key: cord-0926275-y06wxvg7
authors: Hay, M.; Ryan, L.; Huentelman, M.; Konhilas, J.; Hoyer-Kimura, C.; Beach, T. G.; Serrano, G. G.; Reiman, E. M.; Blennow, K.; Zetterberg, H. M.; Parthasarathy, S.
title: Serum Neurofilament Light is Elevated in COVID-19 Positive Adults in the ICU and is Associated with Co-Morbid Cardiovascular Disease, Neurological Complications, and Acuity of Illness
date: 2021-05-02
journal: nan
DOI: 10.1101/2021.04.28.21256277
sha: 5d632aa182a0855b6f5f3922f031d7a09fb76b15
doc_id: 926275
cord_uid: y06wxvg7

In critically ill COVID-19 patients, the risk of long-term neurological consequences is just beginning to be appreciated. While recent studies have identified that there is an increase in structural injury to the nervous system in critically ill COVID-19 patients, there is little known about the relationship of COVID-19 neurological damage to the systemic inflammatory diseases also observed in COVID-19 patients. The purpose of this pilot observational study was to examine the relationships between serum neurofilament light protein (NfL, a measure of neuronal injury) and co-morbid cardiovascular disease (CVD) and neurological complications in COVID-19 positive patients admitted to the intensive care unit (ICU). In this observational study of one-hundred patients who were admitted to the ICU in Tucson, Arizona between April and August 2020, 89 were positive for COVID-19 (COVID-pos) and 11 were COVID-negative (COVID-neg). A healthy control group (n=8) was examined for comparison. The primary outcomes and measures were subject demographics, serum NfL, presence and extent of CVD, diabetes, sequential organ failure assessment score (SOFA), presence of neurological complications, and blood chemistry panel data. COVID-pos patients in the ICU had significantly higher mean levels of Nfl (229.6+163 pg/ml) compared to COVID-neg ICU patients (19.3+5.6 pg/ml), Welchs t-test, p =.01 and healthy controls (12.3+3.1 pg/ml), Welchs t-test p =.005. Levels of Nfl in COVID-pos ICU patients were significantly higher in patients with concomitant CVD and diabetes (n=35, log Nfl 1.6+.09), and correlated with higher SOFA scores (r=.5, p =.001). These findings suggest that in severe COVID-19 disease, the central neuronal and axonal damage in these patients may be driven, in part, by the level of systemic cardiovascular disease and peripheral inflammation. Understanding the contributions of systemic inflammatory disease to central neurological degeneration in these COVID-19 survivors will be important to the design of interventional therapies to prevent long-term neurological and cognitive dysfunction.

Neurological complications following SARS-CoV-2 infection and been reported by a number of investigators (1), (2) , (3) . In addition, there is overlap of the risk factors in patients with severe COVID-19 and patients at risk for Alzheimer's Disease Related Dementias (ADRD) and vascular contributions to cognitive impairment and dementia (VCID). These include age, hypertension, diabetes, cardiac disease, hypercholesterolemia, and pulmonary disease.

Recent studies have shown that of ICU admitted COVID-19 patients in France, 84% showed some level of neurological impairment during their hospital admission (4) and there is a high prevalence of neurological involvement in critically ill patients suffering from COVID-19 (5), (6) , (7) , (8) . It has been suggested that SARS-CoV-2 results in damage to the CNS damage via a surge of systemic inflammatory cytokines called Cytokine Storm Syndrome (CSS) (9) , (10) .

Neurofilament light protein (Nfl) is one of the 3 primary neurofilament isotypes that have been shown to increase in both the cerebrospinal fluid (CSF) and blood in the presence of axonal damage and neurodegeneration (11) . Levels of serum Nfl have been found to be elevated in subjects with a number of neurological degenerative diseases (12) , (13) , (14) , (15), as well as those with acute conditions such as hypoxic brain injury (16) , cardiac and related surgeries (17) , (18) , and traumatic brain injury (19) . Recent studies have found increased levels of Nfl in COVID-19 patients (20) , (21) ,(22) but the effects of concomitant systemic inflammatory disease such as cardiac disease or diabetes on neuronal injury as measured by NfL in COVID-19 patients has not been previously reported. In the present study we examined the relationship between levels of serum Nfl in COVID-19 positive ICU patients and the presence of cardiovascular disease, diabetes, and acute neurological complications.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint

One-hundred patients admitted to the ICU in Tucson, Arizona, USA, were included in this study. Of these, 89 patients were positive for COVID-19 with an average age of 60.8. Of these 89, full clinical data was available on 50 patients. All 50 of the COVID-19 positive (COVIDpos) ICU patients were mechanically ventilated. We compared the COVID-19-pos ICU Nfl data to 11 patients who were admitted to the ICU but were COVID-negative (COVID-neg), with an average age of 65.1. Of these 11, full clinical data were available on 5 patients. Of these five patients, four were mechanically ventilated and one was on a high flow nasal cannula system.

