Journal Logo

F-26 Free Communication/Poster - Cycling Physiology: JUNE 1, 2007 1: 00 PM - 6: 00PM ROOM: Hall E

The Reliability of the Mechanical Power Output and Peak Power Output Estimation Equations in Cycling


Board #82 June 1 2:00 PM - 3:30 PM

Galilea, Pedro A.; González-Haro, Carlos; Drobnic, Franchek; Escanero, Jesús F

Author Information
Medicine & Science in Sports & Exercise: May 2007 - Volume 39 - Issue 5 - p S445-S446
doi: 10.1249/01.mss.0000274762.36146.84
  • Free

PURPOSE: To assess the reliability of the most important published equations that estimate the mechanical power output and the peak power output (PPO) in cycling efforts.

METHODS: Thirty-four endurance modality cyclists (mountain bikers (n=4) and road cyclists (n=11)) and triathlets (n=19) of both sexes performed two incremental tests to exhaustion in velodrome, at which the mechanical power output was measured by means of the SRM (Schoberer Rad Messtechnik®, Germany) system. The measurements, belonging to all the aerobic range, were compared with the estimations that provided 9 published equations (Eq1: Di Prampero et al. 1979, Eq2: Whitt, 1971, Eq3: Whitt and Wilson, 1982, Eq4: Menard, 1991, Eq5: Basset et al., 1999, Eq6: Kyle, 1991, Eq7: Olds et al., 1993, Eq8: Olds et al., 1995, Eq9: Candau et al., 1999) by means of the standardized mean prediction error (SMPE), the systematic error (SE) and the random error (RE). Moreover, the Bland-Altman methodology was used to compare the PPO measured by the SRM system with that given by the equations. When the Bland-Altman methodology was not possible to be applied, the equations absolute reliability referring to PPO was calculated by means of the intraclass correlation coefficient (ICC) with 95 % CI, the standard error of the estimate (SEE) and the coefficient of variation (CV).

RESULTS: The results showed that, in the estimation of the mechanical power output, Eq1 and Eq9 presented a moderate SMPE and a low RE. On the other hand, when comparing the PPO, the Bland-Altman methodology could not be applied in any case due to heteroscedasticity (the Kendall coefficient was significant and remained after the log transformation data). However, the PPO estimations showed an acceptable CV (from 12.3 to 14.4%), and a high ICC (Eq1: 0.880, Eq2: 0.847, Eq3: 0.862, Eq4: 0.792, Eq5: 0.847, Eq6: 0.838, Eq9: 0.853), particularly Eq7 and Eq8 (0.911 and 0.892, respectively). Despite this, the SEE were also high (Eq1: 109 W, Eq2: 89 W, Eq3:67 W, Eq4: 166 W, Eq5: 194 W, Eq6: 220 W, Eq7: 69 W, Eq9: 70 W), except for Eq8, where it was acceptable (46 W).

CONCLUSIONS: The equations proposed by Candau (1999) and Di Prampero et al. (1979) are the ones that estimate the mechanical power output all along the aerobic range the best. The Olds et al. (1995) equation is the one that estimates the PPO the best.

© 2007 American College of Sports Medicine