Cognitive Dysfunction After Acute COVID-19

More than 300 million cases of COVID-19 have been confirmed worldwide, with estimates projecting the actual burden of illness to be much higher. Besides the immediate morbidity and mortality of the pandemic, a spectrum of persistent symptoms has been observed in many individuals after their acute illness.¹ Known by various names—long COVID, postacute COVID syndrome, postacute sequalae of COVID-19 (PACS)—the syndrome has been reported consistently worldwide and constitutes an overlapping constellation of symptoms involving many organ systems. We have chosen to use the acronym PACS to describe this syndrome throughout this article for consistency.

Cognitive dysfunction lasting longer than 12 weeks after the acute illness is one of the most alarming and durable symptoms reported, with a recent meta-analysis suggesting a prevalence of 22% among all people who have had COVID-19.² People describe a variety of cognitive symptoms, including brain fog, which can represent a wide spectrum of subjective perception and effects on daily functioning. Our understanding of this phenomenon has been confounded by numerous superimposed variables (eg, COVID-19–related social upheaval with widespread quarantines, social isolation) along with the inherent difficulties in measuring specific underlying pathologies. Here we summarize frameworks for understanding and providing support for people with PACS.

Patterns of Persistent Cognitive Dysfunction

Several studies have been conducted in an attempt to define the neuropsychologic profile of people with PACS; however, there has been notable variability in findings and particularly in correlation between objective and subjective measures. Short-term complaints centering on cognitive function associated with COVID-19 infection were recognized early during the COVID-19 outbreak. Defining the underlying patterns of impairment has been of consistent interest to increase understanding of the intersection of the acute, postacute, and chronic phases of these complaints and their bearing on broader risks of static or progressive cognitive decline over time.

On standardized cognitive testing, patterns of cognitive impairment in those with PACS do not resemble traditional neurodegenerative patterns of cognitive impairment or consistent burden of impairment seen in other populations. In a study of 53 individuals diagnosed with COVID-19 infection, overall rates of borderline or impaired cognitive scores were similar to historic rates found in general population studies.³ This study suggested that premorbid vascular and psychologic factors were not reliably correlated with cognitive test results 6 months after COVID-19 but were more significantly correlated with concurrent levels of emotional distress. A similar study of 20 individuals found a small burden of objective cognitive impairment on formal testing, but high premorbid rates of psychologic diagnoses, such as anxiety and depression. Some studies have not found a correlation between behavioral health and post–COVID-19 cognitive complaints. The relationship remains poorly understood but is observed frequently.⁴

Another study, reporting data derived from 50 individuals between 6 and 12 months following COVID-19 infection found a consistent and statistically significant pattern of impairments affecting attention, executive functioning, episodic memory, and visuospatial processing. The effect sizes seen in this study were relatively small but were spread across younger and older patients, which may have magnified the personal and social effects.⁵

Studies to evaluate cognitive symptoms among individuals during the pandemic regardless of COVID-19 exposure have been performed using different methodologies and had mixed results. Two studies using standardized testing in healthy individuals found measurable worsening in executive functioning and motor inhibition in one and in working memory in the other. Other studies have relied on self-reported measures or questionnaires, which may amplify imprecision in reporting of specific patterns of symptoms or defining their severity and effects.⁶

Mechanisms of Cognitive Dysfunction in PACS

The underlying causes of persistent symptoms after COVID-19 are not understood, although several parallel and compelling hypotheses have been proposed in an attempt to unify neurologic as well as broader systemic symptoms associated with PACS. Mechanisms of neural involvement have centered around retrograde trafficking of viral antigens in olfactory or vagal axons, direct leaking of spike proteins across the blood–brain barrier, and angiotensin-converting enzyme 2 receptor–mediated binding (which also has been proposed as a unifying mechanism of systemic, multiorgan syndromic impairments observed in COVID-19). Overall, lack of measurable chronic infection suggests a larger role of virus-induced systemic inflammatory processes along with vascular and metabolic-mediated tissue injuries rather than direct viral injury.⁷

[¹⁸F]Fluorodeoxyglucose PET studies of the brain have shown decreased metabolic activity in angiotensin-converting enzyme 2–enriched areas throughout the CNS, including olfactory, visual, cortical, and brainstem or cerebellar structures. These findings were observed subacutely and trended toward improvement at 6 months after the acute illness.⁸ Studies seeking to measure patterns of persistent neurodegeneration have found overall normalization of most biomarkers of neuronal injury by 6 months after acute COVID-19, despite persistence of neuropsychiatric symptoms.⁷ Further study and time will be required to differentiate the relative contribution to persistent symptoms from subacute postviral proinflammatory effects vs comorbidities during a prolonged phase of recovery.

A persistent theory is that causes of cognitive issues at later stages of recovery are multifactorial and may be related indirectly or unrelated to COVID-19 infection. Other authors have suggested that longstanding observations of cognitive dysfunction in people with severe cardiopulmonary diseases can provide insight into COVID-19–related cognitive dysfunction. Similar uncertainty exists regarding relationships between chronic fatigue and cognition in other illnesses, highlighting the difficulty in delineating these experiences.⁹ The same studies often find trends toward poorer cognitive performance in people experiencing other neurologic symptoms, such as headache or anosmia.¹⁰ However, the significance of this relationship is unclear with regard to its implications for underlying pathogenesis.

