Cognitive profile, neuroimaging and fluid biomarkers in post-acute COVID-19 syndrome | Scientific Reports – Nature.com

In the present study, we conducted a comprehensive cross-sectional and longitudinal assessment of individuals with PACS regarding their cognition, mental health status, neuroimaging, and fluid biomarker profile. This approach offers a broad view of PACS patients, which is particularly valuable considering the limited previous research in this area. We evaluated PACS patients with subjective cognitive complaints and their evolution over a 6-month period. Our findings revealed cognitive impairment affecting executive function in more than two-thirds of participants and verbal memory in over one-third. Additionally, prevalent mental health issues included apathy (64%), moderate-severe anxiety (57%), and severe fatigue (35%). Visual memory impairment correlated with total gray matter and subcortical gray matter volume, as well as regional GM reductions in the hippocampus and thalamus. Notably, markers of neuronal damage and inflammation were within normal limits. Importantly, overall health and cognitive evaluations showed no significant change over time. Furthermore, altered executive function and verbal memory, common in PACS, persisted in most subjects without any link to alterations in their biomarker and imaging profiles.

Our study stands out for three main reasons: First, it assesses cognitive deficits through comprehensive neuropsychological evaluations. Whereas the literature on PACS is replete with studies on cognition using screening tools like the Mini-Mental State Examination or the Montreal Cognitive Assessment53,54, our research incorporates detailed neuropsychological evaluations conducted by an experienced neuropsychologist. Second, our study is distinguished by its longitudinal design. Beyond describing the alterations in patients suffering from PACS, we repeated the same analyses six months later to assess their progression and track the evolution of these key health indicators over time. Third, our study is notable for the breadth of areas evaluated: it meticulously examines cognition, mental health, brain structure, and markers of inflammation and neuronal damage concurrently, tracking their longitudinal evolution.

Cognitive evaluations in PACS showed that attention-executive and verbal memory were the most affected domains (Fig.1), which has been described in previous published works; however, the pattern of alterations was broader and more heterogeneous between patients4,5,6,7,52,53. The sample had a high premorbid intelligence and would not be expected to perform below average on cognitive testing. Despite their cognitive reserve, known as a protective mechanism against neurological impairments, these individuals nonetheless experience cognitive deficits. This would likely lead to difficulties in effectively managing daily work and life responsibilities, adversely affecting their quality of life. At a 6-month follow-up, we determined that only the FCSRTDelayed Free Recall of verbal memory scores improved significantly from baseline, using LME models. Nevertheless, if we consider the percentage of normal evaluations (defined as the proportion of tests within clinical limits of normality), there was a significant improvement with time. Considering that most participants with abnormal results were close to the threshold for normal performance, even a slight improvement in these tests could lead to reaching normal threshold values. The improvement in verbal memory and the achievement of normality in the neuropsychological tests indicate a positive trajectory toward normal cognitive functioning. Conversely, high levels of anxiety, apathy, and fatigue present at the beginning of the study remained unchanged. The slight improvements observed in cognition did not strongly affect participants' clinical outcomes or quality of life. It is plausible that the persistence of executive function deficits, ongoing psychological symptoms, and chronic fatigue significantly influenced the overall lack of enhancement in participants' well-being.

Our results are in line with previous published works showing both improvements and persisting cognitive deficits in PACS55,56. Previous research has also documented a decline in executive functions among participants who initially presented with severe cognitive impairment57. The repeated administration of a cognitive test four times over a six-month period in our study raises concerns about the potential influence of a learning effect on the results. To mitigate the learning effect, participants received standardized instructions and practice trials during the baseline assessment to familiarize themselves with the cognitive test procedures, potentially minimizing the influence of initial unfamiliarity or anxiety on test performance. In future research, it is advisable to schedule evaluations at more significant intervals to allow for a more comprehensive study of PACS.

Participants also reported depressive symptoms, anxiety, apathy, fatigue, and low scores in general health. These symptoms did not improve during this 6-month study (Table 2). Given that our analysis demonstrated a significant relationship between one memory test and stratification in anxiety scores, we believe that the coexistence of cognitive and mental health symptoms could not be interpreted as causality. Additionally, the modest sample size in our study may have limited our ability to detect subtle differences in other categories. Recent literature has reported mixed findings regarding the associations between psychiatric comorbidities and cognitive impairment in individuals with PACS. For instance, one recent study identified a significant association between depression symptom severity and cognitive impairment severity among PACS patients58, while findings for post-traumatic stress disorder and anxiety were inconclusive. Conversely, another study found no association between depression, anxiety, total general health status, fatigue, and cognitive profiles59.

