To the Editors:
Cardiopulmonary exercise testing (CPET) has evolved as a useful tool for evaluation of exercise capacity in apparently healthy subjects. In the clinical arena, CPET is widely used to judge exercise “normalcy” in individuals with a suspected disorder and patients with several potential causes of exercise limitation 1. In order to establish “abnormality”, however, it is crucial to obtain representative frames of reference to interpret the (lack of) appropriateness of the systemic responses to exertion. Unfortunately, this is not a trivial task in the nonathletic subject as there are multiple confounding factors, especially behavioural characteristics such as the level of regular physical activity. Not surprisingly, there are only a few sets of reference values for clinical CPET interpretation which have survived the proof of time and are currently used worldwide 1.
In this context, the study recently published in the European Respiratory Journal by Koch et al. 2 is welcomed. After evaluating a large sample of apparently healthy males and females with a wide range of age and body dimensions, Koch et al. 2 provided a comprehensive set of prediction equations for the main CPET variables. The authors carefully avoided some well-known confounding factors, especially those related to past or current medical conditions, and the statistical analysis was unusually sophisticated.
There is, however, one major shortcoming in the study by Koch et al. 2 which might hamper its application in clinical practice. Unfortunately, the participants were not randomly selected from the general population, i.e. they freely volunteered to the study as part of a larger investigation on health-related outcomes in Germany. Consequently, it is conceivable that the more active subjects participated, an effect that is likely to be more relevant for the older groups. In fact, the authors stated that “the influence of physical activity was not consistently significant throughout the investigated groups” 2, which suggests that the elderly group were as active as the younger subjects. The hypothesis that the study has been biased to evaluate subjects who were more active than the sedentary, general population is consistent with the higher prevalence of nonsmoking and nonhypertensive subjects in the group of volunteers compared with the complete population (p<0.05). Moreover, the age-related decline in peak oxygen uptake (V′O2) was lower than previously reported by most of the previous studies and the predicted values for subjects aged >40 yrs were systematically higher than estimated by other commonly used equations. Therefore, age-related decline in predicted V′O2 from age 20–25 to 65 yrs has been previously estimated to average 20–25% in sedentary subjects; in contrast, Koch et al. 2 reported only a 15% decrease in males and females. As a consequence, figure 5 from the study by Koch et al. 2 shows that the median peak V′O2 values predicted by three other equations for subjects aged ≥65 yrs were in the lower quartiles or close to the 5th percentile in males and females, respectively. Collectively, these findings seem to indicate that the reference values of Koch et al. 2 are of limited value for the evaluation of exercise normalcy in the specific sub-population of sedentary elderly subjects in whom cardiopulmonary diseases are more prevalent 3 and CPET could be clinically more useful.
We have previously reported the findings of a similar, albeit smaller, study in which the subjects were randomly selected from a database of >8,000 subjects 4. Although this study feature increased enormously the complexity of the investigation, it eventually proved essential to obtain truly representative data for clinical interpretation of CPET. For instance, occasional volunteers were submitted to the same evaluation protocol but results were not considered on the final analysis. As expected, they were more active, fitter and leaner than the randomised subjects. In fact, if data from the nonrandomised subjects had been included in the analysis, predicted peak V′O2 values would be almost 20% higher, i.e. values quite similar to those reported by Koch et al. 2. In our view, this is the main reason that explains why our prediction equations provide the lowest peak V′O2 values amongst other sources of reference values, e.g. fig. 5 from Koch et al. 2. Moreover, we developed reference values for several effort-independent, submaximal relationships obtained in the incremental phase of exercise 5. Considering that these variables are less influenced by maximal aerobic capacity, Koch et al. 2 might have excellent material in their hands to further contribute to the field.
In conclusion, it is our opinion that the “ideal” set of reference values for clinical interpretation of CPET is still to be generated. Although such investigation would certainly share many of the characteristics of the study by Koch et al. 2, the issue of subject randomisation will be critical to improve our confidence on the limits of normality for key CPET variables.
Statement of interest
None declared.
- © ERS Journals Ltd
