Sunday, October 19, 2014

Summary of the Low Carb Diet Trials: Part 1

Over the past several weeks I looked at the low carb trials, specifically those with at least 100 participants and running for at least 12 months.  The links to those blog posts are below:

Overall, the trials had a good balance between men (796) and women (1064).  All of the studies set ambitious targets for the low carb group as a daily carbohydrate limit of 20-50g or even 100g is a major difference to what most people regularly eat.  In almost all the trials the low carb group returned to eating ~30-40% carbohydrate at the final time point (except in Stern and Brinkworth).  The targets for the low fat group were not too different from what people usually eat, but almost always featured deliberate calorie restriction (CR).

There wasn’t much difference between the low carb and low fat groups, although the low carb group far more often had better outcomes than the low fat group with one exception (more adverse symptoms during first 6 months in the Foster trial).

* In the Gardner trial the low carb group often had a trend for better results compared to the other groups but didn’t reach statistical significance.  This may have been at least partially due to there being 4 groups in the trial, which makes it more difficult for one group to reach statistical significance.

Y = Low carb had significantly better result than low fat
- = Not significantly different regarding low carb vs. low fat
N = Low carb had significantly worse result than low fat

Some trials (Stern, Shai, Bazzano) fairly consistently found a better outcome with the low carb diet while others (Dansinger, Brinkworth, Lim, Iqbal) fairly consistently found no difference.  This doesn’t seem to be related to gender or the intensity of the intervention.  I’ve speculated on the factors responsible below:

Low Carb > Low Fat Trials
Low Carb = Low Fat Trials
Possible Reasons Why
Possible Reasons Why
Fairly good adherence in LC group.  Very poor adherence in LF group
Protein intake higher in LC group
Self-selected adherence for last 10 months, therefore likely poor adherence
Protein intake higher in LC group
Diets were isocaloric
Protein intake much higher in LC group

Very poor adherence to diets in both groups (no difference between groups regarding diet)

Protein intake may be a major factor in explaining this.  In 3/4 of the trials in the first column the low carb group had a higher protein intake (we aren’t given any dietary intake information in Foster).  Whereas only 1, maybe 2 (but not a big difference), of the trials in the second column had a higher protein intake in the low carb group and the diets in Brinkworth were isocaloric which explains the lack of difference between the groups anyway.

* When considering these results it’s important to remember that the carbohydrate intake in the low carb group was ~30-40% of total calories (rather than 20-50g) in all trials except Stern (251g to 120g carbs at 12 months) and Brinkworth (9% carbs at 12 months).  This means that generally these trials didn’t really test a true low carb diet (which I define as ~25% of total calories or ~100-150g) or a very low carb or ketogenic diet (where carbohydrate intake is ~10% of total calories/~50g or less).

No comments:

Post a Comment