COVID-19 is not associated with a putative marker of neuroinflammation: A diffusion basis spectrum imaging study

Abstract

COVID-19 remains a significant international public health concern, with its underlying mechanisms not yet fully elucidated. Recent studies suggest the potential for SARS-CoV-2 infection to induce prolonged inflammation within the central nervous system. However, the evidence primarily stems from limited small-scale case investigations. To address this gap, our study capitalized on longitudinal data from the UK Biobank. This dataset encompassed pre- and post-COVID-19 neuroimaging data from a cohort of 416 individuals (M _age=58.6; n=244 female), including 224 COVID-19 cases (M _age=59.1; n=122 females). Employing an innovative non-invasive Diffusion Basis Spectrum Imaging (DBSI) technique, we calculated putative indicators of neuroinflammation (DBSI-RF) for both gray matter structures and white matter tracts in the brain. We hypothesized that SARS-CoV-2 infection would be associated with elevated DBSI-RF and conducted linear regression analyses with adjustment for age, sex, race, body mass index, smoking frequency, and data acquisition interval. After multiple testing correction using false discovery rate, no statistically significant associations emerged between COVID-19 and neuroinflammation variability (all p _FDR>0.05). Nevertheless, several brain regions displayed subtle differences in DBSI-RF values between COVID-19 cases and controls. These regions are either part of the olfactory network (i.e., orbitofrontal cortex) or functionally connected to the olfactory network (e.g., amygdala, caudate), a network that has been implicated in COVID-19 psychopathology. It remains possible that our study did not capture acute and transitory neuroinflammatory effects associated with COVID-19 due to potential symptom resolution before the imaging scan. Future research is warranted to explore the potential time- and symptom-dependent neuroinflammatory relationship with SARS-CoV-2 infection.