Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Look up keyword
Like this
0Activity
0 of .
Results for:
No results containing your search query
P. 1
20- Versus 30-s Wingate Anaerobic Test

20- Versus 30-s Wingate Anaerobic Test

Ratings: (0)|Views: 4 |Likes:
Published by Angel Petrov

More info:

Published by: Angel Petrov on Nov 06, 2012
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

11/06/2012

pdf

text

original

 
Eur J Appl Physiol (2007) 100:645–651DOI 10.1007/s00421-007-0454-3
 123
ORIGINAL ARTICLE
Cross-validation of the 20- versus 30-s Wingate anaerobic test
C. Matthew Laurent Jr. · Michael C. Meyers ·Clay A. Robinson · J. Matt Green
Accepted: 19 March 2007 / Published online: 12 April 2007
©
Springer-Verlag 2007
Abstract
The 30-s Wingate anaerobic test (30-WAT) isthe most widely accepted protocol for measuring anaerobicresponse, despite documented physical side e
ects. Abbre-viation of the 30-WAT without loss of data could enhancesubject compliance while maintaining test applicability.The intent of this study was to quantify the validity of the20-s Wingate anaerobic test (20-WAT) versus the tradi-tional 30-WAT. Fifty males (mean
§
SEM; age=20.5
§
0.3years; Ht=1.6
§
0.01m; Wt=75.5
§
2.6kg) wererandomly selected to either a validation (
 N 
=35) or cross-validation group (
 N 
=15) and completed a 20-WAT and30-WAT in double blind, random order on separate days todetermine peak power (PP; Wkg
¡
1
), mean power (MP;Wkg
¡
1
), and fatigue index (FI; %). Utilizing poweroutputs (relative to body mass) recorded during each sec-ond of both protocols, a non-linear regression equation(
20WAT+10
=31.4697 e
¡
0.5
[ln(
 X 
second
 /1174.3961)/2.6369
2
];
2
=0.97; SEE=0.56Wkg
¡
1
) successfully predicted(error
»
10%) the
W
nal 10s of power outputs in the cross-validation population. There were no signi
W
cant di
er-ences between MP and FI between the 20-WAT thatincluded the predicted 10s of power outputs (20-WAT+10)and the 30-WAT. When derived data were subjected toBland–Altman analyses, the majority of plots (93%) fellwithin the limits of agreement (
§
2SD). Therefore, whencompared to the 30-WAT, the 20-WAT may be considereda valid alternative when used with the predictive non-linearregression equation to derive the
W
nal power output values.
Keywords
Leg power · Work capacity · Cycle ergometry ·Sprint test
Introduction
The most commonly employed protocol for the measure-ment of anaerobic response is the 30-s Wingate anaerobictest (30-WAT). Developed almost 30years ago, the 30-WAT involves a maximal exertion bout on a cycle ergome-ter to evaluate peak power (PP), mean power (MP), andfatigue index (FI; Bar-Or1987; Bar-Or etal.1977). When performing a 30-WAT, a subject typically exhibits a sharprise in power output, reaching peak power within the
W
rstfew seconds. Typically, subjects are unable to maintain thisoutput, leading to an exponential decline in power through-out the remaining duration of the test (Bar-Or etal.1977;Marquardt etal.1993). During this prolonged period of maximal e
ort, the accumulation of [H
+
] and lactate asbyproducts of anaerobic glycolysis results in a drop inblood pH (lactic acidosis). The increased acidity impairsenzyme activity involved in energy metabolism andreduces maximal muscle
W
bre recruitment (Allen etal.1992; Davis1985). In addition, the acute increase of blood glucose as a substrate for glycolysis during maximal exer-cise can result in hypoglycemia (Vincent etal.2004).Together these responses may result in unwelcome physicalside e
ects, including fatigue, headache, dizziness, and
C. Matthew Laurent Jr. (
&
) · J. Matt GreenDepartment of Kinesiology, The University of Alabama,Box 83012, Moore Hall, Tuscaloosa, AL, 35487-0312, USAe-mail: cmlaurentjr@bama.ua.eduM. C. MeyersHuman Performance Research Laboratory,Department of Sports and Exercise Sciences,West Texas A&M University, Canyon, TX 79016, USAC. A. RobinsonDepartment of Agriculture, West Texas A&M University,Canyon, TX 79016, USA
 
