Rédiger un PREreview

A recent Science perspective by Magkos, Forde, and Robinson critically examined five randomized controlled trials on ultraprocessed foods (UPFs) and concluded that the evidence for a processing-specific effect on energy balance is weak. Their analysis exposed a deeper problem: the reliance on the Wishnofsky rule – a fixed tissue energy density of 7,700 kcal/kg – to translate weight change into energy imbalance. This rule, though widely used, is a statistical abstraction, not a physical constant. Here, I argue that the Wishnofsky rule constitutes a structural blind spot that prevents energy balance model (EBM)-based analyses from adequately accounting for water shifts, nitrogen balance, and the varying composition of tissue change. I apply the mass balance model (MBM) to reinterpret the five UPF trials and demonstrate that MBM – by tracking mass flows directly in grams and decomposing weight change into its true components (fat, lean mass, glycogen, water) – resolves the very ambiguities that the Science perspective identified. I conclude by outlining what an MBM-informed UPF trial should measure and argue that future feeding studies must move beyond the Wishnofsky rule to direct mass accounting to yield mechanistically interpretable results.