467
Day to Day Variation in Time Trial Cycling Performance M. S. Hickey, D. L. Costill, G. K. McConell, I I Widrick, H. Tanaka Human Performance Laboratory, Ball State University, Muncie, IN 47306
Introduction
Abstract M. S. Hickey, D. L. Costill, G. K. McConell,
.1 1 Widrick and H. Tanaka, Day to Day Variation in Time Trial Cycling Performance. mt J Sports Med, Vol 13, No 6, pp 467—470, 1992.
In the laboratory setting, assessment of performance is often used as the criterion measure of exercise capacity in physiological testing. Testing methodologies have traditionally included sub-maximal performance rides to exhaustion at a fixed percentage of VO2max (3—Il). Coyle and associates (5,6) have reported that, in well-trained cyclists, time
In an attempt to assess the reproducibility of laboratory cycling performance, eight well-trained (\TO2max
= 4.6 0.2 1 min )
male
cyclists completed 12 trials in-
volving 4 successive performance rides at each of three total work outputs (approximately 1600, 200, and 14 kilojoules, respectively). These trials, deisgnated as long, medium, and
short trials (LT, MT, ST), represented exercise bouts of 105.12±0.41, 12.03±0.17 and 0.55±0.11 minutes, respectively. The trials, conducted on a computerized cycle
to fatigue in prolonged cycling at a constant work output is highly reproducible, suggesting that this method represents a reliable means of assessing the effects of various dietary/training manipulations. More recently, however, Krebs and Powers (8) reported poor test-retest reliability in time to fatigue, defined here as a 10% decline in power output below the designated target value. These data suggest that the specific definition of fatigue or exhaustion used in an experimental design may influence the reproducibility of the testing protocol.
ergometer in an isokinetic mode, were separated by a minimum of 72 hrs. All trials for each subject were completed at the same time of day. In all trials, subjects were allowed to select the pace in order to complete the ride in the shortest possible time. The mean coefficient of variation (CV) for performance time in each trial was: LT = 1.01%, MT respectively. The CV for per0.95%, and ST = formance time in ST was significantly greater than the CV in either LT or MT. In LT, performance time was significantly faster, and the mean % JO2max was significantly
Williams et a!. (17) have suggested that a distinction be made between endurance capacity, i.e., time to exhaustion, and endurance performance, or the ability to complete a target work bout in a minimal amount of time. From a metabolic standpoint, these protocols appear to differ in terms of the influence of exogenous carbohydrate supplementation, as Williams, et al. (17) observed no overall difference in 30 km treadmill time between glucose, fructose, and water. However, it must be noted that the vast majority of exercise feeding studies have reported changes in endurance capacity, and relatively little information exists as to the influence of pre-competition
higher in trial 4 versus trials 1 —3. There was no order effect
feedings or supplementation during exercise performance
in the MT or ST rides. The CV for mean V02 (1. min ), mean % JO2max, and RER during the LT rides were 3.02
%, 3.64%, and 3.53 %, respectively. These data suggest
that trained cyclists have the ability to reproduce endurance performance with a CV of approximately 1.0% in a timetrial protocol. The observation that performance was significantly improved in LT-4 while no differences were observed between LT 1—3 suggests that psychological factors (i.e., the
last ride") may contribute to significant improvements in
endurance performance independent of experimental manipulations.
Key words
____________
Exercise performance, cycling, time trial,
reproducibility
(time-trial oriented) protocols (17). We believe that the assessment of exercise performance through the use of time-trial type protocols represents a logical extension of the available data on carbohydrate supplementation and exercise capacity, but feel
that the efficacy of this type of protocol must be established prior to its widespread application. Thus, the intent of this investigation was to examine the reliability of time-trial cycling performance in a laboratory setting. Methods
__________
Eight well-trained male cyclists served as subjects in this investigation. The men had participated in experiments in this laboratory in the past, and were familiar with the experimental procedures. All subjects were licensed, competi-
tive cyclists (USCF Cat. 2/3, Tn-Fed.). Prior time trial experience and personal bests at a standard (40 km) distance varied in the subject pool. All subjects were currently riding between 350 and 500km week- at the time of this investigation. After being fully informed of the risks and stresses associated with the project, the subjects gave their written consent
Int. J. Sports Med. 13 (1992) 467—470 Georg Thieme Verlag Stuttgart New York
to participate. Approval of the institutional review board for human subjects research was secured prior to the initiation of
Downloaded by: Queen's University. Copyrighted material.
