SARS-CoV-2 is associated with changes in brain structure in UK Biobank

  1. Gwenaëlle Douaud
  2. Soojin Lee
  3. Fidel Alfaro-Almagro
  4. Christoph Arthofer
  5. Chaoyue Wang
  6. Paul McCarthy
  7. Frederik Lange
  8. Jesper L. R. Andersson
  9. Ludovica Griffanti
  10. Eugene Duff
  11. Saad Jbabdi
  12. Bernd Taschler
  13. Peter Keating
  14. Anderson M. Winkler
  15. Rory Collins
  16. Paul M. Matthews
  17. Naomi Allen
  18. Karla L. Miller
  19. Thomas E. Nichols
  20. Stephen M. Smith

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

There is strong evidence of brain-related abnormalities in COVID-19 1–13 . However, it remains unknown whether the impact of SARS-CoV-2 infection can be detected in milder cases, and whether this can reveal possible mechanisms contributing to brain pathology. Here we investigated brain changes in 785 participants of UK Biobank (aged 51–81 years) who were imaged twice using magnetic resonance imaging, including 401 cases who tested positive for infection with SARS-CoV-2 between their two scans—with 141 days on average separating their diagnosis and the second scan—as well as 384 controls. The availability of pre-infection imaging data reduces the likelihood of pre-existing risk factors being misinterpreted as disease effects. We identified significant longitudinal effects when comparing the two groups, including (1) a greater reduction in grey matter thickness and tissue contrast in the orbitofrontal cortex and parahippocampal gyrus; (2) greater changes in markers of tissue damage in regions that are functionally connected to the primary olfactory cortex; and (3) a greater reduction in global brain size in the SARS-CoV-2 cases. The participants who were infected with SARS-CoV-2 also showed on average a greater cognitive decline between the two time points. Importantly, these imaging and cognitive longitudinal effects were still observed after excluding the 15 patients who had been hospitalised. These mainly limbic brain imaging results may be the in vivo hallmarks of a degenerative spread of the disease through olfactory pathways, of neuroinflammatory events, or of the loss of sensory input due to anosmia. Whether this deleterious effect can be partially reversed, or whether these effects will persist in the long term, remains to be investigated with additional follow-up.

Article activity feed

  1. Version published to 10.1038/s41586-022-04569-5
  2. Version published to 10.1101/2021.06.11.21258690v4 on medRxiv
  3. Version published to 10.1101/2021.06.11.21258690v3 on medRxiv
  4. NCRC

    Our take

    This study, available as a preprint and thus not yet peer-reviewed, included 394 patients with COVID-19 and 388 individually-matched controls, and evaluated brain imaging data from the UK Biobank collected prior to the COVID-19 pandemic and follow-up scans from February-June 2021. Of 297 hypothesis-driven imaging derived phenotypes (IDPs) and 2,022 exploratory IDPs, most longitudinal differences between the groups were in regions of the brain related to olfactory and gustatory systems, as well as some related to memory function, which are consistent with some of the common symptoms of COVID-19. Results may be limited by residual confounding and it’s not yet clear whether these are acute and reversible or longer-lasting changes due to COVID-19.

    Study design

    case-control;retrospective-cohort

    Study population and setting

    This was an ambidirectional cohort study, leveraging existing imagining data from the UK Biobank imaging study, which had completed 42,729 brain scans before the COVID-19 pandemic and included follow-up imaging data on 798 participants after the beginning of the COVID-19 pandemic (782 of whom had usable data and 394 of whom had COVID-19 between visits). The follow-up imaging study included people who had a COVID-19 diagnosis based on primary care data, hospital records, antigen tests linked through the Public Health datasets in England, or two concordant positive home-based lateral flow kits, and a group of controls from the remaining UK Biobank imaging participants who didn’t have a history of COVID-19 and were individually matched by age, ethnicity, date of birth (+/- 6 months), location of imaging assessment, and date of first imaging assessment (+/- 6 months). 2,260 brain imaging-derived phenotypes were used to describe different aspects of brain structure and function, based on three structural MRI scans (T1, T2 fluid attenuation inversion recovery, and susceptibility weighted MRI) as well as diffusion MRI, resting MRI, and task MRI. The analysis focused primarily on analyzing 332 pre-specified brain regions of interest based on expectations from animal models and post-mortem findings, including those related primarily to olfactory and gustatory function. Beyond that, the full set of imaging derived phenotypes were explored. All analyses were focused on longitudinal differences, considering the difference in imaging between scans (adjusted for baseline function) and were adjusted for multiple comparisons.