Multifactorial Contributors to Cognitive Impairment

Several comorbidities arise frequently enough that they are worth consistent consideration and screening in people with cognitive problems related to PACS. Sleep difficulties are reported frequently and are important to explore in detail, particularly given their intersection with prominent post–COVID-19 complaints, such as fatigue and cognitive difficulties. Causative considerations exist as well, including the effects of circadian disruptions from lifestyle interruptions during the pandemic, respiratory stress, and depression or other mood disorders affecting sleep quality. Obstructive sleep apnea (OSA) shares some mechanistic considerations with PACS (eg, relative hypoxia, sleep cycle disruptions, vascular and oxidative stress), and although the effect of COVID-19 on the prevalence of OSA is not well-defined, the presence of OSA likely is a magnifier of most neuropsychologic post–COVID-19 syndromes.¹¹

Mood disturbances are highly prevalent in people with PACS, with reports of depression and anxiety in a variable proportion of people, with estimates ranging from 23% to more than 30%. Some studies estimated a prevalence of mood disturbances in the general population of more than 30% in the midst of the pandemic, which underscores the difficulty in dissecting out illness-related and psychosocial risk factors for these comorbidities. Post-traumatic stress disorder (PTSD)–related symptoms were reported in up to 43% of individuals, with a background estimate of between 7% and 10%. The risks were increased in, but not exclusive to, those surviving critical illness associated with COVID-19 infection. Persistence of PTSD after COVID-19 may mirror the experience noted in the wake of the initial SARS outbreak.¹²

COVID-19 and Dementia

Encephalopathy in association with acute COVID-19 is a well-known complication, with a recent large study showing a prevalence of 9.2%. Poor outcomes were seen in about 58%, with higher odds noted among older adults and those with presumed fatal comorbidity, Glasgow Coma Scale score <9, vasopressor requirement, renal replacement therapy during intensive care unit stay, or CNS ischemic or hemorrhagic complications.¹³ Approximately one-third of people who survive acute COVID-19–associated encephalopathy have cognitive deficits at the time of hospital discharge.¹⁴ Long-term outcomes of these individuals are being characterized; studies reporting 6-month follow up of individuals hospitalized for COVID-19 show poor correlation between subjective cognitive complaints and objective measures.¹⁵

A commonly recognized clinical situation is unmasking of dementia in individuals who develop acute COVID-19. This clinical pattern can be seen in acute illness and acute encephalopathy, perhaps magnified by the effects of social isolation. There is great interest in understanding the risk of dementia in people with cognitive dysfunction secondary to PACS. Some studies, such as the NIH RECOVER initiative, are seeking to help answer this question, but long-term follow-up will be required to better understand consequences of the illness.

Therapeutic Approaches to Cognitive Dysfunction

No single intervention to treat the persistent symptoms experienced by those with PACS is accepted as widely beneficial. Interventions remain symptom-oriented and supportive. The heterogeneity of severity, duration, and nature of symptoms poses a challenge to the evaluation and diagnosis of PACS, although consistent trends toward improvement in clinical cohorts as well as biomarker studies is encouraging. For people experiencing progressive or worsening symptoms over time, it is important to reconsider the diagnosis and consider comorbidities.

For many people, lack of a clear COVID-19 diagnosis compounds the challenge, given the possibility of false-negative or false-positive SARS-CoV-2 test results, or not having been tested during their illness. This underscores the importance of inclusive diagnostic evaluations and careful consideration of overlapping patterns of symptoms. Progressive cognitive symptoms should prompt consideration of broad structural and metabolic contributors.

The multifactorial nature of these symptoms offers several targets for intervention. Screening for broad patterns of primary and secondary health effects, including physical symptom and mental health burden, lifestyle, and health maintenance factors, is essential. Multidisciplinary care and use of appropriate rehabilitation services, including dedicated cognitive rehabilitation and appropriate physical therapy, are of broad benefit to cognitive and physical symptoms, and any barriers to access of this care must be navigated. Sleep apnea, circadian rhythm disruption, or other disorders affecting sleep quality should be screened for and addressed as appropriate based on the individual’s pattern of symptoms.^16,17

The literature lacks consensus on and robust findings of benefit for most nutritional and vitamin supplementation interventions to improve or support cognition; relative benefits of probiotics and a variety of nutritional supplements used for this purpose have been reported in smaller, nonrandomized trials.¹⁸ Discordant findings of benefits from pharmacologic interventions demonstrate lack of adequate organized data and likely reflect the clinical heterogeneity of PACS. The use of stimulants has not been studied in a systematic fashion but can be extrapolated as a potential source of help for some people with clear inattentive patterns of cognitive dysfunction. However, given lack of long-term outcome data, early reliance on these medications may mask the existence of other unaddressed contributors to symptoms.

Conclusions

Our understanding of the burden of COVID-19 on individuals and society continues to develop. Early and ongoing efforts at quantifying and characterizing patterns of cognitive dysfunction in people with PACS continue to yield mixed results, highlighting the diverse nature of the clinical entity as a common end point of a multifactorial process. Symptomatic recovery requires addressing direct effects of illness, including the psychosocial factors that arise from living through a global pandemic, a protracted state of illness, financial insecurity, and reduced access to factors that promote physical and cognitive health, such as physical activity, cognitive stimulation, and social interaction.

Key Messages

• PACS with cognitive impairment occurs frequently and has significant impacts on the daily functioning of affected individuals.
• In patients with worsening cognitive symptoms, the presence of an underlying degenerative process “unmasked” by acute illness should be carefully considered, though the causative relationship between neurodegenerative disease and COVID-19 infection will require more time to fully understand.
• Most individuals benefit from careful evaluation of comorbidities that compound their persistent symptoms, including comorbid mood disorders, sleep derangements and physical limitations.
• For all patients, proactive supportive care for physical, cognitive and emotional symptoms can enhance symptomatic and functional improvement.