Accepted after revision: May 5, 1992
468 mt. .1 Sports Med. 13 (1992)
M. S. Hickey, D. L. Costill, U. K. McConell, J I. Widrick, H. Tanaka Table 1 Coefficient of Variation in Performance Time and Work Output Variable
Time KJ Watts
0.94
V02 (I min l) V02 (% max)
LT
0.50 % 3.02 % 3.64 %
MT
0.53 %
ST
1.35 % * *
RER
interval (KM
Data are expressed as the mean of each individual's CV for the variables shown. KJ = total kilojoules per ride. * Expired gas was not collected in these trials. + p < 0.05 vs LT, MT. Watts represents the mean power Output in each trial.
data collection. The age, height, weight, and O2max for the subjects averaged (± SE): 22.1 (± 1.2) yrs., 180 (± 1.1) cm, 73,9
(±4.9)kg, and 4.6(±0.2)l.min'. A VO2max test was performed on a Cybex cycle ergometer (MET 100). A standard two-stage warm-up consisting of 5 minutes at 150 watts and 3 minutes at 250 watts was immediately followed by a graded exercise test which consisted of 25 W increments each minute. Expired gas was collected and analyzed using a computer-based system. VO2max
was taken as the average of the two highest consecutive data points. In all tests, VO2max plateaued with increasing workloads.
Subjects completed four rides at each of three total work outputs (i.e., 1600, 200, and l4kJ, which were equivalent to 40, 5 and 0.5 miles as determined by the ergometer). All trials were completed on an isokinetic controlled cycle ergometer (Cybex MET 100) which enabled the subject to selfpace each trial. Subjects were able to adjust cadence in 5rpm increments during the course of the test, and all such changes were recorded. The objective of the test was to finish as quickly as possible, with time being the criterion measure of perform-
ance. All trials were initiated with a standardized 10 minute warm-up at 150 watts. Bouts were conducted under controlled environmental conditions (22.5±0.5 'C, 46.6±3.6% RET). In order to limit external variables which might affect performance, the ergometer screen was covered in such a way that subjects could view only the cumulative work completed. Additionally, the subjects were unaware of elapsed time either during or after the trials. Thus, subjects completed all bouts with no reference to time and no knowledge of pacing. For the purposes of this investigation, we believed that every effort to minimize the number of confounding variables was necessary in order to assess the cyclist's capacity to reproduce performance without the influence of knowledge of pacing relative to other bouts, etc. The trials were completed on separate days, with a minimum of 72 hrs between bouts, and were conducted at the same time of day for each subject. Exercise was restricted
the day prior to each ride. In addition, prior to the 1600 kJ, or long trial (LT) rides only, subjects received a carbohydrate supplement (236gm. day-') that was consumed in addition to the normal diet for the 48 hr period prior to each trial, in an attempt to normalize muscle glycogen levels.
During each LT ride, subjects received a total of 150 ml of cool tap water every 15 minutes in an attempt to standardize hydration state. Heart rate was monitored telemetrically and recorded at 5 minute intervals throughout the ride. Split times and whole body ratings of perceived exertion using the Borg scale (2) were taken at the equivalent of 200 kJ intervals as determined by the ergometer. Similarly, expired gas was collected in Douglas bags at 200, 400, 600, 800, 1000, 1200, and 1400 kJ during each ride and analyzed to determine oxygen uptake, mean percentage of TO2max, and RER for each ride. Approximately two weeks after the completion
of the LT rides, subjects returned to the laboratory and completed a series of short performance rides. The medium trial (MT) and short trial (ST) rides were conducted as described for
the LT trials, with the exception that no expired gas was collected in these trials. The MT and ST rides were completed in blocks, i.e., all four ST rides, then all four MT rides, or vice versa. The order of these blocks were randomized. For all trials, work output was calculated from the equation MPH = (watt!
