10.14288/1.0077409Carr, Robin VictorCritical time in the PWC 170 test: the influences of work load duration, work load intensity, and state of trainingThe University of British Columbia1980enArchivalObject1The purpose of this study was to examine the concept of "critical time" (i.e. the time required to achieve steady-state heart rates) in the administration of the PWG 170 three-stage submaximal bicycle ergometer test. Specifically, the problem involved determining the effects of four different work load duration protocols on 'D' scores (which reflected the relative attainment of steady-state heart rates) and on PWG 170 scores. The combined effect of the order and relative intensity of the work loads on the 'D' scores was also studied, as was the effect of state of training on both 'D* scores and PWG 170 scores. Eight endurance-trained and eight untrained college males, aged 18 to 30, took a preliminary test to verify placement into their groups and to determine work loads for the experimental study. Each subject then underwent four experimental PWG 170 tests. Each test consisted of three periods of bicycle ergometer work of increasing intensity with the duration of the work period set at 3, 4, 5, or 6 minutes for the four test variations. There was an interval of at least two days between tests, which were administered in a counterbalanced Latin-square design. The pedalling cadence on the Monark bicycle ergometer was 50 r.p.m., and the warm-up consisted of ‘0’ work load for two minutes. Continuous monitoring of the subject's E.K.G. permitted calculationsoffaverage heart rate for every 15-second interval. These heart rates, and the associated work loads, provided the raw data for this study. Linear regression was used to determine the PWG 170 scores, while an asymptotic regression program was chosen to predict steady-state heart rate for each subject at each work load in all four tests. These predicted steady-state heart rates were then subtracted from the last 15-second average heart rates for all work periods to yield a 'D' score. This 'D* score then gave an indication of the extent to which steady-state heart rates were achieved. The hypotheses were tested through the use of two-way and three-way ANOVA's and preplanned orthogonal comparisons. The original analysis showed a trend toward increasing PWG 170 scores with shorter duration work periods, but the effect was not significant at the .05 level. However, after a careful analysis of the results, one of the trained subjects was classified as an "outlier" (one whose data contributes too much variance to be considered representative) and another ANOVA was run with this subject's aberrant data deleted. The statistical results were now very different, with the protocols effect highly significant (p<.001), and explained well by a linear function (p<.00l). On the basis of these ambiguous findings, confident conclusions regarding the protocols effect must await further study. The first ANOVA showed no evidence of an interaction effect between state of training and the protocols effect, however the second, 'post hoc ANOVA' (with subject "outlier" deleted) found a significant difference which suggested that trained athletes may have their PWG 170 scores overestimated more than untrained subjects as a result of shorter duration protocols. There was a highly significant protocols effect in the 'D' scores (p<.00l) which was explained almost entirely by a linear function (p<.00l). This data therefore tends to support the 'post hoc ANOVA' for the PWG 170 scores, since these scores are obviously dependent on the extent to which steady-state has been achieved. Although the 'D' scores suggested that the 4-minute protocol might be optimal for achieving steady-state values, this assumes that an asymptotic first-order model accurately predicts steady-state heart rates. In this light, the lack of a significant effect of state of training or work load number/intensity on critical time, shown by this study, must be interpreted with caution. Further study with 'D' scores based on second-order models may uncover significant main and interaction effects.