Agreement Between Measured And Estimated Energy Expenditure During Exercise: Modulating Effects Of Sex And Race

Abstract

The ACSM metabolic equations have been used to estimate energy expenditure (EE) during exercise; yet limited evidence supports their accuracy in overweight/obese and sedentary individuals, especially during aerobic exercise training. PURPOSE: Evaluate the relationship between measured and estimated EE over a 24-wk exercise intervention and determine moderating demographic and anthropometric relationships. METHODS: Data from the EMECHANIC study were analyzed (N = 108; 77 female; 71 Caucasian). Participants (sedentary) were randomized to aerobic exercise (8 or 20 kcal/kg body mass/week). EE was measured via indirect calorimetry at absolute (2 mph, 0% grade) and relative (65-85% VO2max) intensities pre-intervention, every 2 weeks for the initial 12 weeks, and monthly thereafter. An EE offset factor (EOF) was computed at each intensity by expressing measured EE as a percentage of the ACSM estimated EE. An EOF lower than 100% indicates an overestimation of EE when using the equations. RESULTS: Absolute and relative EOF decreased during the intervention (time effect, P < 0.01). Exercise dose had no effect on EOF (P > 0.50). A sex effect was found for absolute EOF (P < 0.01), with women exhibiting lower values throughout the study compared to men (94.4 ± 1.9% vs 99.9 ± 3.0%; mean ± 95%CI), but not for relative EOF (P = 0.11). No sex by time interactions were observed for absolute or relative EOF (P > 0.35). African Americans (AA) demonstrated lower absolute (90.2 ± 2.9% vs 98.3 ± 1.9%, P < 0.01), but not relative EOF (87.3 ± 2.0% vs 88.3 ± 1.3%, P = 0.39), compared to Caucasians. No race by time interactions were observed for absolute or relative EOF (P > 0.35). At absolute intensity, sex and race differences in EOF are mirrored in VO2 (P < 0.03), but not respiratory exchange ratio (RER; P > 0.15). Although relative EOF was not different, RER was higher in AA and VO2 was lower in women (P < 0.03). Higher body weight resulted in a lower absolute and relative EOF at all time points (P < 0.01). Age had no effect on absolute or relative EOF (P = 0.91; P = 0.07, resp). CONCLUSION: Calculated EE may be overestimated in women and AA, especially at lower exercise intensities. Studies that use ACSM estimated EE in these populations (ex. Calorie balance studies) may overestimate prescribed EE resulting in lower than anticipated daily EE.