Energy expenditure of constant- and variable-intensity cycling: power meter estimates

Purpose: The objective of this study is to compare the effects of constant-and variable-intensity cycling on gross efficiency (GE) and to compare estimates of energy expenditure (EE) made using indirect calorimetry (CAL) with estimates derived from commercially available power meters. Methods: Nine national team female road cyclists completed a GE test (GE(test) = 4 min at approximately 45%, approximately 55%, approximately 65%, and approximately 75% maximal aerobic power (MAP)) before and after 10.5 min of either constant- (CON)- or variable-(VAR)-intensity cycling averaging approximately 55% MAP. GE measured before, after, and during CON and VAR cycling was compared. Total EE (kJ) for 10.5 min of VAR cycling was estimated using indirect CAL and compared with estimates on the basis of mechanical power [Schoberer Rad Messtechnik (SRM)] using the group mean GE, each athlete's mean GE, and each athlete's power to GE regression. Results: There was no effect of VAR on GE(tests) (P = 0.74). GE reduced from 19.1% +/- 0.4% (mean +/- SE) during the pretrial GE(tests) to 18.7% +/- 0.4% during the posttrial GE(tests) (P < 0.05) in both conditions. Differences in GE (mean +/- SD) measured during CON (18.4% +/- 1.6%) and VAR cycling (18.6% +/- 1.1%) were trivial (P = 0.28). SRM-based estimates of EE were most accurate when using individual athlete's power GE regression using Pre- and Post-VAR GE(test) data combined (Delta(%) +/- 90% CI, 0.3 +/- 0.8; R-2 0.98,