The Psychiatry of Long COVID: Brain Fog, Depression, and Beyond

Introduction

The 2024 National Academies definition describes Long COVID as an infection-associated chronic condition present for at least three months. The illness may be continuous, relapsing and remitting, or progressive, with effects spanning one or more organ systems. Importantly, a documented positive SARS-CoV-2 test is not required for the diagnosis, and no single biomarker is currently definitive. [1-3]

This broad definition is highly useful in clinical practice because the psychiatric presentation of Long COVID rarely stays neatly inside one diagnostic boundary. A single patient might report slowed thinking, insomnia, palpitations, low mood, headache, sensory overload, post-exertional symptom worsening, loss of occupational identity, and a pervasive fear that no clinician believes their illness is real. Psychiatry is often asked to determine whether the core problem is depression, a neurocognitive syndrome, trauma, dysautonomia, medication burden, a primary sleep disorder, or some complex combination of these factors.

The least helpful approach is forcing the clinical picture into a false choice between a purely medical or purely psychiatric explanation. Long COVID can comfortably coexist with major depressive disorder, generalized anxiety disorder, posttraumatic stress disorder, insomnia, substance-related problems, or a preexisting psychiatric condition. Simultaneously, it can produce cognitive and functional symptoms that are not adequately explained by any standard psychiatric diagnosis. [4-11]

Defining the Role of Psychiatry: Avoiding Dismissal and Diagnostic Capture

The most useful psychiatric formulation begins with two core propositions. First, depression, anxiety, trauma-related illness, sleep disorders, and suicide risk deserve active diagnosis and evidence-based treatment. Second, the presence of one of those conditions does not prove that fatigue, brain fog, orthostatic symptoms, pain, or post-exertional worsening are purely psychogenic. [4,5]

This distinction is far from semantic. In a US survey of 14,767 adults with prior SARS-CoV-2 infection, daily cognitive symptoms were reported by 56.7% of respondents with post-COVID-19 condition, compared to 27.1% of respondents who considered themselves fully recovered. Crucially, cognitive symptoms remained associated with functional interference and lower odds of full-time employment even after accounting for depressive symptoms. [7]

Psychiatry therefore has four linked tasks in this population:

1. Characterize the cognitive complaint: Translate the vague term “brain fog” into the specific affected domains, triggers, time course, and functional consequences. [4,6]
2. Diagnose treatable psychiatric syndromes: Determine whether the patient meets formal criteria for a depressive, anxiety, trauma-related, sleep-wake, substance-related, psychotic, or other disorder, rather than simply equating elevated screening scores with a definitive diagnosis. [5]
3. Identify mimics and red flags: Actively look for delirium, focal neurologic disease, seizure, medication toxicity, sleep apnea, endocrine or metabolic illness, substance effects, and acute suicide risk. [4-6]
4. Coordinate function-centered care: Align psychiatric treatment with physical pacing, rehabilitation, workplace accommodations, family education, and the management of coexisting medical symptoms. [2,4-6]

Translating Brain Fog into Actionable Clinical Domains

“Brain fog” is a patient-centered description, not a unitary medical diagnosis. Commonly described problems include slowed information processing, reduced sustained attention, difficulty shifting attention, working-memory overload, executive dysfunction, word-finding difficulty, forgetfulness, and disproportionate mental fatigue after cognitively demanding activity. [4,6-8]

A large community study in England found small but objectively measurable group-level cognitive deficits after COVID-19. These deficits were notably greater among participants whose persistent symptoms had not resolved. While these findings support the biological and functional plausibility of the syndrome, they do not provide an individual diagnostic test and should never be converted into a deterministic prognosis for a single patient. [8]

The clinical pattern of cognitive impairment can be highly inconsistent. A patient may perform perfectly well during a short, quiet office screen yet completely fail when required to multitask for several hours, process rapid conversations, manage frequent interruptions, or recover after sustained mental effort. Brief cognitive instruments were designed mainly to detect more global impairment, not necessarily the fluctuating processing-speed, attentional, and fatigability problems reported in Long COVID. A normal score therefore narrows the differential diagnosis but does not invalidate the patient’s complaint. [4,6]

The most informative cognitive history asks what the patient could do before infection, what specifically fails now, how long performance can be sustained, what happens later that day or the next day, and whether external structure changes performance. Collateral information from family or employers can be incredibly valuable when the cognitive change affects medication management, finances, driving, work quality, or basic self-care. [4,6]

