How food structure and processing govern metabolizable energy: a critical review of the food-matrix determinants of energy availability and the limits of the Atwater accounting.

Food energy is usually reported using Atwater factors: composition-based coefficients that treat metabolizable energy as a function of macronutrient mass rather than food structure. Yet a growing food-science literature shows that metabolizable energy can diverge materially from Atwater predictions according to the integrity of the food matrix and the processing applied to it. This review brings energy-accounting and food-structure science into direct contact. It defines where the Atwater system remains useful and where its structural blind spot becomes consequential; clarifies the frequent conflation of digestibility, bioaccessibility, bioavailability and metabolizable energy, and distinguishes physical resistance from chemical inhibition of digestive enzymes while treating the two as coupled rather than independent. It then maps major processing operations to their structural and energetic consequences, and compares methods used to quantify metabolizable energy and nutrient release, including human balance studies, ileostomy and intubation studies, standardised in vitro digestion, dynamic models and in silico approaches. Tree nuts provide the strongest empirical case. Measured metabolizable energy ranges from about 5% below the Atwater value for pistachios to more than 30% below for whole almonds, a divergence that is explained by matrix integrity and lipid encapsulation. Four energetic consequences of matrix disruption are considered: net metabolizable energy, rate of energy release, thermic effect of feeding and colonic salvage. The review concludes that energy values based on composition alone are insufficient for foods whose structure strongly constrains digestion, and that food-composition data, labelling, feeding research and reformulation would benefit from structure-aware measures of delivered energy.