Didn't this study have a terribly small sample size? 13 adults were divided into 3 groups. That is 4 people in each group. There must be major individual differences how their bodies react to different diets during a four week period. I don't see how this is statistically significant.