Differentiating Depression, Anxiety, Trauma, and Sleep Disturbances

The mental health burden associated with Long COVID is very real, although prevalence estimates vary widely depending on the Long COVID definition used, sampling methods, acute illness severity, variant era, vaccination status, and whether studies measure general symptoms or formal diagnoses. In a nationally representative US sample of 25,122 adults, those reporting post-COVID-19 condition were approximately twice as likely to report depression or anxiety symptoms. Treatment uptake was similar across groups, but cost-related barriers to care were notably more common in the Long COVID group. [9]

Electronic health record studies also show associations between Long COVID diagnoses and subsequent coded mental health disorders. These studies are clinically important but remain vulnerable to detection bias, coding differences, healthcare utilization patterns, and residual confounding. They should not be read as absolute proof that one specific biological mechanism directly caused every psychiatric outcome. [10]

Longitudinal data from patients hospitalized during early pandemic waves show that cognitive, depressive, anxiety, and fatigue symptoms can persist, worsen, or newly emerge two to three years later, carrying significant occupational consequences for some patients. However, that cohort was heavily enriched for severe early-wave illness and cannot be generalized without qualification to contemporary, nonhospitalized cases. [11]

Several clinical overlaps are especially easy to misread:

• Fatigue is not automatically depression: Major depression is supported by a sustained syndrome that includes depressed mood or anhedonia alongside associated cognitive, vegetative, and self-evaluative symptoms. Long COVID fatigue may instead be tightly linked to physical exertion, fluctuate with orthostatic stress, and produce delayed post-exertional worsening. The two conditions can certainly coexist. [2,4,5]
• Palpitations are not automatically panic: Panic attacks are abrupt episodes of intense fear or discomfort with characteristic cognitive and physical symptoms. Positional tachycardia, heat intolerance, presyncope, and reproducible orthostatic worsening should prompt an autonomic assessment rather than premature psychiatric closure. [4,5]
• Poor concentration is not automatically a primary attention disorder: New cognitive inefficiency after an infection requires a thorough review of sleep, mood, medications, pain, autonomic symptoms, neurologic findings, and functional trajectory before a new lifelong neurodevelopmental diagnosis is assumed. [4-6]
• Insomnia can be both driver and consequence: Sleep disruption can worsen attention, pain, emotional regulation, and fatigue, while Long COVID symptoms can simultaneously fragment sleep. Evaluation should consider primary insomnia, circadian disruption, sleep-disordered breathing, restless legs symptoms, medication effects, and maladaptive time-in-bed patterns. [4-6]
• Trauma symptoms may be illness-related or treatment-related: Intensive care, respiratory distress, isolation, loss, and prolonged disability can contribute to PTSD, adjustment disorder, grief, or demoralization. These syndromes warrant dedicated treatment without becoming a blanket explanation for all ongoing physical symptoms. [5,11]

Exploring Biologically Plausible but Clinically Unsettled Mechanisms

Proposed mechanisms for Long COVID include viral persistence, immune dysregulation, vascular or endothelial injury, autonomic dysfunction, altered energy metabolism, organ injury, and complex interactions among these processes. The current evidence supports profound heterogeneity rather than one established pathway that can be reliably measured in every patient. [12]

Longitudinal UK Biobank imaging identified group-level structural and cognitive changes after SARS-CoV-2 infection, including among many participants who were never hospitalized. Those findings strengthen the case that post-infectious cognitive symptoms should not be dismissed as imaginary. However, the cohort was older, the study was not specific to Long COVID, and routine magnetic resonance imaging is not a diagnostic biomarker for brain fog. [13]

Psychiatric symptoms may arise through several routes at once. These include direct or indirect biological effects, sleep disruption, pain, autonomic symptoms, loss of function, social isolation, financial stress, stigma, and the recurrence of a preexisting disorder. A biopsychosocial formulation is useful only when all three components are treated as real contributors. It fails when “psychosocial” becomes shorthand for unexplained or self-generated symptoms. [4,5,9,12]

Conducting a Practical and Comprehensive Psychiatric Assessment

The initial visit should establish the acute infection history, symptom onset, longitudinal pattern, dominant current problems, and degree of functional change. A positive acute test can support attribution but is not strictly required under current US guidance. [1-3]

