Abnormal exercise adaptation after varying severities of COVID-19: A controlled cross-sectional analysis of 392 survivors

The multisystem impairment promoted by COVID-19 may be associated with a reduction in exercise capacity. Cardiopulmonary abnormalities can change across the acute disease severity spectrum. We aimed to verify exercise physiology differences between COVID-19 survivors and SARS-CoV-2-naive controls and how illness severity influences exercise limitation. A single-centre cross-sectional analysis of prospectively collected data from COVID-19 survivors who underwent cardiopulmonary exercise testing (CPET) in their recovery phase (x = 50[36;72] days). Patients with COVID-19 were stratified according to severity as mild [M-Cov (outpatient)] vs severe/critical [SC-Cov(inpatients)] and were compared with SARS-CoV-2-naive controls (N-Cov). Collected information included demographics, anthropometrics, previous physical exercise, comorbidities, lung function test and CPET parameters. A multivariate logistic regression analysis was performed to identify low aerobic capacity (LAC) predictors post COVID-19. Of the 702 included patients, 310 (44.2%), 305 (43.4%) and 87 (12.4%) were N-Cov, M-Cov and SC-Cov, respectively. LAC was identified in 115 (37.1%), 102 (33.4%), and 66 (75.9%) of N-CoV, M-CoV and SC-CoV, respectively (p < 0.001). SC-Cov were older, heavier with higher body fat, more sedentary lifestyle, more hypertension and diabetes, lower forced vital capacity, higher prevalence of early anaerobiosis, ventilatory inefficiency and exercise-induced hypoxia than N-Cov. M-Cov had lower weight, fat mass, and coronary disease prevalence and did not demonstrate more CEPT abnormalities than N-Cov. After adjustment for covariates, SC-Cov was an independent predictor of LAC (OR = 2.7; 95% CI, 1.3-5.6). Almost two months after disease onset, SC-CoV presented several exercise abnormalities of oxygen uptake, ventilatory adaptation and gas exchange, including a high prevalence of LAC.