Exercise Treatments for Long COVID Symptoms: A Review of Controlled Studies

After the onset of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, long-term consequences emerged as latent symptoms for people recovered from the initial infection. Post-Acute COVID-19 Syndrome (PACS) is characterized by persistent, relapsing, or new symptoms that occur 30 or more days after the acute phase of SARS-CoV-2 infection, with a broad spectrum of physical and mental health manifestations that reduce quality of life [1]. PACS was first described in a Delphi study conducted by the World Health Organization (WHO) and then included in the International Classification of Diseases 10 and 11 (ICD-10 and ICD-11). People experiencing symptoms 4 weeks after the onset of the SARS-CoV-2 infection are thought to be affected by COVID-19 late sequelae or long COVID [2].

Long COVID is estimated to affect at least 10% of those recovering, corresponding to about 65 million people worldwide, with more than 200 reported symptoms including fatigue, cognitive deficits, sleep disorders, muscle weakness, and psychological problems [3,4,5]. A recent meta-analysis confirmed persistent symptoms for a minimum of 28 days post-infection, with COVID-19 survivors experiencing unresolved symptoms around four months after infection [6]. Possible mechanisms involve viral persistence, immune dysregulation, and blood cell changes, resembling other post-viral syndromes [7]. A total of 9.8% of patients with mild COVID-19 reported persistent symptoms one year post-infection, with an early phase (30–180 days) marked by hair loss, chest pain, cough, myalgia, and respiratory disorders, and a late phase (180–360 days), where most symptoms resolved [8]. Symptom severity defines three phenotypes (mild, moderate, severe), while clinical clusters (fatigue-like, respiratory, chronic pain, neurosensorial) describe different trajectories, with overlap indicating greater severity and reduced quality of life [9,10]. Stratifying patients by these profiles could help design targeted interventions and rehabilitation [9].

Several risk factors related to PACS were identified, including age (≥40 years old), female sex, frailty, emergency visits, and hospitalization due to COVID-19. Conversely, vaccination appears to reduce long COVID incidence, with a lower risk of dyspnoea [8,11], and increased protection with more doses. Identifying at-risk populations can contribute to the understanding and prevention of long COVID [11], addressing also the reduced quality of life [12].

The heterogeneity of symptoms that occurs in long COVID complicates the diagnosis and management of the condition. Exercise has emerged as a promising intervention [13]. However, evidence indicated some barriers to engaging in exercise and a lack of guidance regarding physical activity, pointing to the necessity for tailored exercise programs in individuals with long COVID [14]. Wright et al. [15] reported a marked decline in physical activity levels among individuals with long COVID and symptom exacerbation in response to physical exertion, highlighting the development of safe and effective exercise interventions [15]. Physically active individuals with a confirmed diagnosis of COVID-19 had a significantly lower risk of hospitalization, fewer hospital days, less respiratory distress, and a decreased need for oxygen support compared to sedentary people post-infection [16]. Similarly, physically active people with long COVID experienced lower hospitalization rates due to symptom exacerbation and a related better management, preventing disabilities and reducing the need for further interventions or medications [17].

Some trials have investigated the effects of physical exercise programs of variable duration and type in individuals with PACS, showing improvements in symptoms and quality of life. Some interventions, lasting from 1 to 2 weeks, assessed moderate-intensity aerobic exercise or a combination of endurance and balance training, yielding improvements in cardiorespiratory fitness and functional status [18], symptoms severity and immune function [19]. An 8-week protocol of aerobic training combined with strengthening exercises improved cardiorespiratory fitness, but neither quality of life nor persistent symptoms (dyspnoea, fatigue) improved [20]. Conversely, strengthening exercises of respiratory muscles improved quality of life but not exercise tolerance or lung function [21]. Finally, the current limitations in robust evidence make it difficult to publish effective interventions for long COVID [22]. Therefore, this systematic review investigated the impact of exercise in long COVID management, with a specific focus on the type of intervention and its effects on physical function and health-related quality of life.

The protocol of the systematic review was previously registered on PROSPERO (International Prospective Register of Systematic Reviews, CRD42024517069). The review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) [23]. The review question was "Is there scientific evidence concerning the effectiveness of exercise interventions in Post-Acute COVID-19 Syndrome"? The patients, interventions, control, outcome, and study design (PICOS) format [24,25] was based on the following:Population

This systematic review included studies focused on adults aged between 40 and 60 years old, with a confirmed diagnosis of PACS regardless of whether they required hospitalization for acute COVID-19 infection. Evidence showed that individuals aged 36–50 years had the highest proportion of long COVID diagnoses [7]. Similarly, a study characterizing PACS reported that adults aged 40–60 years were the most affected [10]. The studies including children, adolescents, and the elderly (over 60 years old), and people with ongoing physical or mental illness were excluded. Moreover, due to the heterogeneity in the definition of PACS, we did not apply restrictions related to diagnostic criteria or to the duration of symptom persistence, in order to include a larger number of studies potentially relevant to exercise interventions.