Clinicians must ask specifically about cognitive and physical post-exertional worsening. Post-exertional malaise can follow minor physical or mental effort, often worsens 12 to 48 hours later, and may last for days or weeks. A delayed crash after sustained concentration can look very different from moment-to-moment distractibility or low motivation. [2]

Common Overlaps That Require Careful Interpretation

Presentation	Clarifying questions	Clinically useful response
Slowed thinking, word-finding, forgetfulness	Which domains fail? What tasks are affected? Is there delayed worsening after cognitive effort?	Use a brief cognitive screen as one data point, document real-world function, review sleep and medications, and consider neuropsychology, occupational therapy, or speech-language pathology when impairment is persistent or consequential. [4,6]
Low energy, withdrawal, reduced activity	Is there anhedonia, pervasive low mood, guilt, hopelessness, or suicidal thinking? Is reduced activity driven by exertional symptom worsening?	Diagnose and treat depression when criteria are met, while assessing post-exertional malaise and other medical contributors rather than using fatigue alone as evidence of depression. [2,4,5]
Palpitations, dyspnea, dizziness, fear	Are episodes positional, heat-related, exertional, or associated with presyncope? Is fear primary or a response to reproducible physiology?	Check orthostatic features and relevant cardiopulmonary or autonomic causes before concluding that episodes are panic attacks. [4,5]
Insomnia or nonrestorative sleep	Is the problem initiation, maintenance, schedule drift, hypersomnia, snoring, witnessed apnea, restless legs symptoms, pain, or medication effect?	Treat insomnia behaviorally when appropriate and investigate sleep-disordered breathing or other sleep disorders when the history indicates. [4-6]
Intrusive memories, avoidance, hyperarousal	Was there intensive care, respiratory distress, isolation, bereavement, or another traumatic exposure?	Assess PTSD and related syndromes directly, then use evidence-based trauma treatment while continuing appropriate medical care. [5,11]
Cognitive or emotional worsening after medication changes	Are sedatives, anticholinergic drugs, antihistamines, cannabis, alcohol, stimulants, or complex polypharmacy contributing?	Reconcile the full medication and substance list, define a target for each agent, and reduce avoidable cognitive, sedating, autonomic, or sleep effects. [4-6]

A focused examination should include mental status, attention to fluctuation, language, motor or sensory asymmetry, gait when relevant, and orthostatic vital signs when symptoms suggest dysautonomia. Testing should be directed by the history and examination rather than ordered as a uniform Long COVID panel. Common baseline evaluations may include a complete blood count, metabolic testing, and thyroid assessment, with additional studies based on specific symptoms and risk factors. [2,4]

Using Screening Tools as Clinical Aids Rather Than Final Verdicts

Instruments such as the Patient Health Questionnaire-9, Generalized Anxiety Disorder-7, and PTSD Checklist can beautifully quantify symptom burden and track change over time. However, they were not designed specifically for Long COVID. Somatic items heavily overlap with fatigue, sleep disturbance, and concentration difficulty. Therefore, diagnosis still requires a thorough clinical interview, assessment of core affective and cognitive features, and evaluation of alternative explanations. [5]

The exact same principle applies to brief cognitive screening. A normal Montreal Cognitive Assessment or Mini-Mental State Examination does not exclude a narrower processing-speed, executive, attention, or mental-fatigability problem. Formal neuropsychological evaluation is most useful when the diagnosis remains uncertain, objective characterization will change treatment or accommodations, safety-sensitive work is involved, or the patient shows progressive or disproportionate decline. [4,6]

Recognizing Red Flags and Determining When to Escalate Care

Long COVID should never become a diagnostic umbrella that delays the evaluation of a new medical emergency. Immediate or urgent assessment is warranted for new focal neurologic deficits, delirium or fluctuating consciousness, seizure, rapidly progressive cognitive decline, severe new headache with neurologic findings, new psychosis or mania, suicidal intent or a plan with access to means, or an inability to maintain basic safety and self-care. [4,5]

Specialist referral is also highly appropriate when cognitive impairment is persistent and functionally consequential, the presentation is atypical, neurologic findings are present, a sleep disorder is suspected, substance use complicates care, or first-line psychiatric treatment fails. Referral should be coordinated carefully rather than used to simply transfer ownership of the whole illness to a single specialty. [4-6]

Management Strategies: Treating Targets and Protecting Daily Function

1. Establishing a Credible and Validating Therapeutic Frame

A useful opening statement is that the symptoms are real, the diagnostic work remains active, and psychiatric treatment is intended to reduce specific burdens rather than explain away the illness. CDC and AAPM&R guidance both emphasize validation, individualized goals, and restoration of function. [2,4]

Document the patient’s priority outcomes in concrete terms. Examples include reading for 30 minutes without a next-day crash, returning to a limited work schedule, sleeping on a stable schedule, reducing panic episodes, resuming medication management, or tolerating family conversation. Functional goals make treatment far more measurable than a global promise to eliminate brain fog.