2.277)2 / (Cybex, Inc. calibration data). Data were analyzed using a repeated measures ANOVA. When significant F ratios were obtained, a Tukey's post-hoc test was used to locate these differences. Statistical significance was accepted at p
Day to Day Variation in Time Trial Cycling Performance M. S. Hickey, D. L. Costill, G. K. McConell, I I Widrick, H. Tanaka Human Performance Laboratory, Ball State University, Muncie, IN 47306
Introduction
Abstract M. S. Hickey, D. L. Costill, G. K. McConell,
.1 1 Widrick and H. Tanaka, Day to Day Variation in Time Trial Cycling Performance. mt J Sports Med, Vol 13, No 6, pp 467—470, 1992.
In the laboratory setting, assessment of performance is often used as the criterion measure of exercise capacity in physiological testing. Testing methodologies have traditionally included sub-maximal performance rides to exhaustion at a fixed percentage of VO2max (3—Il). Coyle and associates (5,6) have reported that, in well-trained cyclists, time
In an attempt to assess the reproducibility of laboratory cycling performance, eight well-trained (\TO2max
= 4.6 0.2 1 min )
male
cyclists completed 12 trials in-
volving 4 successive performance rides at each of three total work outputs (approximately 1600, 200, and 14 kilojoules, respectively). These trials, deisgnated as long, medium, and
short trials (LT, MT, ST), represented exercise bouts of 105.12±0.41, 12.03±0.17 and 0.55±0.11 minutes, respectively. The trials, conducted on a computerized cycle
to fatigue in prolonged cycling at a constant work output is highly reproducible, suggesting that this method represents a reliable means of assessing the effects of various dietary/training manipulations. More recently, however, Krebs and Powers (8) reported poor test-retest reliability in time to fatigue, defined here as a 10% decline in power output below the designated target value. These data suggest that the specific definition of fatigue or exhaustion used in an experimental design may influence the reproducibility of the testing protocol.
ergometer in an isokinetic mode, were separated by a minimum of 72 hrs. All trials for each subject were completed at the same time of day. In all trials, subjects were allowed to select the pace in order to complete the ride in the shortest possible time. The mean coefficient of variation (CV) for performance time in each trial was: LT = 1.01%, MT respectively. The CV for per0.95%, and ST = formance time in ST was significantly greater than the CV in either LT or MT. In LT, performance time was significantly faster, and the mean % JO2max was significantly
Williams et a!. (17) have suggested that a distinction be made between endurance capacity, i.e., time to exhaustion, and endurance performance, or the ability to complete a target work bout in a minimal amount of time. From a metabolic standpoint, these protocols appear to differ in terms of the influence of exogenous carbohydrate supplementation, as Williams, et al. (17) observed no overall difference in 30 km treadmill time between glucose, fructose, and water. However, it must be noted that the vast majority of exercise feeding studies have reported changes in endurance capacity, and relatively little information exists as to the influence of pre-competition
higher in trial 4 versus trials 1 —3. There was no order effect
feedings or supplementation during exercise performance
in the MT or ST rides. The CV for mean V02 (1. min ), mean % JO2max, and RER during the LT rides were 3.02
%, 3.64%, and 3.53 %, respectively. These data suggest
that trained cyclists have the ability to reproduce endurance performance with a CV of approximately 1.0% in a timetrial protocol. The observation that performance was significantly improved in LT-4 while no differences were observed between LT 1—3 suggests that psychological factors (i.e., the
last ride") may contribute to significant improvements in
endurance performance independent of experimental manipulations.
Key words
____________
Exercise performance, cycling, time trial,
reproducibility
(time-trial oriented) protocols (17). We believe that the assessment of exercise performance through the use of time-trial type protocols represents a logical extension of the available data on carbohydrate supplementation and exercise capacity, but feel
that the efficacy of this type of protocol must be established prior to its widespread application. Thus, the intent of this investigation was to examine the reliability of time-trial cycling performance in a laboratory setting. Methods
__________
Eight well-trained male cyclists served as subjects in this investigation. The men had participated in experiments in this laboratory in the past, and were familiar with the experimental procedures. All subjects were licensed, competi-
tive cyclists (USCF Cat. 2/3, Tn-Fed.). Prior time trial experience and personal bests at a standard (40 km) distance varied in the subject pool. All subjects were currently riding between 350 and 500km week- at the time of this investigation. After being fully informed of the risks and stresses associated with the project, the subjects gave their written consent
Int. J. Sports Med. 13 (1992) 467—470 Georg Thieme Verlag Stuttgart New York
to participate. Approval of the institutional review board for human subjects research was secured prior to the initiation of
Downloaded by: Queen's University. Copyrighted material.