    Summary of main findings

    The final study population included 394 patients who had COVID-19 (median age 59.1 years, 57% female) and 388 controls (median age 60.4 years, 57% female), all of whom had imaging prior to and during the COVID-19 pandemic (average time between scans was 3.1 years for both groups). Only 15 of the participants with COVID-19 were hospitalized during their infection. Of 297 olfactory and gustatory regions that passed quality assurance tests, only 8 showed significant differences between groups after adjusting for multiple comparisons. These included reduced grey matter thickness or volume over time in the primary or secondary cortical gustatory and olfactory areas in the left hemisphere in the COVID-19 patients compared to controls. In the exploratory analysis of 2,022 imaging derived phenotypes that passed initial quality checks, the longitudinal difference for COVID-19 cases vs. controls was only significant for four measures: patients with COVID-19 had more prominent reductions the ratio of brain volume to total intracranial volume, reductions in cortical thickness of the parahippocampal gyrus and lateral orbitofrontal cortex, as well as increases in lateral ventricle volume. Many of these changes are linked to memory or olfactory function.

    Study strengths

    Brain imaging data was available from prior to the COVID-19 pandemic and during the COVID-19 pandemic. Controls were individually matched to COVID-19 cases based on several important characteristics.

    Limitations

    Follow-up imaging data during the COVID-19 pandemic were only available for a subset of the people originally evaluated in the UK Biobank imaging study. It is not clear how the COVID-19 cases or controls were selected for follow-up. They had limited power to evaluate differences in imaging between hospitalized and non-hospitalized COVID-19 patients. Despite matching and considering baseline differences, residual confounding, due to clinical comorbidities or other factors is still possible. It is not clear whether the observed changes are a long-term consequence of infection that will persist over time or if they are a more acute manifestation of COVID-19 that will resolve.

    Value added

    This is the first study that includes comprehensive brain imaging data on COVID-19 patients and control from before and during the COVID-19 pandemic.

  5. Version published to 10.1101/2021.06.11.21258690v2 on medRxiv
  6. ScreenIT

    SciScore for 10.1101/2021.06.11.21258690: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    Ethicsnot detected.
    Sex as a biological variablenot detected.
    RandomizationWe randomly permuted the patient-vs-control regressor relative to the ΔIDPs, with 5,000 random permutations.
    Blindingnot detected.
    Power Analysisnot detected.
    Cell Line AuthenticationAuthentication: Separate masks for the left and right olfactory bulbs were generated by manually drawing a binary mask for the right olfactory bulb on an averaged template-space T2 FLAIR volume generated from 713 subjects, and mirroring this to obtain the mask for the left (having confirmed by visual inspection that symmetry in this region allowed for this to be effective).

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    The core UK Biobank pipeline already takes the magnitude data from two echoes to provide a quantitative mapping of T2* signal decay.
    Biobank
    suggested: (HIV Biobank, RRID:SCR_004691)
    We extracted these ROI masks from the FreeSurfer processing and applied them to the T2* and diffusion images to generate additional subcortical IDPs.
    FreeSurfer
    suggested: (FreeSurfer, RRID:SCR_001847)

    Results from OddPub: Thank you for sharing your code.

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    Limitations of this study include the lack of stratification of the patients beyond the information of whether they had been hospitalised (information on O2 saturation levels and details of treatment or hospital procedures is currently available on only a few participants), lack of clinical correlates, small number of participants from Asian, Black or other ethnic background other than White, (minority of) cases and (all of) controls’ COVID-19 status identified through lateral flow test kits that have varied diagnostic accuracy50. For those cases, no distinction is possible at present to determine whether a positive test is due to COVID-19 or thanks to vaccination, so potential cases identified only through lateral flow test in vaccinated participants were not included. Another issue inherent to the recruitment strategy of UK Biobank, based on participants volunteering after being contacted at home for a possible re-imaging session, is the high number of mild cases. This can be seen however as a strength of this study: the majority of the brain imaging publications so far having focussed on moderate to severe cases of COVID-19 (doi.org/10.3389/fnagi.2021.646908), there is a fundamental need for more information on the cerebral effects of the disease even in its mildest from. It is also worth noting that the UK Biobank COVID-19 re-imaging study is an ongoing effort, and self-report on symptoms and treatments, as well as further distinction between vaccination and COVID-19 stat...

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: We did not find any issues relating to colormaps.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    Results from scite Reference Check: We found no unreliable references.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  7. Version published to 10.1101/2021.06.11.21258690v1 on medRxiv