646Eur J Appl Physiol (2007) 100:645651
 123
nausea (Jacobs etal.1982). Subject awareness of these sidee
ects during anaerobic testing may result in submaximale
ort, high attrition rates, or unsuccessful subject comple-tion of subsequent sport performance testing (Jacobs etal.1982; Marquardt etal.1993). In 1987, Vandewalle etal. reported high correlationcoe
Y
cients between data collected at the 20th and 30th sec-onds of a WAT. Consequently, they proposed to curtail theduration of the WAT in order to diminish the stress of thetest and the involvement of the aerobic metabolism. How-ever, this previous study did not compare the 20-WAT and30-WAT but the data collected at the 20th and 30th secondof the same test. Hence this previous study failed to provethat results of a 20-WAT would have been the same asresults collected during the
W
rst 20s of a 30-WAT becauseof a possible submaximal e
ort strategy during a stressful30-WAT. More recently, Marquardt etal. (1993) reportedhigh intraclass correlations (
>0.95) between results of a20 and 30 Wingate tests in 14 subjects and concluded “thata 20-s Wingate test may be a valid and less strenuous alter-native to the 30-s Wingate test.”Since its
W
rst publication, the 30-WAT has been used inhundreds of studies published in the international literatureand it would be interesting to be able to compare the resultsof a future 20-s anaerobic test with all these experimentaldata. Moreover, it is likely that many athletes have per-formed several traditional 30-WATs and, subsequently,could be reluctant to perform new tests whose results can-not be compared with previous scores. The prediction of 30-WAT indices from 20-WAT performances is not indi-vidual but statistical in the study by Vandewalle etal.(1987a). In that study, the statistical prediction of theamount of work performed at the 30th seconds (an equiva-lent of MP) from 20-s data was probably accurate(
=0.989), however, the regression coe
Y
cient betweenpedal rate at 20th and 30th was lower (
=0.882). In thepresent study we validate a 20-WAT by comparing the datacollected during 20 and 30s all-out exercises in 50 sub- jects. Moreover, we propose a method which enables anindividual and accurate prediction of 30-WAT indices fromthe data of a 20-WAT (20-WAT+10).
Methods
SubjectsThe participants for this study included 50 male, college-aged students (mean
§
SEM; age=20.5
§
0.3years;Ht=175.8
§
1.2cm; Wt=75.5
§
2.6kg) with no knowncardiovascular/pulmonary disease, metabolic disorders, ormedical contraindications to exercise as determined by self-response medical history form and interview. Subjects wererandomly selected students who volunteered from univer-sity level physical education classes. Subjects were encour-aged not to participate in any strenuous exercise during the24h prior to testing, and initially reported to the laboratory4h postabsorptive. Subjects were fully informed of thenature of the study and provided written, informed consentin accordance with the accepted guidelines of the Institu-tional Review Board of the university and the AmericanCollege of Sports Medicine (1997).ProceduresAll subjects initially reported to the laboratory whereuponeach individual had their height (m) and body mass (kg)determined using calibrated physician scales. Anaerobicresponse was quanti
W
ed utilizing both a 20-WAT and a 30-WAT, performed in a double blind, randomly determinedorder to measure peak power (Wkg
¡
1
), mean power(Wkg
¡
1
), and fatigue index (%). The maximum powerachieved during the
W
rst 5s of the test is de
W
ned as PP,while MP is de
W
ned as the average power achievedthroughout the trial, and the FI re
X
ects the percent powerdecline during the trial (Bar-Or1987; Bar-Or etal.1977). Both WAT protocols were conducted while pedaling acalibrated Monark model 824E cycle ergometer (Monark AB, Varberg, Sweden) with integrated laser-based sensorand computer software (Sports Medicine Industries, Inc.,St. Cloud, MN). Prior to both tests, each subject began awarm-up phase consisting of alternating three 30-s intervalsof active rest (pedaling against no resistance at 60rpm)with three 30-s intervals pedaling against increasing resis-tance of 25, 50, and 75% of test resistance (Vanderfordetal.2004). Prior research has indicated that this form of pre-test loading elicits optimal power production duringsupramaximal exercise (Burnley etal.2005). Test resis-tance was calculated by multiplying the subject’s bodymass (kg) by 0.075. Following completion of the warm-up,the subject continued to pedal at 60rpm with no resistancefor another 2min until initiation of the WAT. Following a10-s countdown, the resistance was immediately added andthe subject was verbally encouraged to pedal as fast as pos-sible for 20 or 30s. Relative power outputs (Wkg
¡
1
) weremeasured during each second of the two trials, and PP, MP,and FI were calculated according to accepted procedures(Bar-Or1987; Bar-Or etal.1977). Within each subject, the 20-WAT and 30-WAT werecompleted at the same time of day, no less than 48h and nomore than 7days apart in order to ensure optimal recoveryand minimize any confounding anthropometric changes,respectively. In accordance with previously publishedresearch regarding test–retest reliability of the WAT (Bar-Or1987; Granier etal.1995; Kaczkowski etal.1982; Vandewalle etal.1987a), a PP increase
¸
8% was considered
 
