The decline in resting energy expenditure (REE) during weight reduction is due to a loss of body mass, to a lesser thermic effect of food, and to a decline in the metabolic rate within the body's tissues. The latter is considered to reflect a metabolic adaptation, generally referred to as ‘adaptive thermogenesis’. In their recent review Clinical significance of adaptive thermogenesis,1 GC Major, E Doucet, P Trayhurn, A Astrup and A Tremblay argue that adaptive thermogenesis is quantitatively significant in compensating ‘at least partly, for the prescribed energy deficit’. In the famous Minnesota starvation study, the decline in basal energy expenditure per unit of active tissue was 15.5% (not 25% as stated in Major et al.1) after 24 weeks of semi-starvation.2 The 40% overall decrease in basal metabolic rate (BMR) was therefore mostly attributable to weight loss (16.8 kg). One would in fact rarely expect a decline in the resting metabolic rate per unit of tissue to compensate for more than some 25% of a given energy deficit while on a restricted diet.
The authors' statement that adaptive changes in thermogenesis can ‘in some cases [be] quantitatively sufficient to overcome the prescribed energy restriction’ is therefore truly extraordinary! It is undoubtedly the point that will be most noted! The data supporting this claim (Table 1) were obtained during a study of obese adults participating in a weight reduction program, with a prescribed reduction in energy intake of 2900 kJ/day.3 Twelve of the 15 men, and 15 of the 20 women received 60 mg fenfluramine daily. REE was determined by indirect calorimetry, by analysis of a 15-min collection of expired air through a mouthpiece while the nose was clipped,3 a procedure now generally considered to be unreliable. Table 1 describes the maximal observed individual deviation between the observed REE and the REE predicted for the reduced body weights reached after 8 weeks, which amounted to −2606 in men and −2696 kJ/day in women subject. The deviations from predicted REE averaged −953 in men and −614 kJ/day in women at that time. The s.e.m. of ±230 and ±173 described in Doucet et al.3 imply s.d. of approximately ±860 and ±755 kJ/day. The values cited in Table 1 differ from the mean by 1653 and 2082 kJ/day, or 1.9 and 2.8 times the s.d. By applying the Chauvenet criterion4 for the analysis of outlying results, the deviations cited in Table 1 would be identified as spurious.
This is a preview of subscription content, access via your institution