CoChat

Cardiorespiratory fitness assessment using risk-stratified exercise testing and dose-response relationships with disease outcomes

Open access

Published version of a preprint

The DOI matches its Crossref record; 54 of the 54 references it cites were found in a registry or an index.

Abstract

Cardiorespiratory fitness (CRF) is associated with mortality and cardiovascular disease, but assessing CRF in the population is challenging. Here we develop and validate a novel framework to estimate CRF (as maximal oxygen consumption, VO 2 max) from heart rate response to low-risk personalised exercise tests. We apply the method to examine associations between CRF and health outcomes in the UK Biobank study, one of the world’s largest and most inclusive studies of CRF, showing that risk of all-cause mortality is 8% lower (95%CI 5–11%, 2670 deaths among 79,981 participants) and cardiovascular mortality is 9% lower (95%CI 4–14%, 854 deaths) per 1-metabolic equivalent difference in CRF. Associations obtained with the novel validated CRF estimation method are stronger than those obtained using previous methodology, suggesting previous methods may have underestimated the importance of fitness for human health.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41598-021-94768-3

Cites

54 of 54 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Conceptual framework and design for validation study. (A) Overview of the five exercise tests performed by validation study participants (3 UKB-CRF tests (flat protocol, low ramp protocol, high ramp protocol), 1 steady-state test unique to the validation study, and 1 maximal exercise test to measure VO2max). X-axes: Time; Y-axes: Work rate (WR). Tests were completed consecutively, and work rates were individualised according to standardised criteria (See `Experimental procedure and equipment’ in Methods). UKB-CRF test and steady-state test data were used for method development. Maximal exercise test data were withheld from method development and used for validation purposes only. (B) Conceptual plot of WR-to-VO2 response during steady-state and ramped exercise tests. VO2 increases linearly at a rate proportional to the rate of change in WR (i.e. ramp rate) until VO2max is reached (in an exhaustive test). The WR-to- VO2 relationship (line slope) changes depending on the ramp rate of the test. As ramp rate decreases, the WR when VO2max is achieved approaches the maximal WR for an exhaustive steady-state test. Note that VO2 is extrapolated to maximal values for demonstrative purposes, but in the validation study ramped and steady-states tests were non-exhaustive. (C) Exemplar HR data (blue scatter and grey line; upper panel), WR data (red line; lower panel), and test phase annotation for ramp test. (D,E) Feature extraction for ramp phase using simple linear regression model and for recovery phase using first-order exponential decay model (see Supplemental Methods).
Figure 1 Conceptual framework and design for validation study. (A) Overview of the five exercise tests performed by validation study participants (3 UKB-CRF tests (flat protocol, low ramp protocol, high ramp protocol), 1 steady-state test unique to the validation study, and 1 maximal exercise test to measure VO2max). X-axes: Time; Y-axes: Work rate (WR). Tests were completed consecutively, and work rates were individualised according to standardised criteria (See `Experimental procedure and equipment’ in Methods). UKB-CRF test and steady-state test data were used for method development. Maximal exercise test data were withheld from method development and used for validation purposes only. (B) Conceptual plot of WR-to-VO2 response during steady-state and ramped exercise tests. VO2 increases linearly at a rate proportional to the rate of change in WR (i.e. ramp rate) until VO2max is reached (in an exhaustive test). The WR-to- VO2 relationship (line slope) changes depending on the ramp rate of the test. As ramp rate decreases, the WR when VO2max is achieved approaches the maximal WR for an exhaustive steady-state test. Note that VO2 is extrapolated to maximal values for demonstrative purposes, but in the validation study ramped and steady-states tests were non-exhaustive. (C) Exemplar HR data (blue scatter and grey line; upper panel), WR data (red line; lower panel), and test phase annotation for ramp test. (D,E) Feature extraction for ramp phase using simple linear regression model and for recovery phase using first-order exponential decay model (see Supplemental Methods).CC BY · PubMed Central
Figure 2 Scatterplots (top row) and Bland–Altman plots (bottom row) demonstrating agreement between work rates measured at the respiratory compensation point (RCP) and work rates estimated from flat ramp tests (left column), low ramp tests (middle column), and high ramp tests (right column) using the most comprehensive prediction equation from the multilevel framework (M1 for ramp tests; M4 for flat test). r: Pearson’s correlation coefficient, rho: Spearman’s rank correlation coefficient. RMSE: Root-mean-square error.
Figure 2 Scatterplots (top row) and Bland–Altman plots (bottom row) demonstrating agreement between work rates measured at the respiratory compensation point (RCP) and work rates estimated from flat ramp tests (left column), low ramp tests (middle column), and high ramp tests (right column) using the most comprehensive prediction equation from the multilevel framework (M1 for ramp tests; M4 for flat test). r: Pearson’s correlation coefficient, rho: Spearman’s rank correlation coefficient. RMSE: Root-mean-square error.CC BY · PubMed Central