2. Treating Syndromal Psychiatric Disorders According to Established Standards

When major depression, an anxiety disorder, PTSD, insomnia disorder, or another psychiatric condition is present, use standard evidence-based psychotherapy and pharmacotherapy tailored to the diagnosis, prior response, comorbidities, preferences, and safety. Psychotherapy can address mood, trauma, avoidance, grief, identity disruption, coping, and behavior patterns without implying that Long COVID is caused by faulty thinking. [5]

No medication is approved specifically to treat Long COVID. Psychotropics should therefore be prescribed for a defined target syndrome or symptom, with explicit monitoring for benefit and harm, rather than presented as disease-modifying Long COVID therapy. [3-5]

Medication selection must account for fatigue, sleep, cognition, orthostatic symptoms, pain, appetite, cardiac risk, drug interactions, and the possibility that anticholinergic or sedating burden is worsening function. Complex regimens deserve periodic deprescribing review, especially when multiple clinicians have added agents for overlapping symptoms. [4-6]

3. Prioritizing Sleep as a Core Clinical Domain

Stabilizing sleep may improve mood, attention, pain tolerance, and daily predictability even when it does not resolve the underlying Long COVID. Use cognitive behavioral therapy for insomnia when appropriate, regularize sleep and wake timing, review activating and sedating substances, and evaluate suspected sleep apnea, restless legs syndrome, hypersomnia, or circadian disorders. [4-6]

4. Approaching Cognitive Rehabilitation with Honest Expectations

The 2026 trial evidence does not support a simple yes-or-no verdict on cognitive rehabilitation. In RECOVER-NEURO, a 5-arm multicenter trial of 328 adults at 22 US sites, online cognitive training, a structured cognitive rehabilitation program, and transcranial direct current stimulation strategies did not show differential benefit on the prespecified primary outcomes. [14]

In CICERO, a smaller UK trial randomized 78 adults with objective impairment in at least two cognitive domains. A 10-week individualized, goal-oriented cognitive rehabilitation program improved patient-selected functional goal attainment at 3 months, and the difference persisted at 6 months. The sample was small and less diverse than the broader Long COVID population, participants could not be blinded, therapist contact differed, and the primary outcome was functional goal attainment rather than proof of universal objective cognitive normalization. [15]

A reasonable synthesis is to avoid generic promises while still offering individualized rehabilitation when the deficit is functionally important and the patient can participate without provoking sustained symptom worsening. Occupational therapy, speech-language pathology, and neuropsychology can help with external memory systems, task simplification, reduced multitasking, error-proofing, environmental modification, and graded return to meaningful cognitive tasks based on tolerance. [4,6,14,15]

5. Respecting the Reality of Post-Exertional Malaise

Cognitive work can absolutely be exertion. A fixed escalation plan that repeatedly pushes through delayed symptom flares may worsen function in patients with post-exertional malaise. Activity planning should identify the patient’s current envelope, distribute demanding tasks, include recovery time, and increase load only when the pattern is tolerable rather than because a calendar requires it. [2,4]

This is very different from reinforcing complete inactivity. The aim is calibrated participation that avoids repeated boom-and-bust cycles, preserves essential activity, and expands capacity when the individual trajectory permits. [2,4]

6. Integrating Work and Disability Accommodations into Treatment

Cognitive symptoms frequently become visible first at work, where speed, multitasking, noise, interruptions, and sustained attention are difficult to reproduce in a quiet clinic. Potential accommodations include written instructions, fewer simultaneous tasks, a quiet workspace, flexible scheduling, planned rest breaks, reduced hours, remote work when feasible, and a gradual return with predefined reassessment. [4,6,7,11]

Documentation should connect symptoms to specific functional restrictions rather than merely listing diagnoses. Examples include reduced tolerance for sustained screen time, impaired accuracy under interruption, delayed post-exertional worsening after meetings, or inability to maintain a full schedule without prolonged recovery. [4,6]

