1. Home>
2. Specialties>
3. Cardiology>
4. Summary and Comment

Low-Carb vs. Low-Fat Diets: Randomized Data

Collectively, 2 studies showed modestly greater weight loss with a low-carb diet, but subjects had trouble adhering to both low-carb and low-fat diets.

In 2 randomized trials, researchers addressed the potential merits of much-debated low-carbohydrate diets.

In the first trial, 132 severely obese people (mean body-mass index, 43; 39% with diabetes) were randomized either to a diet that included 30 g/day or less of carbohydrates or to a calorie- and fat-restricted diet (calorie deficit of 500 per day with 30% of total calories from fat). For the 40% of subjects who dropped out before 6 months, baseline weights were used as final weights.

Mean weight loss by 6 months (the primary endpoint) was significantly greater in the low-carb group (–5.7 kg) than in the low-fat group (–1.8 kg), with a similar difference when dropouts' baseline weights were not used. The 2 groups did not differ significantly in 6-month changes in total-, LDL-, or HDL-cholesterol levels, but the low-carb group showed significantly greater 6-month reductions in mean triglyceride level (–20% vs. –4%) and in mean fasting glucose level (–9% vs. –2%). BP changes did not differ significantly between the groups.

In the other trial, 63 nondiabetic subjects who were not severely obese (mean BMI, 34) were randomized either to a low-carb diet (initially, 20 g/day or less of carbohydrates) or to a high-carb, low-fat, low-calorie diet (1200-1500 cal/day for women, 1500-1800 cal/day for men) with 60% of calories from carbohydrates, 25% from fat, and 15% from protein. For the roughly 40% of subjects who dropped out before 12 months, baseline weights were used as final weights.

Although weight loss was significantly greater in the low-carb group than in the low-fat group at 6 months, weight gain in both groups between months 6 and 12 diminished the 12-month benefit in the low-carb group (mean change from baseline, –4.4% [low-carb] vs. –2.5% [low-fat], P=0.3); the between-groups difference was similar when dropouts' baseline weights were not used. The 2 groups did not differ significantly in 12-month changes in total and LDL cholesterol, but the low-carb group showed a significantly greater 12-month reduction in triglyceride level (–28% vs. +1%) and a significantly greater increase in HDL level (+18% vs. +3%). Insulin-sensitivity and BP changes did not differ significantly between the groups.

Comment: Collectively, these studies show modestly greater weight loss with a low-carbohydrate diet than with a low-fat diet. More notably, subjects had trouble adhering to any diet. An editorialist defends the validity of the dropout analyses, but the larger question of whether patients will stay on these regimens remains. Another interesting finding in the second study is the low-carb group's marked HDL increase, despite no significant weight-loss benefit. Overall, these studies suggest a possible benefit with the low-carb approach. Still, the amount of good evidence is inadequate to the task of addressing the burgeoning obesity epidemic.

— Harlan M. Krumholz, MD, SM

Published in Journal Watch Cardiology August 8, 2003

Citation(s):

Samaha FF et al. A low-carbohydrate as compared with a low-fat diet in severe obesity. N Engl J Med 2003 May 22; 348:2074-81.

• Original article (Subscription may be required)
• Medline abstract (Free)

Foster GD et al. A randomized trial of a low-carbohydrate diet for obesity. N Engl J Med 2003 May 22; 348:2082-90.

• Original article (Subscription may be required)
• Medline abstract (Free)

Ware JH. Interpreting incomplete data in studies of diet and weight loss. N Engl J Med 2003 May 22; 348:2136-7.

• Original article (Subscription may be required)
• Medline abstract (Free)

Bonow RO and Eckel RH. Diet, obesity, and cardiovascular risk. N Engl J Med 2003 May 22; 348:2057-8.

• Original article (Subscription may be required)
• Medline abstract (Free)