[kelseyfrog]

I know this is a bit hyperbolic to prove a point but it brings to the surface a point which so far goes unaddressed: the modulation and variation of metabolic efficiency.

Between the consumption of glucose and gasoline, there exists a spectrum of efficiency in the conversion of thermodynamic calories to useful bodily energy.

Take glucose for example. It's thermodynamic and metabolic calorie values are nearly equal (15.5MJ/kg or 3.7kcal/g). On the opposite end is gasoline, thermodynamically it's 46 MJ/kg but it's 0kcal/g if one were to ingest it. There simply isn't a metabolic pathway to extract energy from gasoline in human physiology.

I have no reason to doubt that the ratio of thermodynamic to metabolic efficiency varies amongst individuals, over time, and in the presence or absence of other compounds. When 9-14kcal/day supports the maintenance of 1kg of body mass in equilibrium, it's clear that, ceteris paribus[1], minute changes in metabolic efficiency can explain rather large differences in human mass.

1. Truly. Even/especially in the absence of differences in consumption.