A multifaceted clinical picture
Around 10% of people develop Long COVID following a SARS-CoV-2 infection; globally, this amounts to at least 65 million people. More than 200 symptoms have been described. Common symptoms include fatigue (severe exhaustion), post-exertional malaise (PEM; worsening following exertion), cognitive impairment, shortness of breath, dysautonomia (dysfunction of the autonomic nervous system), headaches, muscle and joint pain, and disturbances in the sense of smell (Davis et al., 2023). Children are also affected; the prevalence is estimated at 1–3% of SARS-CoV-2 infections (Basaca et al., 2025).
Which biological mechanisms are being investigated?
A clinical review groups potential underlying mechanisms supported by strong evidence into three main categories (Greenhalgh et al., 2024). The first comprises virus-related mechanisms: replication-competent SARS-CoV-2 or components of the virus may persist in tissues. Other viruses, such as Epstein–Barr virus, may also be reactivated.
The second category comprises dysregulated immune and inflammatory responses. Findings include altered immune cells, elevated cytokine levels (signalling molecules of the immune system), and antibodies directed against the body’s own structures. Immune activation and inflammatory responses may remain detectable for more than 180 days after infection (Aid et al., 2025). The complement system, which forms part of the immune defence, may also become dysregulated and contribute to tissue damage.
The third category concerns changes in the inner lining of blood vessels (the endothelium) and in blood clotting. Findings include inflammation, an increased tendency for blood to clot, and very small blood clots known as microthrombi.
Other possible mechanisms are also being investigated. These include inflammatory processes in the nervous system, reduced energy production in the mitochondria (cell components often referred to as the cell’s “powerhouses”), disturbances in iron metabolism, oxidative stress caused by free radicals, and changes in the microbiome (Gupta et al., 2025; Skevaki et al., 2025).
What is known about treatments?
Overall, evidence for treatments remains limited. A systematic review assessed 24 studies involving 3,695 participants. It found evidence of moderate certainty for specific outcomes associated with three approaches. Moderate-certainty evidence means that confidence in the findings is moderate. The true effect is likely to be close to the estimate, but future research could still change this estimate substantially. This rating reflects the certainty of the findings—not the magnitude of the effect or an individual’s likelihood of improvement.
An online cognitive behavioural therapy programme reduced fatigue by an average of 8.4 points on the fatigue scale used and improved concentration. In a study of an online-supported combination of physical and mental rehabilitation, 16.1 percentage points more participants in the treatment group achieved a meaningful improvement in their overall health than in the comparison group; depressive symptoms and quality of life also improved. The third approach, intermittent aerobic exercise performed three to five days per week for four to six weeks, improved physical function more than continuous exercise. No convincing scientific evidence was found for other therapeutic approaches, including various medications (Zeraatkar et al., 2024).
These findings apply to the specific programmes and treatment outcomes studied. In particular, the exercise findings do not support a general exercise recommendation for people with PEM. Exercise that is not carefully managed may exacerbate inflammatory processes and delay recovery. For people with PEM, symptom-guided pacing, in which activities are adjusted to the energy available, is probably safer than a standardised programme involving a gradual increase in exertion (Greenhalgh et al., 2024).
An international guideline published in 2026 recommends vaccination or antiviral medication during the acute phase to prevent Long COVID. For treatment, it recommends multispecies probiotics, cognitive behavioural therapy for fatigue, and personalised rehabilitation after PEM has been ruled out. Nirmatrelvir–ritonavir is not recommended for treating Long COVID, nor are glucocorticoids recommended for persistent respiratory symptoms or disturbances in the sense of smell. The certainty of the evidence supporting the recommendations ranges from very low to moderate (Cao et al., 2026).
Care and unanswered questions
Current care focuses on energy management, individually tailored multidisciplinary rehabilitation, and the treatment of specific symptoms. An ongoing therapeutic relationship and a coordinating healthcare professional are important (Brode & Melamed, 2024).
The available data are insufficient to reliably assess the effectiveness of experimental treatments (Livieratos et al., 2024). High-quality randomised controlled trials are needed in which participants are randomly assigned to a treatment or comparison group. Biomarkers (measurable biological characteristics) and biological subtypes of Long COVID also need to be identified. Future studies should include population groups that have so far been underrepresented and actively involve people with Long COVID (Skevaki et al., 2025; Davis et al., 2023).
Although robust scientific data on treatments and medications are still lacking, the growing understanding of the biological mechanisms underlying Long COVID represents a major advance. These findings provide the basis for identifying and investigating promising therapies.
