Title: Methodological considerations for meal-induced thermogenesis: measurement duration and reproducibility Running title: Methodology of meal-induced thermogenesis Key words: meal induced thermogenesis, postprandial, thermic effect of food, reproducibility, energy expenditure Authors:

1 Meal-Induced Thermogenesis (MIT) research findings are highly inconsistent, in part, due to 2 the variety of durations and protocols used to measure MIT. We aimed to determine: 1) the 3 proportion of a 6 h MIT response completed at 3, 4 and 5 h; 2) the associations between the 4 shorter durations and the 6 h measure; 3) whether shorter durations improved the 5 reproducibility of the measurement. MIT was measured in response to a 2410 KJ mixed 6 composition meal in ten individuals (5 male, 5 female) on two occasions. Energy expenditure 7 was measured continuously for 6 h post-meal using indirect calorimetry and MIT was 8 calculated as the increase in energy expenditure above the pre-meal RMR. On average, 76%, 9 89%, and 96% of the 6 h MIT response was completed within 3, 4 and 5 h respectively, and 10 the MIT at each of these time points was strongly correlated to the 6 h MIT (range for 11 correlations, r = 0.990 to 0.998; p < 0.01). The between-day CV for the 6 h measurement was 12 33%, but was significantly lower after 3 h of measurement (CV = 26%, p = 0.02). Despite 13 variability in the total MIT between days, the proportion of the MIT that was complete at 3, 4 14 and 5 h was reproducible (mean CV: 5%). While 6 h is typically required to measure the 15 complete MIT response, 3 h measures provide sufficient information about the magnitude of 16 the MIT response and may be applicable for measuring individuals on repeated occasions. 17 18