Figure 3 Scatterplots (top row) and Bland–Altman plots (second row) demonstrating agreement between directly measured VO2max and VO2max estimated from flat tests (left column), low ramp tests (middle column), and high ramp tests (right column) using the most comprehensive equation from the multilevel framework (M1 for ramp tests; M4 for flat test). Below these (bottom row), the box plot demonstrates agreement across all ramp rates tested using estimates from the multilevel framework, the simple linear regression method, and the two-point estimation method. r: Pearson’s correlation coefficient, rho: Spearman’s rank correlation coefficient. RMSE: Root-mean-square error.
Figure 3 Scatterplots (top row) and Bland–Altman plots (second row) demonstrating agreement between directly measured VO2max and VO2max estimated from flat tests (left column), low ramp tests (middle column), and high ramp tests (right column) using the most comprehensive equation from the multilevel framework (M1 for ramp tests; M4 for flat test). Below these (bottom row), the box plot demonstrates agreement across all ramp rates tested using estimates from the multilevel framework, the simple linear regression method, and the two-point estimation method. r: Pearson’s correlation coefficient, rho: Spearman’s rank correlation coefficient. RMSE: Root-mean-square error.CC BY · PubMed Central
Figure 4 Scatterplots (top row) and Bland–Altman plots (bottom row) demonstrating short- and long-term test–retest reliability. r: Pearson’s correlation coefficient, rho: Spearman’s rank correlation coefficient. RMSE: Root-mean-square error. Lamda: Regression-dilution coefficient. SE: Standard error. Short-term reliability data are from the validation study (n = 87, follow-up ~ 10 days) and long-term reliability data are from the repeat-measures sub-study in UK Biobank (n = 2877, follow-up ~ 2.8 years).
Figure 4 Scatterplots (top row) and Bland–Altman plots (bottom row) demonstrating short- and long-term test–retest reliability. r: Pearson’s correlation coefficient, rho: Spearman’s rank correlation coefficient. RMSE: Root-mean-square error. Lamda: Regression-dilution coefficient. SE: Standard error. Short-term reliability data are from the validation study (n = 87, follow-up ~ 10 days) and long-term reliability data are from the repeat-measures sub-study in UK Biobank (n = 2877, follow-up ~ 2.8 years).CC BY · PubMed Central
Figure 5 Hazard ratio and 95% confidence interval (CI) for prospective log-linear associations (Cox regression) between fatal and non-fatal outcomes in the UK Biobank with cardiorespiratory fitness in metabolic equivalents (METs, per 3.5 ml O2·kg−1 min−1) computed from the multilevel framework and simple linear regression methods. Event-rate per 100,000 person-years. AF: atrial fibrillation; COPD: chronic obstructive pulmonary disease; CVD: cardiovascular disease; IHD: ischaemic heart disease; RD: respiratory disease. COPD incidence mostly reflects severe COPD since only ~ 25% of cases end up in hospital. Mortality and incidence event-rates differ between fitness prediction methods owing to different inclusion criteria at the estimation level.
Figure 5 Hazard ratio and 95% confidence interval (CI) for prospective log-linear associations (Cox regression) between fatal and non-fatal outcomes in the UK Biobank with cardiorespiratory fitness in metabolic equivalents (METs, per 3.5 ml O2·kg−1 min−1) computed from the multilevel framework and simple linear regression methods. Event-rate per 100,000 person-years. AF: atrial fibrillation; COPD: chronic obstructive pulmonary disease; CVD: cardiovascular disease; IHD: ischaemic heart disease; RD: respiratory disease. COPD incidence mostly reflects severe COPD since only ~ 25% of cases end up in hospital. Mortality and incidence event-rates differ between fitness prediction methods owing to different inclusion criteria at the estimation level.CC BY · PubMed Central
Figure 6 Hazard ratio and 95% confidence interval (CI) for nonlinear associations (cubic splines, Cox regression) between fatal and non-fatal outcomes in the UK Biobank with cardiorespiratory fitness in metabolic equivalents (METs, per 3.5 ml O2·kg−1·min−1) computed from the multilevel framework and simple linear regression methods. Hazard ratios were computed relative to a fitness reference point of 8.0 METs. AF: atrial fibrillation; COPD: chronic obstructive pulmonary disease; CVD: cardiovascular disease; IHD: ischaemic heart disease; RD: respiratory disease. Mortality and incidence counts (superimposed histograms) differ between fitness estimation methods owing to different inclusion criteria at the estimation level.
Figure 6 Hazard ratio and 95% confidence interval (CI) for nonlinear associations (cubic splines, Cox regression) between fatal and non-fatal outcomes in the UK Biobank with cardiorespiratory fitness in metabolic equivalents (METs, per 3.5 ml O2·kg−1·min−1) computed from the multilevel framework and simple linear regression methods. Hazard ratios were computed relative to a fitness reference point of 8.0 METs. AF: atrial fibrillation; COPD: chronic obstructive pulmonary disease; CVD: cardiovascular disease; IHD: ischaemic heart disease; RD: respiratory disease. Mortality and incidence counts (superimposed histograms) differ between fitness estimation methods owing to different inclusion criteria at the estimation level.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 54 references