Blood samples for the Nfl assay were obtained on the day of admission to the ICU. The Nfl assay on samples from the ICU patients was performed in the Clinical Neurochemistry Laboratory at the Sahlgrenska University Hospital using the Single molecule array (Simoa) NFlight Advantage and HD-X Analyzer, (Quanterix, Billerica, MA). A single batch of reagents was used; intra-assay coefficients of variation were below 6.8% for all analytes. The Nfl assay on samples from the healthy control cohort were performed by PBL Assay Science, New Jersey, USA, and analyzed using the same Simoa NF-light Advantage kit on an HD-X Analyzer, as per the manufacturer's instruction (Quanterix, Billerica, MA).

Upon admission, patient's medical histories and clinical data including primary admission diagnosis, their SOFA score (the SOFA score (23) is derived from scores from six organ systems ranging from 0 (no organ dysfunction) to 4 (severe organ dysfunction) and the individual organ scores are then summed to a total score between 0 and 24), the Glasgow Coma Score (assessed without sedation), Systemic Inflammatory Response Syndrome (SIRS) evaluation, sepsis evaluation, pulmonary function including ventilation rates and volumes, days . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) 

Experimental values are expressed as mean ± SE unless otherwise indicated.

Comparison of the Nfl levels and log Nfl levels between the 2 ICU groups and the control were analyzed with Welches' t-test with significance determined at a p value < .05. The Welches' ttest was chosen due to the unequal variances and unequal sample sizes between the COVIDpos and COVID-neg groups. One COVID-pos patient Nfl value of 14,555 pg/ml was considered an outlier and not included in the statistical analysis. Comparison of log Nfl levels across subgroups were analyzed with Kruskal-Wallis ANOVA with p < .05 considered significant. The Kruskal-Wallis ANOVA was chosen due to the non-normal distribution of Nfl levels across groups. Associations between log Nfl levels and other co-morbidities and clinical laboratory variables were analyzed using Pearson correlation and a p value < .05 indicated statistical significance. Linear regression was used to generate a best-fit line. Data were analyzed using GraphPad Prism 8.0.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint

This study and use of patient samples was approved by the University of Arizona Institutional Review Board (IRB# 1410545697). Participant informed consent was provided either directly by the patient or, if the participant was incapacitated, from the patient's legally authorized representative.

All ICU patients had either a rt-PCR confirmed positive or negative result for SARS-CoV-2 infection. The three groups analyzed included 89 COVID-pos ICU patients, 11 COVID-neg ICU patients, and 8 healthy controls. Both ICU patient groups were recruited from Tucson, AZ between April-August 2020. The healthy adult participants were recruited independently from emergency care personnel in the hospital. Age deciles for each group are detailed in Table 1 .

The average age of the COVID-pos ICU patients was 60.8, the average age of the COVID-neg ICU patients was 65.1, and the average age of the healthy controls was 51.7. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint Demographic and initial admission clinical data for both COVID-pos and COVID-neg ICU patients are detailed in Table 2 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint Neurological complications (neurocx) were assessed upon hospital admission and during ICU stay. Of the 50 COVID-pos patients, 5 were diagnosed while in the ICU with acute encephalopathy, 5 with ICU delirium, 4 with sleep apnea, 2 with seizure, 5 with depression, one with cerebral edema, 2 with stroke, and one with facial droop. None of the 5 COVID-neg ICU patients had any noted neurological complications.

In COVID-pos patients in the ICU, Nfl was significantly higher than those observed in COVID-neg ICU patients and healthy control ( Figure 1A) . To determine if the levels of Nfl in groups were related to subject age, we first performed Pearson correlation analysis of all log Nfl values compared to all group ages ( Figure 1B) . In this case, the levels of Nfl were significantly positively correlated with age, (r = .335, p =.005, Pearson), as has been previously reported (12) , (13) . However, when we performed correlation analysis of Nfl values compared to age in COVID-pos ICU patients alone (Figure 1C) , there was no significant correlation of Nfl levels with the age of the COVID-pos patient (r = .006, p = .96).