In our study of PACS participants, we observed a complex interplay between cognitive and mental health symptoms, with prevalent cognitive impairment alongside high levels of anxiety, apathy, and fatigue. This underscores the intertwined nature of cognitive and mental health domains in PACS, where cognitive deficits may coexist with psychiatric symptoms. We next sought to stratify participants by levels of anxiety, depression, apathy, fatigue, or quality of life scores according to their questionnaire scores. Participants displaying moderate or severe anxiety showed lower results in the ROCFT Recall subtest (adjusted p-value=0.0014). No significant differences were observed in cognitive tests between participants with normal and abnormal values of the other stratification categories. No associations were detected between longitudinal changes in cognitive and mental health measures. Conversely, a previously published work found that changes in executive functions were significantly associated with changes in depressive symptoms57.

While we can hypothesize that mental health issues may impede cognitive symptom improvement, it is worth noting that these mental health issues could be a consequence of the cognitive impairment as described elsewhere60. Understanding this relationship is crucial for informing treatment approaches; interventions targeting cognitive rehabilitation should consider the impact of comorbid psychiatric symptoms. Integrated interventions addressing both cognitive and psychiatric symptoms concurrently may optimize patient outcomes in PACS. Furthermore, both types of symptoms may be influenced by fatigue, which was nearly universal and severe in 35% of participants.

Fatigue has been implicated in various aspects of cognitive function, including attention, processing speed, and executive function, and has been associated with cognitive impairment in other medical conditions such as fibromyalgia and chronic fatigue syndrome. Moreover, fatigue often coexists with psychiatric symptoms such as anxiety and depression, contributing to the complex interplay between cognitive and mental health domains. In a recent study by Delgado-Alonso et al.61 investigated the relationship between subjective cognitive complaints, cognitive function, fatigue, and neuropsychiatric symptoms using various analytical methods. The study found that fatigue played a central role as the main mediator between objective and subjective cognition, while the impact of depression was indirect and mediated through fatigue. The lack of symptom improvement in PACS during the study suggests complex underlying factors. Possible reasons include the chronic nature of PACS, ongoing inflammation, and the interplay between cognitive and mental health symptoms. This highlights the need for personalized, multidisciplinary treatment approaches. Strategies may include pharmacological interventions, cognitive rehabilitation, psychotherapy, and lifestyle modifications.

Our study revealed a specific connection between cognitive deficits and brain changes in individuals with PACS. The ROCFT Recall test, a measure of memory and visual-spatial abilities, was the only cognitive test that showed abnormalities. These abnormalities were connected to both overall and specific areas of brain volume loss, specifically in the GM and WM globally, and in particular regions like the hippocampus and thalamus. This finding is significant because it identifies a direct relationship between certain cognitive deficits and changes in brain structure among individuals with PACS. The fact that these links were observed globally in GM and WM volumes, as well as in specific regions critical for memory and cognition (the hippocampus and thalamus), underscores the potential impact of COVID-19 on brain health. However, the fact that these associations were limited to certain brain regions and were only detected with the ROCFT Recall test suggests that the structural brain changes in PACS might be more nuanced than previously understood. While other studies, such as the one by Dez-Ciranda et al.5, have also found connections between cognitive deficits and MRI results, the limited scope of these associations in our study points to a potential gap in the literature. Specifically, it raises questions about the extent and significance of brain structural changes in PACS. Although it was beyond the scope of our study, it's noteworthy that some researchers have investigated the utility of functional neuroimaging to deepen our understanding of PACS pathophysiology. Bungenberg et al., in a cross-sectional study, used resting-state functional MRI (fMRI) to examine participants with PACS. They discovered changes in several brain regionsincluding the brainstem, olfactory cortex, cingulate cortex, thalamus, and cerebellumon average seven months after SARS-CoV-2 infection. These alterations were associated with the severity of fatigue and cognitive functioning54. While structural MRI delineates the brain anatomy, fMRI sheds light on the brains dynamic functions. By revealing changes in brain activity and connectivity, fMRI could reveal underlying neural mechanisms of PACS that are not apparent in structural changes alone.