Research works printed but not linked: 2

  1. Pescatello, L. S., Arena, R., Riebe, D., Thompson, P. D. & American College of Sports Medicine Health-related physical fitness testing and interpretation. In ACSM’s Guidelines for Exercise Testing and Prescription (eds Pescatello, L. S. et al.) 88–93 (Lippincott Williams & Wilkins, 2014).
  2. Engelse, W. A. H. & Zeelenberg, C. A single scan algorithm for QRS-detection and feature extraction. Comput. Cardiol. 6, 37–42 (1979).

Cited by

See all 60

Metadata

Authors
Gonzales, T. I., Westgate, K., Strain, T., Hollidge, S., Jeon, J., Christensen, D. L., Jensen, J., Wareham, N. J., & Brage, S.
Affiliations
  • University of Cambridge (GB)
  • Yonsei University (KR)
  • University of Copenhagen (DK)
Corresponding author
Brage, Søren
Venue
Scientific Reports
Published
2021-07-28
Volume
11
Issue
1
Pages
15315-15315
ISSN
2045-2322 (ISSN-L)
OpenAlex
W3184192324
Cited by
59
References
54
Citations by year
2026: 12 · 2025: 15 · 2024: 18 · 2023: 7 · 2022: 6
FWCI
8.73
Type
journal article
Language
en
Open access
Licence
Locations
  • Scientific Reports (publishedVersion, cc-by)
  • medRxiv (acceptedVersion, cc-by-nd)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Research at the University of Copenhagen (University of Copenhagen) (publishedVersion, cc-by)
  • Apollo (University of Cambridge) (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • NORA - Norwegian Open Research Archives (submittedVersion, cc-by)
  • Edinburgh Research Explorer (University of Edinburgh) (submittedVersion, cc-by)
  • Apollo
Funders
  • National Institute for Health and Care Research ([IS-BRC-1215-20014, BRC1215-20014, MC_UU_12015/3, IS-BRC-1215)
  • British Heart Foundation (CH/1996001/9454)
  • Medical Research Council (IS-BRC-1215-20014, MC_UU_12015, [MC_UU_00006/4], MC_UU_12015/3, MC_PC_17228, BRC-1215-20014)
Topics
  • Health Sciences › Medicine › Physiology (medical) › Cardiovascular and exercise physiology
  • Health Sciences › Medicine › Public Health, Environmental and Occupational Health › Physical Activity and Health
  • Health Sciences › Medicine › Cardiology and Cardiovascular Medicine › Cardiovascular Effects of Exercise
Keywords
cardiorespiratory fitness, heart rate response, all-cause mortality, cardiovascular mortality, metabolic equivalents, dose-response relationship, UK Biobank, physical fitness assessment
MeSH terms
Cardiorespiratory Fitness, Cause of Death, Exercise Test, Female, Heart Diseases, Heart Rate, Humans, Male, Middle Aged, Neoplasms, Obesity, Oxygen Consumption, Respiration Disorders, Proportional Hazards Models

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Cardiorespiratory fitness assessment using risk-stratified exercise testing and dose-response relationships with disease outcomes. https://app.cochat.ai/papers/10.1038/s41598-021-94768-3
The paper
Gonzales, T. I., Westgate, K., Strain, T., Hollidge, S., Jeon, J., Christensen, D. L., Jensen, J., Wareham, N. J., & Brage, S. (2021). Cardiorespiratory fitness assessment using risk-stratified exercise testing and dose-response relationships with disease outcomes. Scientific Reports. https://doi.org/10.1038/s41598-021-94768-3

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.