Constant Work Rate Exercise Testing: A Tricky Measure of Exercise Tolerance

Constant work rate exercise testing has become a popular way to study exercise tolerance in COPD patients. Unlike field tests such as the 6-minute walk and shuttle walk tests, recording of cardiopulmonary responses is facilitated. While incremental exercise testing probes the limits of tolerance, constant work rate testing investigates a work rate domain more likely to be encountered in everyday life.

The reason for the rising popularity of constant work rate testing stems, however, from a specific use: to assess the response to interventions posited to improve exercise tolerance. As was well shown by Oga et al. in 2000 ([1]), for a given intervention (in this case, bronchodilator administration) the fractional improvement in constant work rate exercise duration was much greater than the fractional improvement in 6-minute walk test duration or in fractional increase in peak oxygen uptake in an incremental exercise test. (19% vs. 1% vs. 3%, respectively). Who wouldn't want to show a favored intervention in the best light? However, this increased sensitivity to interventions comes at a cost. It is necessary to do a preliminary study in order to choose an individualized work rate that will result in comparable exercise durations among individuals; this allows the response to the intervention to be observed from a similar baseline.

Constant work rate exercise duration is dictated by power-duration considerations ([2],3). There is a hyperbolic relationship between exercise duration and exercise work rate. This hyperbola is determined by the critical power, the work rate below which exercise can be tolerated essentially indefinitely and by the curvature constant, that determines how rapidly exercise duration decreases once the critical power is exceeded. If the critical power and the curvature constant are known, a work rate can be chosen to elicit any chosen constant work rate duration (with the limitation that day-to-day variability in physiologic capabilities and effort will induce some degree of imprecision). However, determining the critical power and curvature constant for an individual is a rather laborious procedure, requiring several constant work rate tests for their determination (but see ([4])). A simpler procedure, and the one that has been generally used, is to individualize the work rate as a percentage of the peak work rate in an incremental exercise test; in studies of the COPD population, a percentage in the range of 75% to 85% has generally been used ([1],5,6,7). The simplicity flows from needing only a single exercise test to establish the work rate for the constant work rate test.

Using the percent of peak work rate as a normalizing strategy to in an attempt to assure similar constant work rate durations among individuals is not a particularly effective strategy, however. Reasons for this include:

• Peak work rate depends on the way in which the incremental test is performed; for a given peak oxygen uptake a faster work rate increase yields a higher peak work rate ([8]) and therefore a higher work rate at a given percentage of peak work rate.

• Ventilatory and gas exchange kinetics influence peak work rate. For example, a ventilatory-limited COPD patient will exercise to a higher peak work rate if ventilatory kinetics are slow.

• The ventilatory requirement at a given work rate in ventilatory limited patients determines both the peak work rate and the critical power. In those with lower ventilatory requirement (e.g., those with lower V_D/V_T), both the peak work rate and the critical power will be higher and the critical power's percentage of peak work rate will likely differ.

• The relation between peak work rate and critical power varies among individuals. Importantly, in healthy individuals critical power is a lower fraction of peak work rate than in COPD patients ([9]). When exercising at a given percentage of peak work rate, those with a high critical power as a fraction of peak work rate will be able to exercise for a longer period.

The foregoing considerations dictate that individuals exercising at a given percentage of peak work rate will display a wide range of tolerated exercise durations. This complicates the interpretation of changes in constant work rate duration. This is because, for an intervention that promotes endurance, the longer the initial duration of the constant work rate test, the greater the increase this intervention will elicit. This is true whether the increase in duration is expressed in absolute terms or as a fraction of the initial duration.

Laviolette et al. ([10]) in this volume utilize constant work rate duration for a different purpose than determining responses to interventions. They interpret systematic differences in constant work rate exercise durations between two groups of COPD patients (specifically between men and women) to infer differences in physiologic capabilities. When examining responses to work rates chosen as a fixed percentage (70%) of peak work rate, they find that women have systematically shorter exercise durations. The paper's key conclusion is that COPD women are disadvantaged relative to COPD men and might have difficulties in performance of everyday activities that require endurance capacity. But is this an appropriate inference?

The difference in constant work rate exercise duration that was detected can be seen to be a particularly unreliable metric of fitness for endurance exercise tasks. Consider that had a group of healthy men (or women) been the group compared to the COPD men, they almost certainly would have demonstrated a shorter constant work rate duration if they exercised at the same fraction of their peak work rate in an incremental exercise test. This can be seen in the work of Neder et al. ([9]), who compared the power-duration relationships of COPD to healthy men. Critical power averaged 65 watts in the COPD men and 110 watts in the healthy men, but this represented 81.8% and 67.5% of peak work rate, respectively. Figure 1 displays the power-duration curves calculated from the average of the individual critical power and curvature constant parameters from the group of healthy subjects and COPD patients studied. Note that the curve derived for the COPD patients indicates that, had a work rate of 70% of peak work rate (56 watts) been imposed it would have been below the critical power and exercise could have continued essentially indefinitely.

Figure 1 The power duration relationships derived from data collected by Neder et al. ([9]) from study of a group of 8 male COPD patients and a group of 10 age-matched healthy male controls. The curves were constructed from the average values of the critical power (CP) and the curvature constants (W') of the individuals in the two groups. Horizontal and vertical lines indicate the exercise durations (T_lim) that would be predicted in response to a constant work rate (WR) equal to 90% of the peak work rate in the two groups. Note that, despite a much lower peak work rate and lower critical power, the tolerated duration at 90% of peak work rate is much greater in the COPD group.

Figure 1 compares the predicted responses to 90% of peak work rate; the healthy subjects would have exercised for 281 seconds while COPD patients would have exercised for 980 seconds. Thus the healthy men, who undoubtedly would be judged to have superior exercise tolerance than COPD men, manifest a much shorter constant work rate exercise duration when exercising at a given fraction of their peak work rate. It seems to follow that whether women with COPD are disadvantaged in their endurance exercise tolerance in comparison to men cannot be assessed solely by this measure.

We may conclude that constant work rate exercise testing is a useful, but tricky, way of assessing exercise capabilities. As is well shown in the work of Laviolette et al. ([10]), it is possible to use constant work rate exercise testing to obtain detailed physiologic insights into the nature of exercise intolerance. However, the subtleties of the power-duration relationship can, at times, complicate interpretation of exercise capabilities.

Declaration of interest

The authors report no conflict of interest. The authors alone are responsible for the content and writing of the paper.