Cognitive dysfunction after covid-19
BMJ 2024; 384 doi: https://doi.org/10.1136/bmj-2023-075387 (Published 01 February 2024) Cite this as: BMJ 2024;384:e075387- Emma Ladds, GP partner and Wellcome Trust DPhil, student1,
- Julie L Darbyshire, research manager and senior scientific officer1,
- Nawar Diar Bakerly, professor in respiratory medicine2,
- Zacc Falope, consultant in rehabilitation medicine3,
- Ian Tucker-Bell, patient advisory group member4
- 1Nuffield Department of Primary Care Health Sciences, University of Oxford, UK
- 2The Northern Care Alliance, Manchester Metropolitan University, University of Manchester
- 3Birmingham Community HealthCare NHS FT
- 4LOCOMOTION consortium (patient author)
- Correspondence to E Ladds emma.ladds{at}phc.ox.ac.uk
What you need to know
Cognitive problems are common after acute SARS-CoV-2 infection and can be disabling and frightening
Symptoms tend to improve, but this may take up to a year. Those with chronically persistent symptoms lasting more than 12 months have a lower chance of improving
Symptoms usually fluctuate
Assessment should be directed to documenting the nature and trajectory of the impairment and excluding alternative diagnoses
Self-management techniques may help patients manage their condition
As of March 2023, when the Office for National Statistics stopped collecting data on this condition, 1.879 million individuals had self-assessed as having long covid—symptoms lasting more than 12 weeks following acute covid-19 infection. Of these, the proportion of individuals with symptoms lasting two years or more is around 42%, suggesting a decline in new cases of long covid but a persistence of those with ongoing symptoms.1 Some systematic reviews and meta-analyses have reported that up to a third of such individuals have persistent symptoms of cognitive impairment,23 but estimates vary widely and are complicated by methodological heterogeneity—eg, study size, assessment approach, follow-up duration, and different sampling frames (from self-reported surveys4 to large retrospective matched cohort studies of health records5), as discussed in a recent meta-analysis.6
The pathological underpinnings and potential therapeutic possibilities for cognitive impairment in long covid are also uncertain. The bulk of evidence to date is mechanistic (using basic science, animal models, or human tissue), observational (using longitudinal cohort studies), or hypothetical (reasoning from basic principles); this literature has been well summarised by the RECOVER Consortium.7 Because of the methodological heterogeneity, even when individual studies have been rigorously conducted, it is difficult to know to what extent their findings can be extrapolated and generalised across those with long covid. A few randomised controlled trials of potential treatments (pharmacological and non-pharmacological) have been completed, and others are under way.7
Cognitive performance is typically conceptualised in terms of functional domains depending upon the task at hand. Domains are hierarchical and inter-related, ranging from more basic sensory and perceptual processes to executive functioning and cognitive control.8 Both psychometric studies and qualitative studies of the patient experience identify problems across one or more of these domains. The most frequent symptoms seem to affect memory, attention, and concentration9 but discrete impairments are also seen in attentional and executive processing,1011 different types of memory,1012 visuospatial processing,13 and language.12 Many patients show deficits in multiple domains,1214 which may or may not correlate with self-reported symptoms.15 Moreover, linked systemic symptoms such as fatigue, insomnia, and a general befuddled state frequently termed “brain fog” are commonly reported. This seems to wax and wane with physical fatigue.16 A higher symptom burden is often associated with concomitant mood disorders or post-traumatic stress disorder, regardless of illness severity.15171819
Longitudinal cohort, case control, and qualitative studies have shown that while many individuals improve gradually over time, cognitive symptoms fluctuate unpredictably, causing uncertainty and (in some) anxiety.9202122 These impairments can have profound impacts on an individual’s life, with significant ramifications for relationships, jobs, and normal daily activities.422 This article explores the impact of these symptoms and offers a practical guide for GPs to help them approach cognitive dysfunction in individuals with long covid as indicated by the evolving and conflicting evidence base. An account from a patient’s perspective is given in box 1.
Patient experience
When I got covid in December 2020 I was worried about what might happen. It was before vaccines, and the news was full of statistics of the hospitalised and those who’d lost their lives. After my 14 day isolation I was relieved to get back to normal life, thinking I had dodged quite a bullet.
I was quickly aware that something wasn’t right.
At the time of my infection I was a drama teacher in a secondary school. I remember being at work in January 2021 and finding it impossible to concentrate on what I had to do. Trying to focus made my head spin, and working on a computer screen was almost impossible. In the classroom I couldn’t process what was happening around me like I used to be able to, there was a sharpness lost, and that had a significant impact on both my teaching ability and classroom management. After several attempts at returning to work it became clear that I wasn’t able to continue, and I left teaching in January 2022.