7. Coordinating Multidisciplinary Care to Prevent Fragmentation

Psychiatry may need to coordinate closely with primary care, neurology, physical medicine and rehabilitation, sleep medicine, cardiology or autonomic specialists, occupational therapy, speech-language pathology, and social work. A shared plan should clearly identify who monitors suicide risk, medication interactions, orthostatic symptoms, sleep, rehabilitation tolerance, and work documentation. [2,4-6]

Common Clinical Pitfalls and Practices to Avoid

• Do not use a normal routine laboratory result, brain scan, or brief cognitive screen to conclude that the patient’s symptoms are not real. [2-4,6]
• Do not let a positive depression or anxiety screen become a substitute for a thorough differential diagnosis. [4,5]
• Do not label reproducible orthostatic or post-exertional symptoms as panic or deconditioning without properly assessing their pattern. [2,4,5]
• Do not prescribe a psychotropic without a defined target, baseline measure, safety plan, and reassessment strategy. [4,5]
• Do not promise that a supplement, stimulant, antidepressant, anti-inflammatory drug, brain-training product, or device has established Long COVID-specific efficacy when that claim is not supported by controlled evidence. [3,12,14,15]
• Do not make return to work contingent on complete symptom resolution; instead, use function-based accommodations and staged reassessment. [4,6,7,11]

Addressing Special Populations and Systemic Contexts

Patients with preexisting psychiatric disorders may experience relapse, medication disruption, or amplified stress after infection. However, their psychiatric history should not lower the threshold for evaluating new neurologic, autonomic, sleep, or medical symptoms. [4,5]

Older adults require particular attention to delirium history, neurodegenerative disease, cerebrovascular risk, sensory impairment, medication burden, and loss of physiologic reserve. Post-hospital and post-intensive-care patients may also have critical illness neuropathy, hypoxic injury, ICU-related trauma, or prolonged deconditioning that fundamentally changes the differential and rehabilitation plan. [4,11,13]

Access to care is a major part of the clinical problem. Cost barriers, limited specialty availability, disability disputes, unstable employment, caregiving demands, and stigma can worsen psychiatric distress and reduce treatment adherence. Telehealth can improve reach but may also be cognitively fatiguing and inaccessible to patients with limited technology, sensory overload, or unstable housing. [4,9]

Acknowledging the Current Limitations of Long COVID Evidence

The Long COVID literature remains notoriously difficult to synthesize because definitions, comparison groups, variant eras, vaccination status, reinfection rates, acute severity, and outcome measures differ wildly. Many studies rely on self-report, specialty-clinic samples, electronic codes, or cross-sectional designs. Hospital cohorts may overrepresent severe acute disease, while online cohorts may overrepresent people with internet access and high symptom engagement. [7-11]

Mechanistic findings are promising but do not yet offer a single diagnostic pathway. Imaging and biomarker studies often identify group-level differences that cannot classify an individual patient with adequate certainty. Treatment evidence is especially limited, and the contrasting 2026 cognitive rehabilitation trials perfectly illustrate why intervention content, patient selection, outcomes, and delivery format matter immensely. [12-15]

Future Directions and Unanswered Questions in Long COVID Research

The next phase of research must stratify patients by cognitive phenotype, post-exertional malaise, autonomic dysfunction, sleep disorder, inflammatory or vascular markers, acute illness severity, and psychiatric comorbidity. Trials should measure both objective cognition and outcomes that patients actually value, including sustained work, household function, error rates, symptom recovery after exertion, and overall quality of life. [4,12,14,15]

Comparative effectiveness studies are also desperately needed for psychotherapy, medication strategies, sleep treatment, cognitive rehabilitation, pacing support, and integrated care models. Recruitment must intentionally include racial and ethnic minorities, lower-income patients, rural communities, people with disabilities, and those unable to tolerate high-burden research protocols. [4,9,15]

Conclusion: Embracing a Nuanced and Function-Centered Approach

The psychiatry of Long COVID is not a search for a psychological explanation of medically unexplained symptoms. It is the disciplined practice of holding several possibilities at once. Cognitive dysfunction may be entirely real despite a normal brief screen. Depression and anxiety may be present without explaining the entire illness. Autonomic, sleep, neurologic, medication, and post-exertional factors may mimic or amplify psychiatric symptoms. Ultimately, functional recovery may require highly coordinated treatment across multiple specialties. [2,4-8]