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

There is a clear and well established association between the presence of CVD and the risk of cognitive impairment, neurodegenerative disease and vascular dementia (26) , (27) , (28) , (29) ,(30),(31), (32), (33) . In the present study, we tested the hypothesis that levels of Nfl in COVID-pos ICU patients would be associated with the presence of CVD and diabetes. As seen in Figure 2F , of the 50 COVID-pos patients, 38 had at least one form of CVD with hypertension being the most prevalent with 29/50 COVID-pos patients having a diagnosis of . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint hypertension. Figure 2D illustrates the distribution and variance of CVD plotted against the level of Nfl for each cardiovascular disease category. As seen in Figure 2A . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint

To determine if the levels of Nfl were related to the presence of neurocx, we compared Nfl levels in COVID-pos ICU patients in those that had at least one reported neurocx to those that had none. Of the 50 COVID-pos ICU patients, 23 exhibited at least one or more neurocx and 27 had no neurocx. Of these 23, all also had at least one or more noted diagnosis of CVD.

The types and distributions of neurocx in the COVID-pos patients are seen in Figure 3A and 3C. None of the 5 COVID-neg ICU patients exhibited any neurocx. Figure 3B . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint

All 50 of the COVID-pos patients exhibited sepsis and multiple organ system failure and systemic inflammatory response syndrome (SIRS), all of which have been associated with post ICU cognitive impairment (24) , (34) . We evaluated the association between COVID-pos Nfl levels and the ICU clinical status and sequential organ failure assessment core (SOFA). As seen in Figure 4A . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint

In our cohort of 100 patients hospitalized in the ICU in Tucson, Arizona between April 2020 and August 2020, 89 tested positive for COVID-19 and 11 tested negative. We measured serum levels of Nfl in all of these patients and compared them to each other and a cohort of healthy controls (n=8). The mean levels of Nfl in the COVID-pos patients were significantly higher as compared to those observed in COVID-neg patients and healthy controls and suggest that COVID-pos ICU patients may have CNS injury. Of the 50 COVID-pos patients from which we had clinical data, 70% also had at least one or more diagnosis of one of the subtypes of CVD that included heart failure, hypertension (defined as > 140-159/90-99 mm Hg), coronary artery disease, valvular disease, arrhythmia, hyperlipidemia, obesity, and smoking . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint history. The levels of Nfl in these patients with CVD were 205% higher than in those COVID-pos patients those with no CVD. Likewise, those COVID-pos patients with diabetes had 102% higher levels of Nfl than those without diabetes. The levels of Nfl were similar in COVID-pos patients with CVD alone and no neurocx as compared to COVID-pos patients with both CVD and neurocx, suggesting that the increased levels of Nfl in COVID-pos patients with CVD was not dependent on the presence or absence of neurocx.

Numerous studies have shown that Nfl levels in both the serum and CSF have been validated to be able to detect brain injury and axonal damage in individuals with neurodegenerative diseases (12) , (13) , (14) , brain trauma (19) as well as cardiovascular disease and cardiac surgery (17), (18) . A number of studies have shown that cognitive impairment and neurodegenerative disease are correlated with vascular disease, inflammation and decreased cerebral brain blood flow (26) , (27) , (28) , (29) ,(30), (31) ,. Mechanisms thought to contribute to cognitive impairment in patients with chronic CVD and diabetes include high levels of systemic inflammation (35) , altered cerebrovascular autoregulation (38) , and microembolism (39) .

Inflammatory processes play an important role in CVD-related increases in circulating inflammatory mediators are seen in the brain.

An early study of 214 COVID-19 patients in China was among the first to report evidence of neurological complications of SARS-CoV2 infection (5) . The symptoms in these patients were categorized into three areas including changes to 1) the central nervous system (CNS) which included patient reports of dizziness, headache, ataxia, impaired consciousness, and acute cerebrovascular disease; 2) the peripheral nervous system (PNS) symptoms (loss of smell, taste and some loss of peripheral sensation); and 3) skeletal muscular dysfunction. In these studies, 41.1% of the subjects exhibited severe COVID-19 disease requiring ventilation.

In these patients, 36.3% also had hypertension and 45.5% had neurological manifestations. It has been suggested that the CNS complications in COVID-19 may be due ACE2 expression in . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint the CNS and the PNS (40) . Given that ACE2 is known to be expressed in brain endothelial cells within the cerebrovascular system and on glia and neurons (41) , it has been suggested that SARS-Cov-2 binding to ACE2 in brain vascular endothelium may result in a compromised blood-brain-barrier (BBB). This compromise would allow entry of the virus into the brain parenchyma leading to virus induced neuronal inflammation and damage (42) , (43) .