This review considered studies without specific restrictions regarding exercise interventions to include all studies that addressed any type of physical exercise. Studies that were not focused on exercise intervention were excluded.

This review considered studies that compared exercise interventions to any other standard management or care of PACS-diagnosed people.

Health-related quality of life (HRQoL) and functional fitness were identified as outcomes. Therefore, the studies including exercise interventions to determine their impact on the quality of life and/or functional fitness of individuals with PACS were included in the systematic review.

This systematic review included only randomized controlled trials (RCTs).

A total of 30,557 articles were identified from databases and registers (Figure 1). Before screening, 4972 duplicates were excluded. All the trials identified from ClinicalTrials.gov were excluded due to incomplete or unpublished results. Non-interventional studies, non-randomized trials, or out-of-scope studies were also excluded. Thirty studies were also excluded from the Cochrane Library Register due to unpublished results. Finally, a total of 25,171 articles were excluded for multiple reasons according to the PICOS framework (i.e., non-long-COVID population, non-interventional studies, non-controlled trials, studies that did not investigate the outcomes of interest, study designs other than RCTs). The remaining 306 articles were screened for title and abstract, and 67 were assessed for eligibility. Among these, 27 articles were excluded because the investigated population was identified as post-discharge patients, survivors, or patients recovered from COVID-19, thus without a diagnosis of PACS; 9 studies did not consider exercise intervention (e.g., electrical stimulation, vocal breath, etc.); and 11 studies were excluded due to the study designs (e.g., protocols, congress abstract, etc.). The remaining 20 articles were included in this systematic review.

The aim of this systematic review was to review the scientific literature to analyse whether any type of exercise intervention could improve health-related quality of life and physical function in people with a PACS diagnosis. A total of 20 RCTs was included in the systematic review, published from 2022 to 2025. The years of publication are consistent with the first characterization of PACS or "long COVID". In the second part of 2020, a post-COVID syndrome was only suspected [47], and then it was preliminarily confirmed with a case series study [48] in which fatigue and dyspnoea were identified as persisting symptoms in the follow-up (i.e., after the discharge of acute COVID-19 illness). The RCTs included in this review recruited participants with a confirmed diagnosis of PACS according to [2]—3 months after SARS-CoV-19 infection. However, two studies referred to PACS involving individuals with persisting symptoms for at least 40 days [31] or 6 weeks [28], and in one study, the duration of symptom persistence was not specified [34]. The heterogeneity in the PACS definition can be explained by the fact that the authors have based their work on earlier studies, which described the condition in different ways [48,49], and a lack of worldwide consensus at the time of study publication. These initial differences in definitions have led to a variety of methods and understandings of how long COVID has been studied and managed. Nevertheless, as knowledge increased, the clinical definition of PACS evolved, and it is still ongoing [50].

The frequency, setting, and delivery of exercise protocols were not comparable to each other, even though the type of exercise protocol was similar among some studies. Notably, in half of the studies, the exercise delivery was totally remote or in combination with a face-to-face program (i.e., telerehabilitation) [28,31,34,37,39,40,41,42,43,45]. This aligns with the guidelines of the time, which recommended avoiding physical contact to prevent contagion. Studies have demonstrated that telerehabilitation can improve quality of life and functional capacity in patients with COVID-19 sequelae, with favourable outcomes in terms of costs and health benefits. These findings suggest that telerehabilitation could be a valuable tool for providing equitable rehabilitation options to a broader patient population [51]. In the remaining studies, where the protocol was administered in person (at the hospital), aerobic activity was the most frequently investigated, but in different forms (e.g., virtual reality, aerobic versus usual care group, interval versus continuous training) [30,32,33]. Ref. [36] evaluated a combination of inpatient and home-based protocol in which the aerobic training was performed at home. Separately, two studies evaluated concurrent training [27,29], but in one protocol, it was also combined with inspiratory muscle training [29]. Regardless of whether the exercise protocol was conducted in person or remotely, all the studies reported positive results or feasibility (i.e., higher percentage of adherence). Aerobic training could be performed as moderate interval or continuous training, showing similar positive results. However, functional training showed slightly better results when compared with aerobic training or as part of concurrent training, whilst virtual reality did not yield greater results than traditional methods. However, findings related to aerobic training, even in combination with other types of training, are consistent with what is reported in the literature [52,53]. Some studies were excluded from this review due to study design (i.e., non-RCTs, protocols, reviews). Moreover, this review did not include water-based exercise, for which some existing literature has reported benefits in managing PACS symptoms and reducing hospitalization over the long term [17,54]. Finally, two studies focused on breathing exercises and were aimed at improving the respiratory function in PACS individuals. Specifically, a 2-week protocol of breathing- and strength-based exercises was clinically effective [31] as well as an 8-week protocol consisting of breathing exercises, pilates, and yoga disciplines at variable intensities [41]. These findings were also previously reported by [55], although their results were not specific to the PACS condition, and the effectiveness of yoga alone remains to be demonstrated. Therefore, a multi-training approach could be effective in improving the recovery of people with PACS [56]. This could be explained by the release of exerkines—namely signalling moieties released during exercise—with a crosstalk effect between multiple body systems, and it serves as a strategy in the treatment of long COVID [57]. Furthermore, it has been demonstrated that physically active people with PACS recover quickly and fully as compared to sedentary individuals with PACS, confirming that exercise can mitigate the debilitating symptoms of PACS [58]. The diversity of control conditions across studies may have affected the observed results. Comparisons of no-intervention or waitlist controls with alternative exercise modalities or educational guidance may overestimate or underestimate the reported benefits. Variations in execution modalities within control groups could limit direct comparisons and the generalizability of results.

