[1] https://www.reddit.com/r/science/comments/6oufus/comment/dkk...
This is good to see. I wonder if unregistered studies will ever be ignored like non-peer-reviewed studies are.
* the "multiply p value by ten" rule would exclude the main results (but not some well known findings)
* paper does not have much discussion of context or past results
* paper does not have any scatter plots or show individual data at all for the primary results; there are scatter plots under additional files for oxidation vs. lean body mass and vs. fat mass and 24 hour energy expenditure vs fat mass
* there is no discussion of the possibility that the small additional calories could have had the same effect even if not sugar
* there is no discussion of the possibility that gut microbial response to the non-sugar drink might cause the differences they see
* a conflicting study used a different artificial sweetener
* the study did not test if the same SSB vs NNSB given at non-meal times would have a similar or different effect
* also all the limitations mentioned in the study
* this is a basic research study, not something directly applicable to real life; if you aren't doing extensive research in the area then drawing any conclusions from this one paper is unlikely to be helpful
Of course, considering the study in connection with the rest of the literature is likely to address some of these considerations and it may be a valuable contribution. Saying "very likely to be correct" does not sound accurate to me.
Since fructose is suspect, comparing trehalose vs. sucrose would be interesting.
the possibility that gut microbial response ... might cause the differences