We also found a significant positive correlation of Nfl levels with ICU clinical status and the SOFA score and four blood chemistry measurements including creatinine, prothrombin time Increases in NT-pro-BNP is a well-known clinical biomarker for heart disease and has also been shown to be affiliated with microvascular disease in the brain, kidney and heart (49) and cardiac complications and mortality rates in COVID-19 patients (50), (51) .

Taken together, our data suggest that increased levels of Nfl in COVID-19 ICU patients are related to not only to the neurological complications seen in these patients, but also to the presence and extent of chronic inflammatory diseases such as cardiovascular disease and diabetes. In older adults, chronic systemic inflammatory-related diseases, such as vascular disease, heart failure and hypertension that lead to increased brain and systemic inflammation and decreased brain perfusion, are known to increase the risk for dementia and the development of VCID (52), (53) , (54) , (55) . The COVID-19 related cytokine storm and high levels of proinflammatory cytokines along with hypoxia due to respiratory dysfunction and concomitant CVD seen in COVID-19 ICU patients are likely to result in short and long-term cognitive dysfunction and may accelerate pre-existing cognitive deficits (56), (57), (43) .

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint

This study has several limitations. First, there are a limited number of subjects in all groups and an even a further limitation on the numbers of subjects we had access to full clinical data. We were not able to obtain consent from all ICU participants which significantly limited our ability to perform full clinical data comparisons between the COVID-pos and COVID-neg subjects. Also, our healthy control sample, while similar in age-range, was small and included no baseline clinical comorbidity data. Future studies with larger cohorts and complete clinical records will be needed to fully understand the relationship between Nfl and ICU patient status.

Lastly, we have no uniform neurological or cognitive assessments in the ICU patients which makes it impossible to correlate these Nfl levels to cognitive function.

Increased levels of Nfl in COVID-19 ICU patients are correlated with both neurological complications and with the presence of cardiovascular disease and diabetes. Nfl may serve as a biomarker for the risk of cognitive and neurological impairment in COVID-19 patients admitted to the ICU.

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted May 2, 2021. ; https://doi.org/10.1101/2021.04.28.21256277 doi: medRxiv preprint

Pathophysiological Clues to How the Emergent SARS-CoV-2 Can Potentially Increase the Susceptibility to Neurodegeneration

Neurological features and outcome in COVID-19: dementia can predict severe disease

Acute Ischemic Stroke in SARS-CoV, MERS-CoV, SARS-CoV-2: Neurorehabilitation Implications of Inflammation Induced Immunological Responses Affecting Vascular Systems

Neurologic Features in Severe SARS-CoV-2 Infection

Neurological Manifestations of Hospitalized Patients with COVID-19 in Wuhan, China: a retrospective case series

Anticipating and Mitigating the Impact of the COVID-19 Pandemic on Alzheimer's Disease and Related Dementias

Neurological and neuropsychiatric complications of COVID-19 in 153 patients: a UK-wide surveillance study

The emerging spectrum of COVID-19 neurology: clinical, radiological and laboratory findings

Clinical Characteristics of Coronavirus Disease 2019 in China

Hypertension, Thrombosis, Kidney Failure, and Diabetes: Is COVID-19 an Endothelial Disease? A Comprehensive Evaluation of Clinical and Basic Evidence

A Dynamic Cross-Disease Fluid Biomarker for Neurodegeneration

Neurofilaments as biomarkers in neurological disorders

Plasma neurofilament light chain in the presenilin 1 E280A autosomal dominant Alzheimer's disease kindred: a cross-sectional and longitudinal cohort study

):e610-e22. 15. Zetterberg H. Is There a Value of Neurofilament Light as a Biomarker for Neurodegeneration in Parkinson's Disease?

Acute Neurofilament Light Chain Plasma Levels Correlate With Stroke Severity and Clinical Outcome in Ischemic Stroke Patients

Associations between mean arterial pressure during cardiopulmonary bypass and biomarkers of cerebral injury in patients undergoing cardiac surgery: secondary results from a randomized controlled trial

Serum neurofilament light in atrial fibrillation: clinical, neuroimaging and cognitive correlates

A Serum Protein Biomarker Panel Improves Outcome Prediction in Human Traumatic Brain Injury

Neurochemical evidence of astrocytic and neuronal injury commonly found in COVID-19

Serum neurofilament light chain levels in the intensive care unit: comparison between severely ill patients with and without COVID-19

Oudemans-van Straaten HM. SOFA and mortality endpoints in randomized controlled trials: a systematic review and meta-regression analysis

Evaluation of delirium in critically ill patients: validation of the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU)

SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Vascular contributions to cognitive impairment