Alongside my difficulties focusing there are problems with memory. Remembering things from one minute to the next is often difficult, and this is not something I ever had an issue with before covid. If I am mid-task or thought and am distracted (my husband asks a question, the phone rings, etc) I can completely forget what I was doing. I lose words, and as someone who works with words, that is both frustrating and at times frightening. My brain mixes up facts and names, and I find myself having to double check such things on a regular basis to avoid embarrassing situations or mistakes.
Like many elements of long covid, these problems fluctuate and are worse when I don’t do so well with my pacing, or have more mental stimulus or physical activity than I am able to cope with. One big challenge is that, as this is a hidden condition, others forget it’s there and so fail to accommodate it.
That covid has impaired my cognitive function in many ways is without question. I can only hope that, in time, a solution is found, and in the meantime I adapt to survive, and do what I can to manage the condition.
It isn’t easy!
Questions patients ask
Why did I develop cognitive problems?
Anybody who has been infected with the SARS-CoV-2 virus can develop long covid, including cognitive impairment, regardless of the severity of infection,23 but a prospective cohort study of 4000 patients logging their symptoms through an app revealed an increased risk of developing long covid in older individuals, women, and those with a higher body mass index or greater burden of pre-existing comorbidities.24 Moreover, systematic reviews and meta-analyses have shown that vaccination against SARS Co-V-2 reduces the risk of developing long covid.2526 Further associations have been seen in a cohort study of patients admitted to hospital between cognitive impairments and objectively assessed persistent impairments in smell or taste27 and sleep apnoea,28 while numerous, but relatively small cross sectional studies using objective clinical assessments have found strong reciprocal associations with fatigue or mood disorders.293031
What is causing my cognitive problems?
There is no straightforward answer to this question, which is understandably frustrating for patients.22 Different areas of the brain are involved in different cognitive processes. Animal and neuroimaging studies in humans with persistent symptoms have shown alterations in the anatomy or functioning of some of these regions correlating with specific impairments,32333435 providing a pathological link with symptoms. However, there are many mechanistic hypotheses. Neuropathological, biomarker, and neuroimaging studies have suggested ongoing viral infection in the brain,3336 viral activation of immune or inflammatory pathways resulting in neuronal destruction or impairment,3537 and dysregulated autonomic function as seen in other conditions.38 Alternative proposals include microvascular damage caused by endothelial dysfunction and hypoxia related inflammation,39 and parallels have also been drawn with cognitive difficulties experienced during chemotherapy treatment40 or in other chronic inflammatory or autoimmune disorders.41
Seek to exclude non-specific cognitive blunting from other underlying causes. Alongside specific metabolic, toxic, or (patho)physiological causes (box 2), poor sleep, diet, smoking, alcohol, and recreational drug use, reduced activity levels, and low exposure to natural light may also contribute. These should therefore be established in the history and improved or mitigated if and where possible.
Other contributors to cognitive impairment
Alzheimer’s disease/dementia
Anxiety
Vitamin B12 deficiency
Chronic hypoxia (eg, chronic obstructive pulmonary disease)
Chronic pain
Depression
Menopause/perimenopause
Migraine
Multiple sclerosis
Myalgic encephalomyelitis
Poor glycaemic control
Parkinson’s disease
Schizophrenia
Sleep apnoea
Thyroid problems
Medications that can contribute to cognitive impairment*
Anticholinergics
Anticonvulsants (eg, gabapentin/pregabalin)
Antihistamines
Antidepressants (eg, amitriptyline/nortriptyline)
Antispasmodics
Anxiolytics
Chemotherapy
Opioids
Sedatives
* Medication review is recommended (including over-the-counter medicines, supplements, and herbal remedies) to address polypharmacy and potential drug interactions
What tests do I need for my cognitive function?
No specific test will diagnose or assess the severity of cognitive impairment in long covid, although multiple measures can provide an assessment of cognition more generally. Several instruments are in development.424344 These include mental evaluation that captures cognitive deficit in the context of other long covid symptoms. The approach to investigation is to exclude “red flag” symptoms that indicate the need for emergency referral,45 consider and manage comorbidities and alternative causes, and then explore specific symptoms.