The current standard of care demands careful phenotyping, active treatment of defined psychiatric syndromes, protection against diagnostic overshadowing, urgent attention to red flags, and function-centered rehabilitation that respects post-exertional worsening. Current evidence simply does not yet justify a single mechanism, biomarker, medication, or cognitive program for every patient. [2-6,12,14,15]

 

References

1. National Academies of Sciences, Engineering, and Medicine. A Long COVID Definition: A Chronic, Systemic Disease State with Profound Consequences. Washington, DC: National Academies Press; 2024. DOI.
2. Centers for Disease Control and Prevention. Long COVID Clinical Guidance. Updated March 9, 2026. Accessed August 19, 2026. Official guidance.
3. Centers for Disease Control and Prevention. Long COVID Basics. Updated May 6, 2026. Accessed August 19, 2026. Official resource.
4. Cheng AL, Herman E, Abramoff B, et al. Multidisciplinary collaborative guidance on the assessment and treatment of patients with Long COVID: A compendium statement. PM R. 2025;17(6):684-708. DOI. PubMed. PMID: 40261198.
5. Cheng AL, Anderson J, Didehbani N, et al. Multi-disciplinary collaborative consensus guidance statement on the assessment and treatment of mental health symptoms in patients with post-acute sequelae of SARS-CoV-2 infection (PASC). PM R. 2023;15(12):1588-1604. DOI. PubMed. PMID: 37937672.
6. Fine JS, Ambrose AF, Didehbani N, et al. Multi-disciplinary collaborative consensus guidance statement on the assessment and treatment of cognitive symptoms in patients with post-acute sequelae of SARS-CoV-2 infection (PASC). PM R. 2022;14(1):96-111. DOI. PubMed. PMID: 34902226.
7. Jaywant A, Gunning FM, Oberlin LE, et al. Cognitive Symptoms of Post-COVID-19 Condition and Daily Functioning. JAMA Netw Open. 2024;7(2):e2356098. DOI. PubMed. PMID: 38353947.
8. Hampshire A, Azor A, Atchison C, et al. Cognition and Memory after Covid-19 in a Large Community Sample. N Engl J Med. 2024;390(9):806-818. DOI. PubMed. PMID: 38416429.
9. Naik H, Tran KC, Staples JA, Perlis RH, Levin A. Psychiatric Symptoms, Treatment Uptake, and Barriers to Mental Health Care Among US Adults With Post-COVID-19 Condition. JAMA Netw Open. 2024;7(4):e248481. DOI. PubMed. PMID: 38662370.
10. Zhang Y, Chinchilli VM, Ssentongo P, Ba DM. Association of Long COVID with mental health disorders: a retrospective cohort study using real-world data from the USA. BMJ Open. 2024;14(2):e079267. DOI. PubMed. PMID: 38309763.
11. Taquet M, Skorniewska Z, De Deyn T, et al.; PHOSP-COVID Study Collaborative Group. Cognitive and psychiatric symptom trajectories 2-3 years after hospital admission for COVID-19: a longitudinal, prospective cohort study in the UK. Lancet Psychiatry. 2024;11(9):696-708. DOI. PubMed. PMID: 39096931.
12. Peluso MJ, Deeks SG. Mechanisms of long COVID and the path toward therapeutics. Cell. 2024;187(20):5500-5529. DOI. PubMed. PMID: 39326415.
13. Douaud G, Lee S, Alfaro-Almagro F, et al. SARS-CoV-2 is associated with changes in brain structure in UK Biobank. Nature. 2022;604(7907):697-707. DOI. PubMed. PMID: 35255491.
14. Knopman DS, Koltai D, Laskowitz DT, et al.; RECOVER-NEURO Clinical Trial Group. Evaluation of Interventions for Cognitive Symptoms in Long COVID: A Randomized Clinical Trial. JAMA Neurol. 2026;83(1):49-59. DOI. PubMed. PMID: 41212544.
15. Vanova M, Patel AMR, Scott I, et al. Cognitive Rehabilitation and Functional Outcomes in Long COVID-Related Cognitive Impairment: A Randomized Clinical Trial. JAMA Netw Open. 2026;9(7):e2620687. DOI. PubMed. PMID: 42384384.