RESUMO
BACKGROUND: Up-to-date normative reference sets for cardiopulmonary exercise testing (CPET) are important to aid in the accurate interpretation of CPET in clinical or research settings. RESEARCH QUESTION: This study aimed to (1) develop and externally validate a contemporary reference set for peak CPET responses in Canadian adults identified with population-based sampling; and (2) evaluate previously recommended reference equations for predicting peak CPET responses. STUDY DESIGN AND METHODS: Participants were healthy adults who were ≥40 years old from the Canadian Cohort Obstructive Lung Disease who completed an incremental cycle CPET. Prediction models for peak CPET responses were estimated from readily available participant characteristics (age, sex, height, body mass) with the use of quantile regression. External validation was performed with a second convenience sample of healthy adults. Peak CPET parameters that were measured and predicted in the validation cohort were assessed for equivalence (two one-sided tests of equivalence for paired-samples and level of agreement (Bland-Altman analyses). Two one-sided tests of equivalence for paired samples assessed differences between responses in the derivation cohort using previously recommended reference equations. RESULTS: Normative reference ranges (5th-95th percentiles) for 28 peak CPET parameters and prediction models for 8 peak CPET parameters were based on 173 participants (47% male) who were 64 ± 10 years old. In the validation cohort (n = 84), peak CPET responses that were predicted with the newly generated models were equivalent to the measured values. Peak cardiac parameters predicted by the previously recommended reference equations by Jones and colleagues and Hansen and colleagues were significantly higher. INTERPRETATION: This study provides reference ranges and prediction models for peak cardiac, ventilatory, operating lung volume, gas exchange, and symptom responses to incremental CPET and presents the most comprehensive reference set to date in Canadian adults who were ≥40 years old to be identified with population-based sampling.