As with all investigations, consider how the result may alter management. A normal, appropriate brain imaging test may reassure those who are concerned but may also identify incidental findings that can lead to unnecessary intervention. Brain imaging using standard computed tomography or magnetic resonance imaging scanning is not routinely recommended in those with cognitive problems following covid-19, particularly those who describe the typical pattern of “brain fog” frequently seen in patients with long covid. However, imaging or targeted cognitive testing (see below) may be indicated in some patients to rule out other neurological conditions. These would typically include individuals with symptoms suggestive of intracranial pathology such as suspicious nausea and vomiting, postural headaches, focal motor or sensory changes, or visual disturbances, or those who may be at higher risk of an underlying neurological disorder. Examples of the latter group may include older individuals concerned about degenerative conditions, those with a family history of hereditary neurological disorders, or with symptoms potentially indicative of an underlying disorder—eg, bradykinesia or recurrent falls (suggestive of Parkinson’s disease and progressive supranuclear palsy, respectively).
Guidelines from the National Institute for Health and Care Excellence on long covid suggest neuropsychometric testing after six months if cognitive function is not improving (and especially if worsening).46 Most new symptoms spontaneously resolve within this timeframe. However, it may be helpful for patients to elicit the level of impairment relative to their “normal” state at presentation and monitor any change (usually improvement) over time. After this point, formal testing may be helpful as the proportion of individuals with long term persistent symptoms seems to increase with time.1 Formal assessment may include the Montreal cognitive assessment (MOCA),47 mini mental state examination,48 or frontal assessment battery49 (among others), which assess attention, working memory, visuospatial skills, executive functioning (MOCA), and verbal fluency (MOCA). However, the usefulness of such tests in assessing post viral change is currently unclear. Moreover, the tests are affected by an individual’s baseline cognitive function and previous education; thus individuals with higher baseline cognitive performance may notice reduced abilities relative to their usual level but still pass standard tests. The fluctuating nature of long covid also makes one-off assessments potentially inaccurate. Make patients aware of these issues when considering referrals for further neuropsychometric testing.
How do people manage their cognitive problems?
Our engagement with patients in the LOCOMOTION patient advisory group and qualitative studies have highlighted practical strategies to cope with impairments. Many patients have found communicating the fluctuating nature and variable impact of their (largely invisible) symptoms particularly challenging. One technique is to explain energy levels in terms of battery life, bank balance, or the number of spoons remaining in a drawer. All three analogies describe energy as finite units used up by activities and therefore leaving the person unable to do all they want or need to do. Some activities can be rejuvenating and replenish the store. Planning days with a mixture of differently draining activities can help minimise fatigue and frustration and maximise performance.22 Similarly, activity and symptom trackers can help identify triggers for mental fatigue that can be avoided or help people prioritise tasks and activities.
Other reported strategies focus on reducing an individual’s general cognitive load (a combination of simultaneous conscious and unconscious cognitive tasks). These “task loads” interact in complex ways but because working memory capacity is limited,50 even distractions that a person is not consciously aware of can measurably degrade performance.51Box 3 outlines different approaches individuals have raised in qualitative interview and focus group studies to minimise their cognitive load.2252 Lessening overall cognitive load can improve functioning in specific domains while also lessening the distress of “failing” at simple cognitive tasks.
Self-management strategies that can reduce cognitive load
Limit distractions when performing tasks by controlling environmental stimuli such as noise and light
Make and use lists
Identify “homes” for things that might easily be lost (eg, glasses, watch, phone, keys, medications, handbag)
Keep a regular routine that includes rest breaks
Break down activities into smaller components
Ask others to help by checking, reminding, and proofreading work
Identify patterns in symptoms to help with scheduling activities
Pace yourself to maintain steady levels of activity and avoid crashes
Try group mindfulness classes, available at many long covid clinics
Try cognitive exercises such as word puzzles or memory exercises
What can my primary care team do for my cognitive symptoms?
Qualitative interview and focus group studies highlight that patients find the process of seeking help and support to be exhausting.225354 However, they also show how beneficial patients find a supportive encounter or relationship with a clinician who hears the patient’s story and responds empathically, even if they are unable to ease the physical symptoms.225354 Alongside this witnessing role, despite a lack of clear evidence, it makes theoretical sense to ensure optimal management of comorbid conditions such as diabetes, blood pressure, postural orthostatic tachycardia, menopause, insomnia, or depression. For example, the pro-inflammatory environment associated with impaired glycaemic control has been hypothesised to contribute to capillary dysfunction and exacerbation of long covid symptoms, including cognitive dysfunction.3955
Community support from physical and occupational therapists may be helpful to support optimal day-to-day functioning.56 However, physical rehabilitation should reflect the evidence from observational studies57 and patient groups58 that post-exertional symptom exacerbation (PESE) and post-exertional malaise (PEM) can be precipitated using traditional “graded exercise therapy,” which takes no account of the person’s energy levels and encourages participants to “push through” fatigue. Rather, physical interventions should incorporate “pacing”—that is, adjusting the level of exertion to accommodate energy levels as self-assessed with the “battery” or “spoons” metaphor described above.
