James E Peterman1, Matthew P Harber2, Mary T Imboden3, Mitchell H Whaley4, Bradley S Fleenor2, Jonathan Myers5, Ross Arena6, Leonard A Kaminsky1. 1. Fisher Institute of Health and Well-Being, Ball State University, Muncie, IN. 2. Clinical Exercise Physiology Laboratory, Ball State University, Muncie, IN. 3. Health and Human Performance Department, George Fox University, Newberg, OR. 4. College of Health, Ball State University, Muncie, IN. 5. Division of Cardiology, Veterans Affairs Palo Alto Healthcare System and Stanford University, Palo Alto, CA. 6. Department of Physical Therapy, College of Applied Science, University of Illinois at Chicago, Chicago, IL.
Abstract
Equations are often used to predict cardiorespiratory fitness (CRF) from submaximal or maximal exercise tests. However, no study has comprehensively compared these exercise-based equations with directly measured CRF using data from a single, large cohort. PURPOSE: This study aimed to compare the accuracy of exercise-based prediction equations with directly measured CRF and evaluate their ability to classify an individual's CRF. METHODS: The sample included 4871 tests from apparently healthy adults (38% female, age 44.4 ± 12.3 yr (mean ± SD)). Estimated CRF (eCRF) was determined from 2 nonexercise equations, 3 submaximal exercise equations, and 10 maximal exercise equations; all eCRF calculations were then compared with directly measured CRF, determined from a cardiopulmonary exercise test. Analysis included Pearson product-moment correlations, standard error of estimate values, intraclass correlation coefficients, Cohen κ coefficients, and the Benjamini-Hochberg procedure to compare eCRF with directly measured CRF. RESULTS: All eCRF values from the prediction equations were associated with directly measured CRF (P < 0.01), with intraclass correlation coefficient estimates ranging from 0.07 to 0.89. Although significant agreement was found when using eCRF to categorize participants into fitness tertiles, submaximal exercise equations correctly classified an average of only 51% (range, 37%-58%) and maximal exercise equations correctly classified an average of only 59% (range, 43%-76%). CONCLUSIONS: Despite significant associations between exercise-based prediction equations and directly measured CRF, the equations had a low degree of accuracy in categorizing participants into fitness tertiles, a key requirement when stratifying risk within a clinical setting. The present analysis highlights the limited accuracy of exercise-based determinations of eCRF and suggests the need to include cardiopulmonary measures with maximal exercise to accurately assess CRF within a clinical setting.
Equations are often used to predict cardiorespiratory fitness (CRF) from submaximal or maximal exercise tests. However, no study has comprehensively compared these exercise-based equations with directly measured CRF using data from a single, large cohort. PURPOSE: This study aimed to compare the accuracy of exercise-based prediction equations with directly measured CRF and evaluate their ability to classify an individual's CRF. METHODS: The sample included 4871 tests from apparently healthy adults (38% female, age 44.4 ± 12.3 yr (mean ± SD)). Estimated CRF (eCRF) was determined from 2 nonexercise equations, 3 submaximal exercise equations, and 10 maximal exercise equations; all eCRF calculations were then compared with directly measured CRF, determined from a cardiopulmonary exercise test. Analysis included Pearson product-moment correlations, standard error of estimate values, intraclass correlation coefficients, Cohen κ coefficients, and the Benjamini-Hochberg procedure to compare eCRF with directly measured CRF. RESULTS: All eCRF values from the prediction equations were associated with directly measured CRF (P < 0.01), with intraclass correlation coefficient estimates ranging from 0.07 to 0.89. Although significant agreement was found when using eCRF to categorize participants into fitness tertiles, submaximal exercise equations correctly classified an average of only 51% (range, 37%-58%) and maximal exercise equations correctly classified an average of only 59% (range, 43%-76%). CONCLUSIONS: Despite significant associations between exercise-based prediction equations and directly measured CRF, the equations had a low degree of accuracy in categorizing participants into fitness tertiles, a key requirement when stratifying risk within a clinical setting. The present analysis highlights the limited accuracy of exercise-based determinations of eCRF and suggests the need to include cardiopulmonary measures with maximal exercise to accurately assess CRF within a clinical setting.
Authors: Matthew P Harber; McKenzie Metz; James E Peterman; Mitchell H Whaley; Bradley S Fleenor; Leonard A Kaminsky Journal: PLoS One Date: 2020-12-01 Impact factor: 3.240
Authors: Robert A Sloan; Marco V Scarzanella; Yuko Gando; Susumu S Sawada Journal: Int J Environ Res Public Health Date: 2021-11-23 Impact factor: 3.390
Authors: Scott B Maitland; Paula Brauer; David M Mutch; Dawna Royall; Doug Klein; Angelo Tremblay; Caroline Rheaume; Rupinder Dhaliwal; Khursheed Jeejeebhoy Journal: Nutrients Date: 2021-11-26 Impact factor: 5.717