We studied the association of two main LC-symptoms, exercise intolerance, and fatigue with an objectively measured exercise capacity. We found, unexpectedly, that neither the mean exercise capacity nor the mean oxygen uptake percentage of the predicted value differed among those with or without subjective exercise intolerance or fatigue or those with a combination of these symptoms. Beta-blocker users excluded, 14% fulfilled the criteria of dysautonomia with slightly lowered exercise capacity and work efficiency. Another 15% fulfilled the criteria of chronotropic incompetence, and they had in mean a moderately lowered exercise capacity and a slightly lowered oxygen uptake and work efficiency in comparison to the subjects without these features. This may suggest a poorer aerobic physical performance during exercise and could be considered in expectations of recovery and help in patient selection and designing the rehabilitation.

Before the LC era, chronotropic incompetence has been seen as predictor of cardiovascular disease and mortality [32, 33]. Although the criteria of chronotropic incompetence vary in literature, it has been reported that 1/3 of patients with heart failure would fulfill this criterion and would be associated with poor quality of life and prominent exertional symptoms [29]. There are several reports on chronotropic incompetence being associated with LC, both in hospitalized and non-hospitalized LC patients. Cardiac functional or structural abnormalities associated with the LC condition probably explain the low HR response [24, 29, 30, 34,35,36,37], smoking and earlier cardiac diseases have been seen as risk factors for chronotropic incompetence in LC [30]. Earlier studies have shown that the maximal VO2 is lower in those who have suffered from a more severe LC disease [38, 39]. Here, as reported above, only 3 of those with chronotropic incompetence had been hospitalized and the other 10 had been treated at home at the acute phase of the LC disease suggesting that chronotropic incompetence does not only develop in those with severe LC disease.

We recognized 13 patients fulfilling the criteria for chronotropic incompetence [29, 30]. They had the lowest oxygen uptake in the mean, 70% of the predicted value, and the lowest Wlast4 (in the mean, 63% of the predicted value), which is in line with earlier studies on chronotropic incompetence [29, 32, 33]. They also had the lowest blood pressure increase during exercise and during the recovery phase which might suggest a lowered cardiac capacity. The higher breathing reserve, despite the RER values similar to the other subjects, suggests that neither breathing problems nor the submaximal exercise level explain the lowered exercise capacity in the subjects with chronotropic incompetence. Instead, possible cardiac limiting factors might play a role in exercise intolerance in this patient group [36, 40, 41], and in the present study, the greatest percentage of ECG findings suggesting ischemia or sympathetic stimulation were found in those with chronotropic incompetence. However, in the present register study the further development of the disease cannot be followed up.

Although the term dysautonomia refers to varying forms of autonomic dysfunction [17, 41], the HR triad used here as the criterion for dysautonomia has also been used in earlier LC studies, and it has been suggested that dysautonomia explains the fatigue symptoms or functional limitations in LC patients [23, 28, 34, 40]. In the present study, twelve subjects fulfilled the criteria of dysautonomia, with increased resting HR, reduced increase of HR from the high resting level, and slow HR recovery. These patients showed slightly lowered maximal working power and working efficiency (Wmax/VO2peak), which is in line with the findings of an earlier study [23]. The results of some earlier studies suggest that dysautonomia is a mild and reversible condition [17, 23].

In the present study, in group EI + F the maximal HR was higher and HR recovery 10 min after exercise slower than in group N, which might indicate accentuated sympathetic activity after the exercise stress. The associated unpleasant feeling may be one reason why these patients feel forced to quit their previous physical activities. Increased sympathetic activation [17,18,19,20,21] or sympathetic excitation and parasympathetic reduction [41, 42] have been suggested to be prevalent in LC. A slow HR recovery after exercise has been reported in LC [22, 42,43,44,45], with an improvement after 5–6 months [42, 44]. An earlier study has suggested that the increased sympathetic tone might be sequelae after the viral LC infection with subintimal inflammation [46,47,48], leading to increased vascular stiffness, probably explaining the autonomic nervous dysfunction seen as delayed HR recovery but being reversible during follow-up [48]. Mental stress or fear may further contribute to increased sympathetic tonus [49].