World Health Organization guidance suggests that in individuals with PESE or PEM, red flags for such deteriorations should be established before commencing exercise—eg, using pulse oximetry or screening tools such as the De Paul symptom questionnaire-post exertional malaise (DSQ-PEM). Moreover, rehabilitative regimens should be personalised, grounded in realistic expectations, and be conducted in a safe and supportive environment.59 Small interventional studies have shown objectively assessed improvements in fatigue, cognition, and overall functioning following in-person programmes that included resistance, endurance, and balance training alongside an education and pacing element6061—although the latter study specifically focused on individuals who had been admitted to intensive care units during their acute illness.61 Large scale observational or interventional studies of physical interventions for brain fog which reflect the heterogeneity of the long covid population are currently lacking.
Clinical trials are under way for a range of therapeutic agents in long covid, with a view to these relieving cognitive symptoms. Trials include antivirals, anti-inflammatory or immunomodulators, cardiac agents, cannabinoids, agents targeting receptors within the central nervous system (eg, aripiprazole and vortioxetine), anticoagulants, and others (summarised by the RECOVER consortium).7 However, to date, no large scale randomised controlled trials support any particular therapeutic approach to treat cognitive impairment in long covid. Numerous small studies have shown variable benefit from use of nutritional supplements,6263 mindfulness supported by essential oils,64 transcutaneous auricular vagus nerve stimulation,65 hyperbaric oxygen therapy,6667 and acupuncture.68 Patients also report symptom improvement following yoga techniques.69 However, these results are inconsistent, subject to selection, information, and publication biases and potential performance effects during testing. Patients may wish to consider these approaches, but an important aspect of the GP’s role is to ensure appropriate expectation management and discuss their lack of routine availability through the NHS.
Given the absence of definitive evidence on effective drug treatments, GPs can provide essential longitudinal support and encouragement for patients. This ongoing therapeutic relationship may help individuals optimise their physical condition, given the strong association between physical and mental symptoms in chronic conditions.70 This might include education around a healthy, balanced lifestyle and attention to optimising sleep patterns, helping them develop their own management strategies—including discussions around energy preservation and pacing, discussing the potential harms of over-investigation and unproved treatments, and managing their expectations around recovery. Moreover, support and guidance around returning to work is vital. Many employers lack a formalised return-to-work policy, or phased returns are too structured to support the fluctuating nature of long covid. Communicating this to employers and supporting flexible returns may be helpful for patients. Patients may also be interested in taking part in research studies and might value being signposted to opportunities available to them.
Will my cognitive problems get better, and what can be done if they don’t?
A retrospective, nationwide cohort study of nearly two million healthcare records in Israel suggested that individuals who suffer persistent symptoms following mild covid-19 infection, including cognitive dysfunction, generally recover within 12 months.71 Further longitudinal studies support this, suggesting that where cognitive symptoms after covid-19 have resolved, they have taken around one year to do so.122072 In particular, executive function commonly seems to resolve within a year.4 Patients with long term persistent symptoms will need ongoing support, encouragement with self-management strategies, and active acknowledgment of the frustration and impact of their condition on daily life.
Patients with severe, progressive, or persistent cognitive impairment following covid-19, or those for whom the diagnosis is in doubt, will likely benefit from referral to a specialist multidisciplinary long covid service or cognitive disorders clinic.73 Similarly, referral to rehabilitation services may help in the context of a comprehensive holistic neuropsychological rehabilitation programme that is team based and goal focused. This may focus on specific rehabilitative approaches such as respiratory physiotherapy or fatigue management strategies and may address cognitive and emotional difficulties to improve everyday function.74
Long covid patients’ accounts of cognitive dysfunction and its impact (adapted from225375)
One of the things I’ve realised is how many things I do in my normal day—and I’m not talking about work—that are cognitive, and that I didn’t previously think of as such. For example, a trip to the supermarket, where the amount of sensory information—just staring at a row of things while looking for the food that you want, remembering where things are in the aisles, and planning your trip so that you don’t have to walk backwards and forwards around the shop—all surprised me. It’s not just “can I walk around the supermarket?” but the planning, getting there, choosing stuff, all of that which is really difficult.