Eur J Appl Physiol (2007) 100:645651647
 123
to be a distorted response, presumably due to a learninge
ect. In these instances, the initial protocol was repeated48h later and used for comparison. In 13 instances, thedi
erence between the two protocols still exceeded thestated criteria and, consequently, these subjects were omit-ted from the study. To assess the presence of negative sidee
ects following testing, a basic approach was taken. Sub- jects were asked to provide a yes/no answer with respect topresence of nausea, light-headedness, leg fatigue, and/orany other physical side e
ects. This approach was advanta-geous in that it was not di
Y
cult to administer and permittedthe evaluation of side e
ects immediately proximal to thetest. This was important as side e
ects may often be acuteand transient and in such case a written survey completedby subjects after recovering from testing would provide lessvalid information.Statistical analysesData were subjected to multivariate analyses of variance(MANOVAs), where appropriate, utilizing the StatisticalPackage for Social Sciences (SPSS, Inc., Chicago, IL) todetermine a signi
W
cant main e
ect between protocols.When univariate post hoc procedures (AVOVA) indicatedsigni
W
cant di
erences between relative MP outputs and FI,the sample of 50 subjects were randomly assigned to eithera validation group (
 N 
=35) or cross-validation group(
 N 
=15). Both linear and non-linear regression analyseswere employed to develop an equation in order to deter-mine the best model to predict the
W
nal 10s of relativepower outputs. These values were derived from power out-puts obtained during seconds 11–20 of the 20-WAT rela-tive to those obtained during seconds 21–30 of the 30-WATfrom the validation group. The non-linear formulaemployed to
W
t the data was a three parameter, peak, log-normal equation of the form
=
a
e
¡
0.5
[(ln (
 X 
 / 
 X 
0
)/ 
b
)
2
],where
a
,
b
, and
 X 
0
are the parameters
W
tted, e
¡
0.5
is anexponential function, ln is the command to perform a nor-mal logarithmic procedure, and
 X 
is the second for whichthe relative power output is being determined. Followingregression analyses, the predicted
W
nal 10s were combinedwith the observed 20-WAT data (20-WAT+10) recorded inthe cross-validation group to demonstrate the validity andapplicability of the derived equation to a target population.ANOVAs were performed to identify signi
W
cant di
er-ences among the PP, MP, and FI of the 20-WAT+10 andthe 30-WAT relative power output. To further assess valid-ity and reliability, the Bland–Altman method of comparisonwas employed to determine the limits of agreement(mean
§
2SD) between the two protocols (Altman andBland1983; Bland and Altman1986) within the cross-vali- dation group. Statistical signi
W
cance was determinedapriori at the
P
=0.05 level. All data are presented asmean
§
SEM unless otherwise noted.
Results
Peak power, mean power, and fatigue indexThe mean relative power outputs recorded during the20-WAT as well as the 30-WAT along with their respective
P
values and percent di
erences are presented in Table1.As expected, a signi
W
cant MANOVA was found for theWilks’ Lambda rank variable (
3,80
=0.723;
P
<0.001;
n
¡
=0.998) between protocols. Subsequent ANOVAsindicated no signi
W
cant di
erences between relative PP out-puts recorded during the 30-WAT and the 20-WAT; how-ever, signi
W
cant di
erences were found with regard torelative MP and FI. Mean relative power outputs were sim-ilar throughout the
W
rst 20s of both trials, however, a sub-stantial power decline was observed during the
W
nal 10s of the 30-WAT (Fig.1).Linear versus non-linear regression prediction modelResults from linear and non-linear regression analysesrevealed that the non-linear model yield similar coe
Y
cientsof determination between models. There was an additionaltwo percent of variation observed was explained whenusing a non-linear versus a linear prediction model(
2
=0.97 vs.
2
=0.95, respectively). Additionally, anoticeable discrepancy was revealed in the standard error of the estimate between models, with the non-linear modelreporting an error of 
»
10% (SEE=0.56Wkg
¡
1
) whencompared to an error of 
»
15% (SEE=0.75Wkg
¡
1
) of thelinear model observed within the measured population.Consequently, the non-linear regression model wasemployed to validate and cross-validate the 20-WAT ver-sus 30-WAT.
Table1
Mean anaerobic powerand capacity for 20-WAT and30-WAT (
 N 
=50)Variable30-WATMean
§
SEM20-WATMean
§
SEM
P
Di
erence(%)Peak power (Wkg
¡
1
)6.8
§
0.26.7
§
0.20.541.5Mean power (Wkg
¡
1
)4.5
§
0.15.0
§
0.20.01*10.0Fatigue index (%)58.3
§
1.746.1
§
1.7<0.01*20.9*Signi
W
cant at
P
<0.05 level

You're Reading a Free Preview

Download
scribd
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->