Several other studies on LC have found that LC patients had reduced peak oxygen consumption (VO2) [13, 16, 23, 30, 50, 51], reduced maximal work rate [23] or ventilatory efficiency [13, 16, 34, 44, 50]. In the present study, the patients with exercise intolerance with or without fatigue had in mean a normal exercise capacity. However, we could not assess whether there was a true decrease of exercise performance because CPX results before the COVID-19 infection were not available.

In the present study, 40% of our patients had reduced exercise capacity measured as Wlast4 lower than 80% of the predicted value. Compared with Sorensen et al. [51], who found that 19% of their patients had lowered peak workload (≤ 84%), the number of patients with lowered exercise capacity is here greater. Concerning oxygen uptake % of the predicted value the results were corresponding, 50% in our study and 36% in the study by Sorensen et al. [51]. In the study by Sorensen [51], all LC patients from their clinic were tested, whereas in our study, only patients with clinical indications (e.g. cardiac symptoms, exercise intolerance, assessment of working capability etc.) were consecutively tested in our laboratory, representing a selected LC population.

It has been suggested, based on CPX results, that LC patients with reduced exercise capacity are deconditioned due to a long-term decrease in physical activity after the acute disease [52, 53]. It is obvious that also in the present patient material accentuated deconditioning explains at least some of the cases of lowered (< 80%) exercise capacity. Physical inactivity and related deconditioning are known to be associated with lower parasympathetic cardiac modulative activity and slowed HR recovery after exercise load [53,54,55]. Deconditioning has been shown to be resolved with exercise training [54, 55], and a recent meta-analysis found that exercise capacity measured as VO2 would improve within 3 to 6 months after the acute COVID-19 infection [24].

In the present study, several patients felt exercise intolerance despite the measured exercise capacity and oxygen uptake were within normal limits. According to our results, the explanation might be that the patients with very good physical condition would after the COVID disease be deconditioned compared to their pre-COVID felt condition. In addition, those with earlier training or other physical activities easily get worried when they feel that they are not in as good condition as before. In addition, psychological and socioeconomic factors might be a functional component in the long-lasting symptoms [56, 57].

The present study shows that CPX helps to assess possible cardiac or respiratory impairment and to identify or rule out diseases with specific treatment. Normal results are encouraging and remove obstacles to safe rehabilitation. Cardiopulmonary exercise testing is useful in assessing the patient’s exercise capacity, but not all LC patients need CPX; clinical selection of patients for referral for testing is important. The CPX results should be interpreted considering the patient’s history and physical activity in daily life. CPX results could also be used to encourage exercise if results are lowered due to inactivity and deconditioning.

The patients reported also other symptoms than fatigue and exercise intolerance. In an LC study by Contreras et al. [50], 55% of their patients reported having LC respiratory symptoms, and during exercise testing, these symptoms occurred in 44% of them. In the present study, the reported respiratory LC-symptoms occurred in 43% of the patients but restricted exercise capacity in only 9% of them. In our study, the most common causes of termination of the exercise testing included leg fatigue or discomfort (59 patients), fatigue (17 patients), and dizziness (16 patients). Breathlessness was the cause of exercise termination only in 9 patients. In comparison, in the Contreras study, 18% of patients had been acutely hospitalized compared to 10% in the present study. Additionally, 4% of their study showed slight exercise hypoxemia, whereas we did not find exercise hypoxemia. Although these are only small differences between the two studies, Contreras et al. had excluded pre-COVID respiratory diseases. However, they did not report FEV1 follow-up of their patients associated with exercise. We had 9 patients with previously diagnosed asthma in good balance, even during exercise, except one patient with increased FEV1-variation. In Contreras et al.’s study, there was a suggestion of hyperventilation during exercise testing based on an increased VE/VCO2 slope, which was not found in the present study. However, hyperventilation, according to the authors’ clinical experience and earlier literature, may start without exercise or be provoked by different methods or situations in different subjects [58]. The difference might also result from patient selection, as Contreras et al. studied dyspnea symptom, and their control subjects had some kind of dyspnea symptoms, whereas the present study focused on exercise intolerance and fatigue. Additionally, a different method of asking about the symptoms during the exercise might influence the results, but this is only speculation.

As it comes to increased sympathetic activity, there were signs of that in HR increase during exercise as well as during recovery (see Fig. 1). A POTS-type (postural orthostatic tachycardia syndrome) reaction is difficult to detect by exercise testing alone, but there was a suggestion of this, as orthostatic testing revealed an increased HR reaction in over 40% of a small cohort of the present patients.