Accepted after revision: May 5, 1992
468 mt. .1 Sports Med. 13 (1992)
M. S. Hickey, D. L. Costill, U. K. McConell, J I. Widrick, H. Tanaka Table 1 Coefficient of Variation in Performance Time and Work Output Variable
Time KJ Watts
0.94
V02 (I min l) V02 (% max)
LT
0.50 % 3.02 % 3.64 %
MT
0.53 %
ST
1.35 % * *
RER
interval (KM
Data are expressed as the mean of each individual's CV for the variables shown. KJ = total kilojoules per ride. * Expired gas was not collected in these trials. + p < 0.05 vs LT, MT. Watts represents the mean power Output in each trial.
data collection. The age, height, weight, and O2max for the subjects averaged (± SE): 22.1 (± 1.2) yrs., 180 (± 1.1) cm, 73,9
(±4.9)kg, and 4.6(±0.2)l.min'. A VO2max test was performed on a Cybex cycle ergometer (MET 100). A standard two-stage warm-up consisting of 5 minutes at 150 watts and 3 minutes at 250 watts was immediately followed by a graded exercise test which consisted of 25 W increments each minute. Expired gas was collected and analyzed using a computer-based system. VO2max
was taken as the average of the two highest consecutive data points. In all tests, VO2max plateaued with increasing workloads.
Subjects completed four rides at each of three total work outputs (i.e., 1600, 200, and l4kJ, which were equivalent to 40, 5 and 0.5 miles as determined by the ergometer). All trials were completed on an isokinetic controlled cycle ergometer (Cybex MET 100) which enabled the subject to selfpace each trial. Subjects were able to adjust cadence in 5rpm increments during the course of the test, and all such changes were recorded. The objective of the test was to finish as quickly as possible, with time being the criterion measure of perform-
ance. All trials were initiated with a standardized 10 minute warm-up at 150 watts. Bouts were conducted under controlled environmental conditions (22.5±0.5 'C, 46.6±3.6% RET). In order to limit external variables which might affect performance, the ergometer screen was covered in such a way that subjects could view only the cumulative work completed. Additionally, the subjects were unaware of elapsed time either during or after the trials. Thus, subjects completed all bouts with no reference to time and no knowledge of pacing. For the purposes of this investigation, we believed that every effort to minimize the number of confounding variables was necessary in order to assess the cyclist's capacity to reproduce performance without the influence of knowledge of pacing relative to other bouts, etc. The trials were completed on separate days, with a minimum of 72 hrs between bouts, and were conducted at the same time of day for each subject. Exercise was restricted
the day prior to each ride. In addition, prior to the 1600 kJ, or long trial (LT) rides only, subjects received a carbohydrate supplement (236gm. day-') that was consumed in addition to the normal diet for the 48 hr period prior to each trial, in an attempt to normalize muscle glycogen levels.
During each LT ride, subjects received a total of 150 ml of cool tap water every 15 minutes in an attempt to standardize hydration state. Heart rate was monitored telemetrically and recorded at 5 minute intervals throughout the ride. Split times and whole body ratings of perceived exertion using the Borg scale (2) were taken at the equivalent of 200 kJ intervals as determined by the ergometer. Similarly, expired gas was collected in Douglas bags at 200, 400, 600, 800, 1000, 1200, and 1400 kJ during each ride and analyzed to determine oxygen uptake, mean percentage of TO2max, and RER for each ride. Approximately two weeks after the completion
of the LT rides, subjects returned to the laboratory and completed a series of short performance rides. The medium trial (MT) and short trial (ST) rides were conducted as described for
the LT trials, with the exception that no expired gas was collected in these trials. The MT and ST rides were completed in blocks, i.e., all four ST rides, then all four MT rides, or vice versa. The order of these blocks were randomized. For all trials, work output was calculated from the equation MPH = (watt!
2.277)2 / (Cybex, Inc. calibration data). Data were analyzed using a repeated measures ANOVA. When significant F ratios were obtained, a Tukey's post-hoc test was used to locate these differences. Statistical significance was accepted at p
