Low-Carb Comes Out Ahead in Diet Comparison

After Two Years, the Results for Low-Carb Are Great

researcher weighing man
Thinkstock Images/Stockbyte/Getty Images

Low-carb? Low-fat? Mediterranean? It seems like every day, there's a new stamp of approval on what diet is best. In a head-to-head match, however, which one would win? A July 2008 study published in the New England Journal of Medicine compared all three.

A group of 322 participants involved in the study were followed closely for two years. On most of the parameters studied, the low-carbohydrate diet was as good as or better than the others.

Let's dig into the most pertinent findings of the study, and my observations about the research and its potential significance for you and other low-carb dieters.

The Study

This study of mostly men in Israel took place over a two-year period. The participants were overweight or mildly obese, with BMIs averaging around 30 to 31. They were randomly assigned to one of three diet groups and given fairly intensive counseling on how to help them follow one of the diets. All study participants worked in the same place, and the workplace cafeteria developed special menus for the three groups that corresponded to the diets they were put on. Offerings were color-coded to make following the assigned diet during the day easy.

The Diets

The three groups followed one of the following diets:

  1. Low Fat Diet: A calorie-restricted diet with a maximum of 30% of the calories coming from fat, 10% from saturated fat, and 300 mg of cholesterol daily. "The participants were counseled to consume low-fat grains, vegetables, fruits, and legumes, and to limit their consumption of additional fats, sweets, and high-fat snacks."
  1. Mediterranean Diet: A calorie-restricted diet with a goal of a maximum of 35% of the calories from fat, including 2 to 3 tablespoons of olive oil and a few nuts each day. The diet was to be "rich in vegetables and low in red meat, with poultry and fish replacing beef and lamb."
  2. Low-Carb Diet: Based on the Atkins Diet, it began at 20 grams of carbohydrate per day, gradually increasing to 120 grams per day "to maintain weight loss." Calories, fat, and protein were not restricted. "However, the participants were counseled to choose vegetarian sources of fat and protein and to avoid trans fat."

    The Results

    • Weight: Of those who finished the 2-year study, the low-carb diet edged out the other two diets, with a sustained average 12 lb. loss compared to 10 lbs. for the Mediterranean diet and 7 lbs. for the low-fat diet. Although the vast majority (86%) of the participants were men, the women had more weight loss on the Mediterranean diet.
    • LDL ("bad") Cholesterol: All three groups showed a similar small drop in LDL cholesterol, though it was not statistically significant. Interestingly, at six months, the low-carb group showed a slight increase in LDL. However, by the end of the study, their results were between the two other groups.
    • HDL ("good") Cholesterol: All groups increased their HDL cholesterol, but the low-carb group had an increase that was significantly greater than the other two.
    • Cholesterol Ratio: Another measurement that has been shown to predict risk of heart disease is the ratio of total cholesterol to HDL cholesterol. The low-carb group showed the greatest improvement in this measure.
    • Triglycerides: The low-fat group showed no significant change in triglycerides over the two years, while the low-carb group showed a significant drop.
    • Inflammation: Measures of C-reactive protein, a marker for inflammation, showed a significant and similar drop in both the Mediterranean and low-carb groups, and no significant change in the low-fat group.
    • Fasting Blood Glucose: The Mediterranean diet showed a drop in fasting blood glucose among people with diabetes, which differed significantly from the slight rise in fasting blood glucose in the low-fat group. The people with diabetes in the low-carb group did not experience a significant change by the end of the study, nor did any of the non-diabetics.
    • Fasting Insulin: All groups showed a decrease, with no significant differences among the groups.
    • Insulin Resistance: One measure of insulin resistance showed more of a decrease in the diabetics in the Mediterranean group than those in the low-fat group, but no other differences were shown.
    • Hemoglobin AIC: In the low-carb group, the H A1C (a measure of average blood glucose) dropped by almost a percentage point in diabetics, the only significant change of any group.

    Observations and Questions

    After studying this research paper and the data presented, I have a few observations and questions:

    • The low-fat diet could not be considered very low in fat, as 30% of the calories were from fat. In fact, the fat intake of subjects on the diet was about the same as before the diet. Although it is the generally accepted standard for a "healthy low-fat diet," advocates of more restrictive low-fat diets correctly point out that this is not a test of the very low-fat diets they advocate. Dr. Dean Ornish is one of these advocates, and the diet he advocates -- for example -- includes, at most, 10% of calories from fat.
    • Likewise, although the low-carb group started out eating a diet very low in carbs, by the six-month mark, they were consuming 40% of their calories from carbohydrates. This is surprising considering they were supposed to be following Dr. Atkins' plan, which instructs the dieter to slowly raise the amount of carbohydrate until weight loss slows to about 1 to 2 lbs. per week. It would appear that they kept raising the amount of carbohydrate until 120 grams was reached, as dictated by the study. It's not surprising, then, that the participants mostly stopped losing weight at about the sixth month. What is more surprising is that they still lost more weight than the calorie-restricted diets at all points along the way, including the 2-year mark. On the other hand, they continued to eat only half of the amount carbohydrate that they ate before started the diet.
    • The other two groups were also eating less carbohydrate than before the study began -- about 50 grams less for the Mediterranean group and 85 grams less for the low-fat group. This was because these groups were consuming fewer calories than before. So, even though the percentage of carbohydrate was the same, they were eating less total carbohydrate. Were any of the changes these participants experienced due to the slight decrease in carbohydrate? We can't say.
    • In the diabetics, the fasting blood glucose was, on average, down more in the Mediterranean diet group, but the Hemoglobin A-1C was down more in the low-carb group. Since both are a measure of blood glucose, why the discrepancy? Well, Hemoglobin A-1C is an indicator of average blood glucose, including after meals. It would seem that, overall, the blood glucose in diabetics was reduced more by the low-carb diet.
    • The low-carb group ate less fiber than any group. Compared to before they started the diet, at 6 months, they were eating 7 grams less. At 24 months, they were eating 10 grams less. Why were they eating less fiber as time went on? It would seem that they were adding low-quality (refined) carbohydrate to their diets. Interestingly, the Mediterranean group only minimally increased their fiber intake, while the low-fat group ate about 5 grams less than before starting the diet (baseline).
    • In all of these changes from baseline, we aren't given the starting amounts, which would have been very helpful in analyzing the study.
    • We are not told why the low-carb group was instructed to seek vegetarian sources of protein and fat, as Atkins was neutral on this point except to seek a balance of fats and avoid trans fats.
    • This study is more well-controlled than most, as participants were provided with intensive education and diet-appropriate lunch food, and measures were taken to check on adherence to the assigned diet.
    • At the very least, it would seem that this two-year study should "tip the balance" as to the question of whether a low-carb diet is healthy and as effective as other options for weight loss. I think that low-carb diets should at least be offered to patients alongside other options.
    • The researchers speculated that monounsaturated fat may have contributed to the positive outcomes in the Mediterranean diet. Others have speculated that the higher fiber level may also be a partial explanation. As a side note, there's no reason why fiber and monounsaturated fat can't be part of a low-carbohydrate diet. Could this be the best of both worlds? Possibly, but more research is indeed necessary to say for certain.

    UPDATE: Results of this study at the 6-year mark!


    Shai, Iris, et al. "Weight Loss with a Low-Carbohydrate, Mediterranean, or Low-Fat Diet." The New England Journal of Medicine. 359:229-241 (July 17, 2008)

    Continue Reading