The quality of life was one of the outcomes investigated in this review because PACS-related symptomatology has been demonstrated to reduce the wellbeing and lifestyle in people experiencing this condition [9,10]. In fact, most of the included studies evaluated whether an exercise protocol could influence HRQoL, and almost all reported related improvements in people with PACS. Ref. [32] did not detect significant changes in quality of life when comparing the same aerobic training but supervised by a general or specialized physiotherapist. However, it should be noted that the assessment of HRQoL was performed with the SF-12, consisting of a mental and physical component, and significant changes were, nevertheless, reported for the physical part of the survey. In this regard, the other studies that used the SF-12 [27,29,35] reported positive results in both the physical and mental components of the survey. However, the exercise interventions were shorter, and the studies were not multicentric; thus, the results were not applicable to a larger and more diverse sample. Studies that used the longer version of the same questionnaire (SF-36) [30,39], with similar sample size and duration of the exercise protocol [39,46], reported significant improvements in HRQoL. The SF-36 appears more sensitive in capturing changes in both physical and mental health domains, while the shorter version may be less responsive to subtle improvements, especially in physical function or fatigue. Compared with the other tools used (EQ-5D-5L and PROMIS), current evidence is still limited on their responsiveness in PACS. Overall, the SF-36 may be preferable for detailed evaluation, whereas the SF-12 can be considered for practical assessment [59,60,61]. Therefore, given that the quality-of-life assessment methods varied across the included studies, the research methodology may have influenced the outcome of these studies. Similarly, the functional fitness assessment varied across studies. Notwithstanding, the outcome-related parameters were improved in all the included studies regardless of the exercise protocol characteristics (excluding virtual reality, which showed no superior change compared to traditional methods) [33]. The overall low risk of bias identified across studies indicates that the quality of the studies is mostly high. However, in the studies where "some concerns" were identified, the reason was primarily attributable to masking procedures, which were not always possible due to the nature of the interventions [27,31,33,34,42].

Among the limitations of this review, the heterogeneity of the intervention protocols should be acknowledged. Most studies were not comparable in terms of frequency, duration, and modality. Therefore, despite overall positive trends, the heterogeneity of protocols prevented quantification of the magnitude of improvements across studies, making it difficult to establish which type of exercise should be considered superior. Moreover, the studies were conducted based on the knowledge available at the time, leading to differences in investigation methodologies and target population (e.g., diagnosis of PACS differed in terms of symptoms duration and clinical definition). Another limitation concerns the variability of outcome assessment tools. Quality of life was measured using different questionnaires (e.g., SF-12 vs. SF-36). Similarly, functional fitness assessments varied across studies. Moreover, these outcomes were not systematically analysed as a primary outcome, which may have underestimated their role in relation to exercise interventions. Finally, although overall dropout rates were below 37%, lack of adherence in some studies may still have affected outcome reliability and should be considered in the design of future studies.

Future research should address methodological biases and include preventive measures as exercise protocols to monitor long-term impacts and to improve understanding and treatment of PACS condition [7]. It is crucial to expand ongoing intervention studies based on current knowledge and generate high-quality evidence for diverse long-COVID populations [4,10]. Moreover, a cohort study [62] compared long-term health outcomes following COVID-19 illness and seasonal influenza, suggesting that the management of long COVID with exercise protocol could be transferable to seasonal influenza, hypothesizing the same improvement in functional fitness and HRQoL parameters. Further research could, therefore, be conducted, focusing on the influence of exercise interventions while recovering from viral infections.

The management of PACS still requires further in-depth studies regarding the treatment of symptoms according to its clinical definition. This review shows that an intervention protocol based on physical exercise is safe and feasible, also via telerehabilitation. The duration of the intervention can range from 2 to 12 weeks, although a higher adherence has been observed with 8-week protocols, three to five sessions per week, 1 h/session. An aerobic activity may provide benefits, as moderate or continuous, with greater effects if combined with strength-based exercises and/or as concurrent training. Further benefits can be achieved with specific respiratory muscle training (i.e., breathing exercises). However, not all exercise types consistently improved HRQoL or symptoms, highlighting the need for individualized protocols. Recommendations should consider clinical phenotype, baseline functional capacity, and adherence potential. Therefore, while exercise interventions appear promising, their prescription should be tailored and carefully monitored rather than universally recommended.