Experimental design

All participants were requested to avoid physical activity, caffeine and dietary supplements in the 24 h prior to testing, and to use motorised transport to reach the laboratory. Anthropometry and body composition (using bioimpedance analysis) were measured as previously described⁽,13⁾. The metabolic measurements were conducted in the morning after an overnight fast, with the participant seated comfortably in a car seat adapted for continuous measurements of REE and RQ by ventilated hood indirect calorimetry (Cosmed Quark RMR; Cosmed srl), as previously described⁽,13⁾. The O₂ analyser (with paramagnetic O₂ sensor) and CO₂ analyser (with an IR digital sensor) offer a fast response time to measure O₂ and CO₂ changes within 120 ms; the range of O₂ measurement is from 0 to 30 % and that of CO₂ measurement from 0 to 10 %, both with an accuracy of ±0·02 %. Prior to each test, the gas analysers were calibrated using a certified gas mix (5 % CO₂, 16 % O₂ and 79 % N₂). The turbine flowmeter has an accuracy of ±2 % and was calibrated prior to each test with a 3 litre syringe. Flow rate during each test was fixed between 30 and 37 litres/min so as to maintain the concentration of CO₂ in air flowing out of the canopy between 0·7 and 1·0 %. In order to avoid baseline drift the Cosmed Quark RMR system automatically recalibrates every 5 min (over a 55 s period). REE was calculated according to the Weir equation⁽,14⁾, and RQ was calculated as the ratio of CO₂ produced to O₂ consumed (i.e. RQ = VCO₂/VO₂). As short-term urinary collections to assess total N excretion may not be representative of the protein oxidised during the measurement itself, they were not obtained in this study, and assumed to be 13 g/24 h; the latter value reflects urinary N excretion of subjects in the post-absorptive (fasted) state⁽,15⁾. It should be noted that this assumption will not significantly influence the relative partition between carbohydrate and fat oxidation determined by indirect calorimetry because (i) the average overnight-fasted RQ of the subjects (about 0·80) is close to the RQ of protein (about 0·82), (ii) the proportion of total oxidation derived from protein is small (12–15 %) and (iii) in response to carbohydrate ingestion (as in our study here), this proportion is likely to be even smaller due to the potential protein-sparing effect of carbohydrate on protein oxidation.

On each test day, upon arrival in the laboratory at approximately 08·00 hours, the subject rested in the seated position for 15–20 min. This was followed by 30–40 min of baseline measurement, during which stabilisation of REE was achieved. Stabilisation was defined as no more than 2 % variability of REE, with no consistent upward or downward trend. The subject then drank, in 4 min, a 500 ml beverage containing distilled water, 10 ml of lemon juice and 60 g of d(+)glucose, d(+)galactose or d(−)fructose (Argos Organics, Chemie Brunschwig SA) in a randomised cross-over design; lemon juice was added in order to mask differences in taste of the different sugars. The ventilated hood was then replaced and post-drink metabolic monitoring continued for a further 150 min. In order to reduce boredom and accompanying stress, the participants were permitted to watch a calm movie or a documentary. All participants were blinded as to the order in which they received the sugar drinks.