It’s the BIG one. The study all the low carb zealots were waiting for to PROVE that low carb was superior to every other method. In this video we discuss the recent meta-analysis by Ludwig et al that concluded low carb diets increase energy expenditure (EE) in long(er) term studies.
Study: https://pubmed.ncbi.nlm.nih.gov/33274750/
This meta-analysis was well done from a statistics perspective, but there are major GAPING holes that aren’t explained.
Major criticisms in video:
Omitting metabolic chamber data when doubly labeled water (DLW) data available. Metabolic chamber is the gold standard for energy expenditure and a DIRECT measurement and DLW is merely a surrogate… so why would Ludwig choose to use DLW when a direct measurement was available? This would be like using skinfold data instead of MRI measurement of body fat even when both were available!
Using DLW data biases the long term studies towards low carb superiority as DLW has NOT been validated for energy expenditure measurements during low carb diets. This is likely due to the change in respiratory quotient (RQ) (ratio of O2/CO2) during low carb diets which causes greater than predicted increase in CO2. Since RQ is used in the calculation to determine energy expenditure from DLW, this causes an overestimation of energy expenditure when using DLW for low carb diet studies. Additionally, the increased loss of body water from low carb diets also exacerbates this overestimation. This is likely why ONLY studies using DLW have shown significant increases in EE, whereas metabolic chamber studies have failed to show the same outcomes.
Citations:
https://www.biorxiv.org/content/10.1101/403931v3
https://pubmed.ncbi.nlm.nih.gov/27385608/
Furthermore, there was evidence of biasing the data to favor low carb diets. One of the studies (Rumpler et al.) in the meta-analysis was 35 days long but Ludwig chose data from the 28 day mark where the difference between low carb diets and low fat diets was the greatest (~200 kcal/d difference) however on day 35 there was no difference between the two diets. No good explaination was provided for why this timepoint was chosen compared to day 35. During days 28-35 the subjects were moved from a deficit to maintenance. This might make sense as to why Ludwig chose this time point if not for his own studies which were done at maintenance and included in the same meta-analysis. The choice to using the data from day 28 instead of day 35 does not make sense and is not adequately explained by the researchers.
Citation:
https://pubmed.ncbi.nlm.nih.gov/1989409/
Further, in other studies (Abbot et al.) Ludwig claims that the low carb diet increased energy expenditure more than the control diet, but this is not supported by the data in the paper and the conclusions from the researchers themselves
Citation:
https://pubmed.ncbi.nlm.nih.gov/2305878/
Further, Ludwig claims that perhaps fat adaptation may be required to experience the increase in energy expenditure from low carb diets and that’s why the experiments more than 17 days demonstrate an increase in energy expenditure. This proposed period of fat adaptation is not supported by any literature that Ludwig cites in the introduction, in fact several of them support the notion that fat adaptation is short (<7 days). Further, Hall et al demonstrated that the small increase in energy expenditure from low carbohydrate diets (~50kcal/d) was during the first week and then returned to normal once the subjects were adaptated, suggesting the OPPOSITE of what Ludwig claims Citation: https://pubmed.ncbi.nlm.nih.gov/27385608/
Finally, there is NO data demonstrating increased fat loss from low carb diets in these studies. In fact, every single study which ‘demonstrated’ increased energy expenditure ‘from’ low carb diets did NOT show any significant increase in fat loss when compared to the low fat diet. In fact, several of the studies showed a slight (not statistically significant) advantage of fat loss to the low fat diets. If low carb diets TRULY cause a REAL increase in energy expenditure, why does it not cause increased fat loss? Because the increase is likely not real and simply due to a data artifact of DLW not being a viable measurement of energy expenditure for low carbohydrate diets.
In summary: the current meta-analysis does not support the researchers conclusions that low carbohydrate diets increase energy expenditure. It merely supports the non-validity of DLW for use in assessing energy expenditure during low carb diets.
