Back

The mass balance model perfectly fits both Hall et al. underfeeding data and Horton et al. overfeeding data

Arencibia-Albite, F.; Manninen, A. H.

2021-03-03 nutrition
10.1101/2021.02.22.21252026 medRxiv
Show abstract

Background & AimsRecently, the validity of mass balance model (MBM) was questioned based on two feeding studies. Thus, we simulated both of these feeding trials. MethodsMBM describes the temporal evolution of body weight and body composition under a wide variety of feeding experiments. This computational study simulated, utilizing MBM, the underfeeding trial by Hall et al. (Cell Metab. 2015;22:427-36) and the overfeeding trial by Horton et al. (Am J Clin Nutr. 1995;62:19-29). ResultsOur simulation results indicate that data from both of these feeding trials perfectly match MBM-based predictions, i.e., MBM gives a remarkably accurate description of experimental data. ConclusionsIt is becoming increasingly clear that our model (MBM) is perfectly able to predict body weight and body composition fluctuations under a wide variety of feeding experiments. "There is a stupid humility that is not at all rare, and those afflicted with it are altogether unfit to become devotees of knowledge. As soon as a person of this type perceives something striking, he turns on his heel, as it were, and says to himself: "You have made a mistake. What is the matter with your senses? This cannot, may not, be the truth." And then, instead of looking and listening again, more carefully, he runs away from the striking thing, as if he had been intimidated, and tries to remove it from his mind as fast as he can. For his inner canon says: "I do not want to see anything that contradicts the prevalent opinion. Am I called to discover new truths? There are too many old ones, as it is..." - Friedrich Nietzsche, The Gay Science (1882)

Matching journals

The top 1 journal accounts for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.