The Diet Wars Continue: Low Carb vs Low Fat Study


A recent study that garnered a lot of media attention compared the effects of a low-carbohydrate diet with a low-fat diet on weight and cardiovascular risk factors.  This low carb vs low fat study titled, Effects of Low-Carbohydrate and Low-fat diets: A Randomized Trial, was published in the top-tier, peer reviewed journal, Annals of Internal Medicine. After blinded randomization, this parallel-group study followed 148 men and women for one year and found that the low-carb diet induced greater weight loss, improved body composition and greater reductions in cardiovascular risk factors than the low-fat diet.

These findings add to a growing body of literature that supports lower carb approaches for weight loss and improved cardiovascular parameters, and collectively chip away at long held assumptions about diet and health that rely on low-fat doctrine. Within the Paleo community, this is welcome news but should low-carb enthusiasts rejoice? What exactly did this study measure and what exactly did it prove?

The Method

In this randomized parallel-group trial, 148 participants ranged in age from 22 to 77, were 88% female, 51% black and had no self-reported heart disease, diabetes, kidney disease or use of prescription weight loss medications. Body mass index ranged from 30 to 45.

Computer-generated randomization assigned 73 participants to the low-fat group and 75 to the low-carb group. The low-carb group was instructed to maintain a net carb intake of less than 40 grams a day and the low-fat group was instructed to keep total fat intake to less than 30% of total calories. There were no calorie limitations and low-carb or low-fat meal replacement bars or shakes were provided to each participant for daily consumption. Dietary counseling was provided throughout the study’s duration and data was collected via recall at the beginning of the study and at 3, 6 and 12 months.

research-results-300x270.jpgThe Results

  • 82% of the low-fat group and 79% of the low-carb group completed the intervention.
  • Physical activity and caloric intakes were similar with the low-carb group averaging 1507 calories a day and the low-fat group averaging 1615 calories a day.
  • The low-carb group achieved an average intake of 28% carbs, 40 to 43% fat and about 28% protein. The low-fat group averaged 40 to 45% carbs, 28% fat and 28 to 32% protein.
  • At the end of the year, the low-carb group lost about 12 pounds, compared to the low-fat group that lost only 4.
  • HDL cholesterol increased significantly more in the low-carb group than in the low-fat group and ratios of total-HDL cholesterol (a strong predictor of CHD) decreased significantly only in the low-carb group.
  • Those in the low-carb group lost most of their weight as fat mass, whereas those in the low-fat group lost more muscle compared to fat.
  • The low-carb group had significantly greater decreases in C-reactive protein (CRP – a measure of inflammation), than the low-fat group.
  • Participants in both groups decreased their waist circumference and insulin levels however, only the low-carb group had an increase in ketones.
  • Participants in the low-carb group had significant decreases in estimated 10-year risk for CHD and the low-fat group did not.

The Highlights

As the authors point out, the relatively large sample size, diverse population and high completion rate lend credence to this study. Periodic twenty-four hour recalls, (though notoriously problematic, were performed by trained and certified staff members) and urinary ketone levels revealed high rates of dietary compliance. The study, a blinded, randomized parallel-group trial is a type of Randomized Control Trial (RCT), the gold standard in study design.

The Criticisms

There are five noteworthy criticisms of this study. First, it’s been asserted that the low-carb group wasn’t really low-carb and the low-fat group wasn’t really low-fat. Second, real world applicability has been questioned as a result of interventions, namely the nutritional coaching and the provision of daily meal replacements. Third, dietary recall data show increased carb intake among the low-carb group as the study progressed suggesting the study wasn’t long enough. Forth, dietary recalls are inherently unreliable and fifth, a meta-analysis of nearly 50 diet studies was published on the same day in the prestigious journal JAMA showing seemingly conflicting results.

Addressing the Criticisms and My Takeaway

fat-and-carbs-300x209.jpgThere are no official definitions for low-carb or low-fat diets.  A diet is low-carb if it falls below current dietary recommendations for carbohydrates and likewise for low-fat diets.  At a 28% carbs, low-carb group was well below the USDA’s current recommendation that carbs comprise 45% to 65% of the diet.  At 28% fat, the low-fat group was actually within current recommendations to limit fat between 20% and 35% of total calories.  Therefore, it could be said that the low-fat diet was not really low-fat.

Regarding the interventions of nutritional coaching and daily meal replacements; yes, extrapolation to real world applications is compromised. However, the results of the study remain unaffected in that they reflect macronutrient intake specifically.

And what of the carbohydrate creep? The carbohydrate intake of the low-carb group decreased from an average of 242 g/d at baseline to 97 g/d at 3 months, and 93 g/d at 6 months, then increased to 127 g/d at 12 months. Yes, it’s possible that if the study lasted longer, the low-carb group would have continued to increase carb intake. A study of longer duration is needed to shed light on this.

The authors of the study admit to the limitations of dietary recalls and took precautions to minimize bias.

Finally, the meta-analysis published in JAMA did not actually contradict the Annals study. In the Annals study, calories were not restricted so that weight loss could be attributed to differences in carb or fat intake.   The JAMA analysis evaluated diets where participants reduced caloric intake making it impossible to extrapolate whether weight loss was due to reduced carbs, reduced fat or reduced calories. Furthermore, the meta-analysis didn’t determine if weight lost in these studies was fat or muscle.

My takeaway? The Annals of Internal Medicine study is a pretty good one. It adds heavy weight (couldn’t resist the pun) to a growing body of evidence that shows current dietary recommendations for carbohydrate and fat to be less favorable for health than once assumed. With it’s emphasis on high quality whole foods and avoidance of processed items and refined carbohydrates, the macronutrient profile of the Paleo diet falls more in line with that of the Annals’ low-carb group than with those of the low-fat group or the standard American diet. So, go ahead Paleo peeps, rejoice.

Selected References:

Annals of Internal Medicine study
Which diet is best for long-term weight loss?