Table 1.
Overview of studies that investigated the physical capacity and exercise tolerance of patients recovering from COVID-19, employing CPET
| Reference | Location | Observation group at inclusion |
Proportion of patients treated in hospital (%) | CPET protocol | Control group | Main findings | |||
|---|---|---|---|---|---|---|---|---|---|
| N | Follow-up | Residual symptoms | |||||||
| Alba et al. [78] | USA | 18 | 258 days (mean) | Dyspnea, exercise intolerance | 33 | Maximal effort on a cycle ergometer | 18 matched uninfected subjects with unexplained dyspnea and/or exercise intolerance | No difference in CPET variables (e.g. VO2peak, peak workload, VE/VCO2 slope) between groups, except for a higher HRpeak in PASC cohort (P = 0.02) | |
| Clavario et al. [70] | Italy | 200 | 3 months | Fatigue, dyspnea, chest pain | 100 | Incremental, symptom-limited on a cycle ergometer | – | Reduced VO2peak (<85% pred.) in 50% | |
| De Boer et al. [77] | USA | 50 | 6 months (mean) | Dyspnea on exertion, chest pain | 10 | Maximal effort Ramp protocol on a cycle ergometer | Subjects from previously published cohorts of patients with metabolic syndrome and moderately active individuals | Reduced VO2max in 32% (<84% pred., of which 56% with HRR < 15 bpm and 63% with low O2 pulse at peak exercise); Higher mean lactate and lower FATox compared with controls (P < 0.05) | |
| Mohr et al. [71] | Germany | 10 | 115 days (mean) | Dyspnea | 60 | NR | – | Gap between preserved mean work rate (94% pred.) and reduced VO2peak (72.3% pred.); elevated AaDO2 in 30% and mean lactate post-exercise (5.6 mmol/l) | |
| Motiejunaite et al. [72] | France | 8 | 3 months | Exertional dyspnea | 0 | NR | – | 100% and 88% incapable of reaching predicted VO2max and workload, respectively; respiratory alkalosis and hypocapnia in 38%; elevated VE/VCO2 ratio in 63%; symptom reproduction at exertion in all subjects | |
| Raman et al. [73] | UK | 58 | 2.3 months (median) | Majority persistent symptoms (e.g. breathlessness, fatigue) | 100 | Symptom-limited ramp protocol on a cycle ergometer | 30 matched uninfected subjects | Reduced VO2peak and oxygen uptake efficiency slope; greater VE/VCO2 slope compared with controls (P < 0.001) | |
| Rinaldo et al. [74] | Italy | 75 | 97 days (mean) | 52% with residual dyspnea | NR | Incremental, symptom-limited on a cycle ergometer | – | Reduced VO2peak (72% pred.) in 55% (of which 32% with HRR <15% and 37% reduced anaerobic threshold) | |
| Singh et al. [75] | USA | 10 | 11 months (mean) | Dyspnea, exercise intolerance | 10 | Invasive CPET, Ramp protocol until 85% of pred. peak HR was reached | 10 matched uninfected subjects with unexplained dyspnea | Reduced VO2peak (<80% pred.) and systemic O2 extraction; greater venous oxygen saturation and VE/VCO2 ratio compared with controls (P < 0.01) | |
| Szekely et al. [76] | Israel | 71 | 91 days (mean) | 67% with persistent symptoms (e.g. fatigue, myalgia) | 4 | Symptom-limited ramp protocol on a cycle ergometer | 35 matched uninfected subjects | Lower anaerobic threshold, O2 pulse and VO2peak; higher arteriovenous oxygen difference compared with controls (P < 0.05); chronotropic incompetence in 75% | |
AaDO2: alveolar–arterial oxygen difference; bpm: beats per minute; FATox: fat oxidation; HR: heart rate; HRR: heart rate reserve; mmol/l: millimoles per liter; NR: not reported; pred.: predicted and VE/VCO2: minute ventilation/carbon dioxide production ratio.