Alzheimer's disease, and dementia in older persons

Cardiovascular risk factors, cortisol, and amyloid-β deposition in Alzheimer's Disease Neuroimaging Initiative

Cerebrovascular atherosclerosis correlates with Alzheimer pathology in neurodegenerative dementias

Atherosclerosis and risk for dementia

Defining Optimal Brain Health in Adults: A Presidential Advisory From the American Heart Association/American Stroke Association

Pathophysiologic relationship between Alzheimer's disease, cerebrovascular disease, and cardiovascular risk: A review and synthesis

Role of amyloid peptides in vascular dysfunction and platelet dysregulation in Alzheimer's disease

Dissecting the Contribution of Vascular Alterations and Aging to Alzheimer's Disease

Inflammation and Coagulation during Critical Illness and Long-Term Cognitive Impairment and Disability

Elevated levels of interleukin 6 and C-reactive protein associated with cognitive impairment in heart failure

Cognitive impairment in heart failure: a systematic review of the literature

Localized cerebral blood flow reductions in patients with heart failure: a study using 99mTc-HMPAO SPECT

Cognitive dysfunction as a major determinant of disability in patients with heart failure: results from a multicentre survey. On behalf of the GIFA (SIGG-ONLUS) Investigators

Cognitive and brain changes associated with ischaemic heart disease and heart failure

and Mas: the other side of the coin

ACE2: from vasopeptidase to SARS virus receptor

Evidence of the COVID-19 Virus Targeting the CNS: Tissue Distribution, Host-Virus Interaction, and Proposed Neurotropic Mechanisms

Neurological Manifestations of COVID-19 (SARS-CoV-2): A Review

Coronavirus disease 2019 and coagulopathy: other prothrombotic coagulation factors

Hematologic disorders associated with COVID-19: a review

Coagulative Disorders in Critically Ill COVID-19 Patients with Acute Distress Respiratory Syndrome: A Critical Review

Coagulopathy in elderly patients with coronavirus disease 2019

D-Dimer and Prothrombin Time Are the Significant Indicators of Severe COVID-19 and Poor Prognosis

Is microvascular dysfunction a systemic disorder with common biomarkers found in the heart, brain, and kidneys? -A scoping review

Meta-analysis Comparing Outcomes in Patients With and Without Cardiac Injury and Coronavirus Disease 2019 (COVID 19)

Impaired cardiac function is associated with mortality in patients with acute COVID-19 infection

Hippocampal Blood Flow Abnormality Associated With Depressive Symptoms and Cognitive Impairment in Patients With Chronic Heart Failure

Cerebral Hypoperfusion and the Energy Deficit in Alzheimer's Disease

Differing associations between Aβ accumulation, hypoperfusion, blood-brain barrier dysfunction and loss of PDGFRB pericyte marker in the precuneus and parietal white matter in Alzheimer's disease

Heart failure and Alzheimer's disease

Immediate and long-term consequences of COVID-19 infections for the development of neurological disease

Neurological Involvement in COVID-19 and Potential Mechanisms: A Review

Competing Interests MH is founder and CEO of

Pinteon Therapeutics, Nervgen, AZTherapies and CogRx, has given lectures in symposia sponsored by Cellectricon, Fujirebio, Alzecure and Biogen, and is a cofounder of

KB has served as a consultant, at advisory boards, or at data monitoring committees for Abcam

HZ is a Wallenberg Scholar supported by grants from the Swedish Research Council (#2018-02532), the European Research Council (#681712), Swedish State Support for Clinical Research (#ALFGBG-720931), the Alzheimer Drug Discovery Foundation (ADDF), USA (#201809-2016862), the AD Strategic Fund and the Alzheimer's Association

the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 860197 (MIRIADE), and the UK Dementia Research Institute at UCL. KB is supported by the Swedish Research Council

TGB is supported by a Covid-19-focused supplement to a National Institute on Aging grant (3P30AG019610-20S1), the Arizona Department of Health Services and the Michael J. Fox Foundation for Parkinson

We would like to acknowledge Heidi Erikson and Trina Hughes (Tucson, Arizona) and AnnaPfister and Irina Nilsson (Mölndal, Sweden) for their expert technical assistance.Ethics approval and consent to participate.

This study and use of patient samples was approved by the University of Arizona Institutional Review Board (IRB# 1410545697). Participant informed consent was provided either directly by the patient or, if the participant was incapacitated, from the patient's legally authorized representative.

Researchers can request for access to anonymized data from the present study for well-defined research questions that are consistent with the overall research agenda for the cohort. Please contact the corresponding author.