Our next approach in this study was to correlate clinical and neuroimaging features of this PACS cohort longitudinally. While a previous study has included both cognitive and neuroimaging assessment of PACS62, to our knowledge, this is the first study to include longitudinal analysis of both cognitive and neuroimaging tests. We found significant positive correlations between both global and focal measures of brain volume/thickness and visual memory scores, but not with other cognitive tests. This correlation indicated that worse visual memory was associated with lower total and subcortical GM volume together with left cerebral WM volume. Furthermore, subcortical GM volumes, especially the hippocampus and thalamus, significantly corresponded with worse visual memory performance. Previous studies also explored the association between GM volume and cognitive symptoms; it has been reported that worse memory and visuospatial test performance is associated with a loss of GM volume5,20. In line with previous studies, our longitudinal analyses revealed no evidence of volume gain in a 6-month period, nor did we find evidence of progressive volume loss broadly. However, we did observe significant gray matter loss in the left pallidum and left transverse cortical thickness. Despite these findings, we do not believe that they hold clinical significance. The observed changes in the left pallidum and left transverse cortical thickness were not associated with any clinical symptoms or functional impairments in our study population. Therefore, we do not interpret these findings as clinically meaningful54.

The majority of previous studies13,14,16,63,64,65,66, have reported high levels of plasma and/or CSF cytokines, NfL and GFAP in the acute or subacute phase of COVID-19 infection that normalize at follow-up, albeit using differing follow-up intervals64,67,68. Some of these studies related these biochemical changes with the severity of the infection or the gravity of neurological symptoms; however, there is no consensus on how fluid biomarkers relate to acute COVID-19 symptom severity, PASC symptoms, or PASC progression/resolution. In our study, the levels of plasma and CSF cytokines, NfL and GFAP were within pre-specified normal limits. Similar results were observed by Boesl et al.69, they found that NfL levels were normal in participants with self-reported cognitive complaints, and GFAP was altered in only 4%. They compared participants with subjective cognitive decline, single domain or multi-domain impairment and found no association between persistent neuronal or astrocytic damage and cognitive impairment. We observed slight differences in some cytokine levels between PACS and control participants, with variations of small magnitude. Furthermore, cytokine levels were either elevated or reduced compared to controls. Given the proximity of all values, even minimal differences in a subset of measurements could potentially lead to clinically significant results. Despite achieving statistical significance, we find this difficult to interpret and potentially inconclusive, and in our opinion, without clinical significance. However, it is worth mentioning that other studies in neurocognitive disorders show relationships between select cytokines with measures of cognitive function, and this warrants further examination. We did not observe significant differences in either GFAP or NfL levels between PACS participants relative to controls. Previous studies14,15,16,70,71 have inconsistent results regarding the association of fluid biomarkers with the severity of infection or neurological symptoms. This variability in findings from past studies may arise from methodological differences, diverse patient populations, and the dynamic nature of the post-acute phase of COVID-19. All the samples were negative for antineuronal antibodies. The absence of antineuronal antibodies in all samples holds clinical significance, suggesting that autoimmunity involving these specific antibodies may not be a predominant factor in the pathophysiology of PACS. This finding implies that cognitive impairment and neurological symptoms observed in PACS may be driven by mechanisms other than direct autoimmune responses targeting neurons.

We next sought to clarify whether these biochemical markers related to neuropsychological test results in PACS patients, as previous studies have inconsistent results regarding the association of inflammatory marker levels and neuropsychological tests. Results have ranged from no association72 to an association between cytokine levels and fatigue or executive functions (Stroop Color Word test)73, or TNF- levels and memory74. In our research, we discovered a surprising positive correlation between higher GFAP levels and enhanced Stroop Word test performance. Despite observing impairment in Stroop Word test results among participants, GFAP levels stayed within normal ranges, suggesting these levels might not substantially affect cognitive performance or act as a cognitive function marker. Contrary to the expected negative correlation between GFAP and cognitive testing, stemming from inflammation's assumed detrimental effects on cognition, our findings suggest otherwise. This could indicate a compensatory or specific role of glial cell activation in supporting cognitive functions, or potentially represent a Type I error. This intriguing result encourages further investigation to confirm these findings and uncover the mechanisms involved. No association was observed between cytokines, NfL, or GFAP levels and global or regional MRI measures after adjusting for multiple comparisons. Finally, we found that patients serum or CSF samples did not immunoreact with brain tissue or live neurons, suggesting that brain autoantibodies are not involved in PACS symptoms. While our study did not reveal any significant abnormalities in markers of neuronal damage, inflammation, or neuroimaging among individuals experiencing cognitive manifestations following COVID-19 infection, several potential pathophysiological mechanisms warrant consideration. It is plausible that subtle, yet to be identified systemic or central dysregulated immune responses or diffuse microvascular or barrier changes could contribute to cognitive manifestations. We might also consider the central role of fatigue in cognitive manifestations. Future research exploring these mechanisms in depth is crucial for a comprehensive understanding of the neurological sequelae of COVID-19 infection.

