Objectives: To gain insights into the control systems underlying human variability in the regulation of body composition during weight recovery, as well as the disproportionate recovery of fat relative to lean tissue, the classical Minnesota Experiment conducted on 32 men subjected to long-term semi-starvation and refeeding was revisited with the following objectives: (1) to determine whether the control of energy-partitioning between lean and fat tissues during weight loss and weight recovery is an individual characteristic, and if a predictor can be statistically identified, (2) to determine whether the reduction in thermogenesis during weight loss persists during weight recovery, and underlies the disproportionate recovery of fat tissue and (3) to integrate the control of energy-partitioning and that of thermogenesis in order to explain the pattern of lean and fat tissue mobilisation and deposition during weight loss and weight recovery.
Methods: Individual data on body weight, body fat, fat-free-mass (FFM), and basal metabolic rate (BMR), assessed during the control baseline period (i.e. prior to weight loss), at the end of 24 weeks of semi-starvation, and at the end of a 12 week period of restricted refeeding, were used to calculate the following parameters: (i) a quantitative index of energy-partitioning, the P-ratio, defined as the proportion of body energy mobilised as protein during weight loss, or as the proportion of body energy deposited as protein during weight recovery, (ii) a quantitative index of changes in thermogenesis, defined as the change in BMR adjusted for FFM (or for both FFM and fat mass) and (iii) the degree of replenishment of fat and FFM compartments, defined as the recovery of body fat and FFM (during refeeding) as a percentage of that lost during semi-starvation.
Results: This re-analysis indicates the following: (i) a large inter-individual variability in P-ratio during both weight loss and weight recovery, but for a given individual, the P-ratio during refeeding is strongly correlated with the P-ratio during semi-starvation, (ii) body composition during the control period is the most important predictor of variability in P-ratio, such that the higher the initial % body fat, the lower the proportion of energy mobilised as protein, and hence the greater the propensity to mobilise fat during semi-starvation and to subsequently deposit fat during refeeding and (iii) at week 12 of refeeding, the change in adjusted BMR is found to be reduced by a magnitude which is inversely proportional to the degree of fat recovery, but is unrelated to the degree of FFM recovery. A quantitative relationship is derived between the P-ratio during refeeding, the % fat recovery, and the P-ratio during semi-starvation.
Conclusions: Evidence is presented here suggesting that (i) human variability in the pattern of lean and fat tissue deposition during weight recovery is to a large extent determined by individual variations in the control of energy-partitioning, for which the initial % body fat is the most important predictor and (ii) the disproportionate gain in fat relative to lean tissue during weight recovery is contributed by a reduction in thermogenesis (i.e. increased efficiency of food utilization) for accelerating specifically the replenishment of the fat stores. These control systems, operating via energy-partitioning and thermogenesis, have been integrated into a compartmental model for the regulation of body composition during underfeeding/refeeding, and can be used to explain the individual pattern of lean and fat tissue deposition during weight recovery in situations ranging from the rehabilitation after malnutrition to the relapse of obesity.