Prediction of cellular ATP generation from foods in the adult human : application to developing specialist weight-loss foods : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Nutritional Science at Massey University, Palmerston North, New Zealand

Abstract

For the accurate prediction of the potential ‘available energy’ of a food at the cellular level (i.e. ATP generation from food) it is necessary to be able to predict both the quantity and location of uptake (upper-tract or colon) for each energy-yielding nutrient. The objective was to develop a valid model (‘Combined Model’) for predicting the (potential) ATP available to the body from absorbed nutrients across the total digestive tract. The model was intended for the adult human under conditions where energy intake ≤ energy expenditure and all absorbed nutrients are catabolised. The development of the model involved two parts: (i) the experimental development of a dual in vivo – in vitro digestibility assay (‘dual digestibility assay’) to predict human upper-tract nutrient digestibility, as modelled by the rat upper digestive tract, and colonic digestibility, as predicted by fermenting rat ileal digesta in an in vitro digestion system containing human faecal bacteria; and (ii) the development of a series of mathematical equations to predict the net ATP yielded during the post-absorptive catabolism of each absorbed nutrient at the cellular level. A strong correlation (r=0.953, P=0.047) was found between total tract organic matter digestibility (OMD), as predicted with the newly developed dual in vivo – in vitro digestibility assay and with that determined in a metabolic study with humans for four mixed diets ranging considerably in nutrient content. There were no statistically significant (P>0.05) differences for mean OMD between the predicted and determined values for any of the diets. The Combined Model (dual in vivo – in vitro digestibility assay + stoichiometric predictive equations) was applied to three meal replacement formulations and was successfully able to differentiate between the diets in terms of both energy digestibility and predicted ATP yields. When the energy content of each diet was compared to that of a baseline food (dextrin), some metabolisable energy (ME) models gave considerably different ratios compared to that predicted by the Combined Model. By way of example, for Diet C a ratio of 0.96 (Atwater and FDA models) was found ii versus 0.75 (Combined Model). Thus, the model has practical application for predicting dietary available energy content, particularly in the research and development of specialised weight-loss foods, where it may be more accurate than some current ME models. Uniquely, the Combined Model is able to define a food in terms of ATP content (mol ATP / g food) using recent estimates of cellular P/O ratios and therefore, directly relates dietary energy intake to the quantity and form (ATP) of energy ultimately delivered at the cellular level.