An interesting finding elucidated by this work is the breakdown of PACS amongst sex. Whereas COVID-19 infects women and men equally, related publications indicate that there is a higher prevalence of females with PACS, with percentages ranging from 63 to 74%13,16,75, in line with these observations, 79% of participants in this study were women. Interestingly, in a study including 377 patients with COVID-19 infection, the female sex was independently associated with PACS within the multivariable analysis75. The higher prevalence of PACS in females suggests multifaceted influences across biological, psychological, and social dimensions. Hormonal differences may affect immune responses and neuroinflammation, contributing to gender-based variations in susceptibility and outcomes. Psychologically, gender-specific stressors and coping mechanisms could impact symptom manifestation, requiring exploration of psychosocial aspects in the post-acute phase. Social disparities, including healthcare-seeking behavior and societal expectations, may further influence the identification and reporting of PACS symptoms. Acknowledging these complexities highlights the importance of customized research and treatment approaches for effectively addressing PACS in females.

A significant limitation of our study is the small sample size, which included only 49 participants at baseline and 46 at the 6-month follow-up visit. This limitation is particularly pronounced concerning CSF samples. The lumbar puncture procedure was designated as optional. Consequently, CSF samples were obtained from only 12 participants. As a result, the interpretation of our findings must be approached cautiously; a larger number of CSF samples would have provided a more robust basis for identifying differences compared to controls, if any. The statistical analyses have been adapted to the reduced data. Thus, it could be generalized our results. However, the unicentric nature of the study, even if limited in the sample size, also provided homogeneity to the data acquisition. Secondly, the present study neither has healthy participant controls nor participants with COVID-19 infection without cognitive complaints for neuropsychological or neuroimaging analyses. This was due to the review of the local Ethics Committee, which considered the inclusion of controls as too high of a demand. This study may face referral bias, as participants were referred by healthcare providers, potentially overrepresenting severe cases. Additionally, considering the heightened fear and anxiety surrounding COVID-19, some participants actively sought assistance and self-referred due to concerns about cognitive symptoms related to the virus. Further research should consider a more diverse and randomized sample to mitigate potential biases in interpreting the severity and prevalence of cognitive symptoms in PACS. Another limitation of the study lies in the absence of baseline cognitive assessments prior to COVID-19 infection, which could have offered valuable insights into pre- and post-infection cognitive changes. However, participants with previous cognitive impairment were excluded. Finally, we believe the current duration of this study was limited and that including a longer endpoint with greater distance between measurement intervals may be more suitable for studying PACS cognitive symptoms. However, the study was designed during the last quarter of 2020, even before the formal definition of PACS, and most studies then were designed with short follow-up periods20,64.

In conclusion, our study showed cognitive impairment, mainly affecting attention/executive and verbal memory functions lasting for at least 6months in individuals with PACS. Cognitive impairment was accompanied by depressive symptoms, apathy, anxiety, fatigue, and low health status. These findings (except for visual memory loss) were not associated with brain structural abnormalities, elevated cytokines, markers of neuronal damage, or neuronal antibodies. Given these findings, a tailored and multidisciplinary approach involving cognitive and mental health interventions is recommended for patient care. Future research is essential for understanding the enduring cognitive trajectory of PACS and the associated biological mechanisms. Longitudinal studies of extended duration will provide insights into the long-term cognitive impact. Addressing the gaps identified by our study's limitations, ongoing research endeavors are crucial to guide clinical interventions and enhance the overall management of PACS.

View original post here:

Cognitive profile, neuroimaging and fluid biomarkers in post-acute COVID-19 syndrome | Scientific Reports - Nature.com