Rebecca (34)
I can’t cope with multiple inputs. If I’m trying to reply to a message on my phone and one of my boys starts speaking to me or there’s something else happening as well—that just really fries my brain. Multi-tasking used to be my superpower, and I was able to do lots and lots of things. As a doctor, I would be dealing with one patient while hearing about another. I’d be remembering, doing something else, juggling lots and lots of things. And now I can’t keep multiple plates spinning, I absolutely can’t. I’ve got to focus on just one thing or I make massive mistakes and it’s like I forget my intentions all the time.
Nadya (54)
I can ask somebody a question and then I’ll ask the exact same question two minutes after and not remember that I’ve asked. I can’t remember significant things that have happened in the past either.
John (48)
I’ve had an in-depth conversation with somebody who isn’t aware of how I am, and they’ve said to me, “you’re going round in circles in your conversation” or “you’re not making a lot of sense,” when I hadn’t realised how repetitive I was being. But those same people can’t seem to cut me any slack, or see how difficult it is. There doesn’t seem to be the understanding—but it would be beyond my comprehension as well if I hadn’t lived it.
Stephen (61)
Seven months in, I don’t know whether I’m going to get my brain back. I’m really, really fearful for the future and whether I’m going to be able to get back to what I want to do. That’s like your identity and yourself. Being an [allied health professional] is a big part of who I am, and if I’ve lost that, I’ve lost a huge part of me.
Rachael (45)
Resources for patients
Healthtalk.org. Hear directly from people who have had a wide range of debilitating and sometimes confusing symptoms. These are people who know what it is like to have long covid. They talk about things they have found difficult and some of the things that have helped. https://healthtalk.org/Long-Covid-In-Adults/overview
Headway. A charity for individuals suffering from a brain injury. Includes advice and support for recognition and management of cognitive dysfunction https://www.headway.org.uk
Research
At the time of writing there are 34 ongoing research studies funded by the NIHR recruiting patients with long covid in the UK. Individuals with cognitive symptoms may like to learn more about participating here: https://bepartofresearch.nihr.ac.uk/results/search-results?query=long%20covid&location=
Resources for professionals
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. PM R 2022;14:96-111
Silver JM MTSJ. Mild traumatic brain injury. Traumatic Brain Imaging. Washington,American Psychiatric Publishing; 2011. 239-64.
See also The BMJ article for general sources of support for healthcare professionals and individuals with long covid45
How patients were involved in the creation of this article
Patient members of the LOCOMOTION quality improvement collaborative contributed the comments in the box “Long covid patients’ accounts of cognitive dysfunction and their impact.” ITB has lived experience of long covid and is a member of the patient advisory group for LOCOMOTION. The paper was read by four additional patients with long covid and modified in response to their feedback. This was largely concerned with ensuring we had adequately captured the patient experience with regard to the complexity and severity of symptoms. We subsequently included box 1 “Patient experience” to ensure the patient voice was adequately heard.
How this article was created
This article is one in a series that offers guidance to general practitioners on persistent symptoms following covid-19. The series began with a general introduction45 and has previously covered orthostatic tachycardia.76 This paper on cognitive dysfunction draws on two sources: a literature search of PubMed conducted using the terms “post-acute covid-19”, and “long covid” and multiple terms relating to impaired cognition; and group discussions with front line clinicians and patient partners (the LOCOMOTION consortium77) who are members of a national quality improvement collaborative for long covid clinics in UK. In box 1 “Patient experience” and the box “Long covid patients’ accounts of cognitive dysfunction and their impact” we give examples of the impact of impairments raised by our patient advisory team and by participants from qualitative research studies previously conducted by one of the authors (EL). We have also provided a practical guide about how to approach cognitive dysfunction in long covid, designed using the established Quick Reference Handbook style guide used by the Association of Anaesthetists and Resuscitation Council UK.
Acknowledgments
We thank patient research participants for the descriptions in box 1, as well as four additional members of the LOCOMOTION project’s patient advisory group for helpful comments as this paper was being drafted.
Footnotes
The paper draws on the clinical experiences and wisdom of the LOCOMOTION consortium. All authors contributed to the narrative and snowball literature search and synthesis of key findings. EL and JLD wrote the first draft which was extensively amended by other authors. All authors offered feedback on the infographic. All authors read and approved the final manuscript. Trisha Greenhalgh provided editorial comments on a final draft.
Competing interests: Several authors and patient contributors are part of the LOCOMOTION consortium and are funded by the NIHR within the LOCOMOTION Consortium. The views expressed in this publication are those of the author(s) and not necessarily those of NIHR or the Department of Health and Social Care.
Funding: This is an ongoing research project funded by the National Institute for Health and Care Research (NIHR) (long COVID grant, Ref: COV-LT2-0016.
Provenance and peer review: commissioned; externally peer reviewed.