Determinants of cardiorespiratory fitness measured by cardiopulmonary exercise testing in COVID-19 survivors: a systematic review with meta-analysis and meta-regression

Background: The relationship between cardiorespiratory fitness and its possible determinants in post-COVID-19 survivors has not been systematically assessed. Objectives: To identify and summarize studies comparing cardiorespiratory fitness measured by cardiopulmonary exercise testing in COVID-19 survivors versus non-COVID-19 controls, as well as to determine the influence of potential moderating factors. Methods: We conducted a systematic search of MEDLINE/PubMed, Cochrane Library, EMBASE, Google Scholar, and SciELO since their inceptions until June 2022. Mean differences (MD), standard mean differences (SMD), and 95% confidence intervals (CI) were calculated. Subgroup and meta-regression analyses were used to evaluate potential moderating factors. Results: 48 studies (3372 participants, mean age 42 years, and with a mean testing time of 4 months post-COVID-19) were included, comprising a total of 1823 COVID-19 survivors and 1549 non-COVID-19 controls. After data pooling, VO2 peak (SMD=1.0 95% CI: 0.5, 1.5; 17 studies; N = 1273) was impaired in COVID-19 survivors. In 15 studies that reported VO2 peak values in ml/min/kg, non-COVID-19 controls had higher peak VO2 values than COVID-19 survivors (MD=6.2, 95% CI: 3.5, 8.8; N = 905; I2=84%). In addition, VO2 peak was associated with age, time post-COVID-19, disease severity, presence of dyspnea, and reduced exercise capacity. Conclusion: This systematic review provides evidence that cardiorespiratory fitness may be impaired in COVID-19 survivors, especially for those with severe disease, presence of dyspnea, and reduced exercise capacity. Furthermore, the degree of reduction of VO2 peak is inversely (c) 2024 Associa & ccedil;& atilde;o Brasileira de Pesquisa e P & oacute;s-Gradua & ccedil;& atilde;o em Fisioterapia. Published by Elsevier Espa & ntilde;a, S.L.U. All rights are reserved, including those for text and data mining, AI training, and similar technologies.