and in diabetics as well as in non-diabetics.
• Among diabetics, glycated hemoglobin levels (a measure of blood glucose levels) dropped significantly only in the low-carb group.

Maximum weight loss was reached after about six months, followed by a partial rebound and a plateau. On the other hand, biomarkers like HDL, C-reactive protein and fasting insulin levels continued to improve throughout the two year trial period. As the authors point out, a healthy diet has benefits beyond weight loss.

Conclusions

Here is how the authors summarize the results of their clinical trial:

"Mediterranean and low-carbohydrate diets may be effective alternatives to low-fat diets. 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." (1).

I don′t see any great difference between the outcomes of the three dietary protocols. This might be partly due to common calorie limits in the low-fat and Mediterranean diet groups. I suspect though that the food intakes in the three groups weren′t quite as different as the investigators believed — true low-fat and low-carb diets are difficult to follow.

Low-fat diets are tasteless — just compare low-fat to regular cheese. They are almost by definition high-carb diets, and excess glucose is simply converted to saturated fatty acids. If you don′t eat fat, you make it from the extra carbs you eat! With a low-fat diet you deprive yourself for nothing.

Low-carb diets, on the other hand, are difficult to follow because so many of our dietary staples are starchy foods — bread, pasta, rice, potatoes, breakfast cereals (not to mention desserts). It′s hard to devise meal plans that don′t contain any of these foods. In any event, low-carb eating doesn′t even make sense. Our brains and muscles need glucose, and glucose is the only fuel that red blood cells can use. We need carbohydrates, but of the low-glycemic variety.

If this study proved anything, it is that real food — the Mediterranean eating plan — works just as well as diet fads like low-fat and low-carb schemes.

References

1. Shai I et al. Weight loss with a low-carbohydrate, Mediterranean, or low-fat diet. N Engl J Med 2008;359(3):229-241. [Full Text]
2. Tom Venuto. What the new "low carb" study REALLY says. [Blog]