Theresa J. Van Lith ¹ HAO LI ¹ Marte van der Wijk ¹ Naomi T. Wijers ² Wouter M. Sluis ³ Marieke J. Wermer ⁴ Frank-Erik De Leeuw ¹ Anton Meijer ⁵ Anil Man Tuladhar ^1*

• ¹ Department of Neurology, Donders Institute for Brain, Cognition and Behaviour, Radboud University, Nijmegen, Netherlands
• ² Department of Neurology, Leiden University Medical Center, Leiden, Netherlands
• ³ Department of Neurology and Neurosurgery, University Medical Center Utrecht, Utrecht, Netherlands, Netherlands
• ⁴ Department of Neurology, University Medical Center Groningen, Groningen, Netherlands
• ⁵ Department of Medical Imaging, Radboud University Medical Centre, Nijmegen, Netherlands

SARS-CoV-2 infection is associated with a decline in functional outcomes and many patients experience persistent symptoms, while the underlying pathophysiology remains unclear. This study investigated white matter (WM) integrity on brain MRI in hospitalized COVID-19 patients and its associations with clinical outcomes, including long COVID.We included hospitalized COVID-19 patients and controls from CORONIS (CORONavirus and Ischemic Stroke), an observational cohort study, who underwent MRI-DWI imaging at baseline shortly after discharge (<3 months after positive PCR) and three months after baseline scanning. We assessed WM integrity using diffusion tensor imaging (DTI) and neurite orientation dispersion and density imaging (NODDI) and performed comparisons between groups and within patients. Clinical assessment was conducted at three and twelve months with functional outcomes including (Modified Rankin Scale (mRS), Post-COVID-19 Functional Status scale (PCFS), Visual Analogue Scale (VAS) and long COVID., for cognitive assessment the Modified Telephone Interview for Cognitive Status (TICS-M) and for mood the Hospital Anxiety and Depression Scale (HADS). Associations between WM integrity and clinical outcomes were evaluated using logistic and linear regression. Results 49 patients (mean age 59.5 years) showed higher overall peak width of skeletonized mean diffusivity (PSMD) (p=0.030) and lower neurite density index (NDI) in several WM regions compared to 25 controls at baseline (p<0.05; FWE-corrected), but did not remain statistically significant after adjusting for WM hyperintensities. Orientation dispersion index (ODI) increased after 3-month follow-up in several WM regions within patients (p<0.05) which remained significant after correction for changes in WMH volume. Patients exhibited worse clinical outcomes compared to controls. Low NDI at baseline was associated with worse performance on the Post-COVID-19 Functional Status scale after 12 months (p=0.018).After adjusting for WMH, hospitalized COVID-19 patients no longer exhibited lower WM integrity compared to controls. WM integrity measured shortly after discharge was generally not associated with clinical assessments, suggesting that factors other than underlying WM integrity play a role in worse clinical outcomes or long COVID.