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

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.

Nine national team female road cyclists completed a GE test (GEtest = 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.

There was no effect of VAR on GEtests (P = 0.74). GE reduced from 19.1% ± 0.4% (mean ± SE) during the pretrial GEtests to 18.7% ± 0.4% during the posttrial GEtests (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 GEtest data combined (Δ(Equation is included in full-text article.)(%) ± 90% CI, 0.3 ± 0.8; R 0.98, P <0.001). Group mean estimates were within approximately 1% of CAL, although individual errors were approximately 11%.

Findings support the use of calibrated power meters for estimating cycling EE. For trained female road cyclists, total mechanical work (kJ) multiplied by 5.3 (GE = 19%) provides a valid estimation of total EE during variable-intensity cycling <75% MAP, although determining each athlete's GE improves accuracy greatly.