Cerebral microstructural alterations in Post-COVID-condition are related to cognitive impairment, olfactory dysfunction and fatigue

Abstract

After contracting COVID-19, a substantial number of individuals develop a Post-COVID-Condition, marked by neurologic symptoms such as cognitive deficits, olfactory dysfunction, and fatigue. Despite this, biomarkers and pathophysiological understandings of this condition remain limited. Employing magnetic resonance imaging, we conduct a comparative analysis of cerebral microstructure among patients with Post-COVID-Condition, healthy controls, and individuals that contracted COVID-19 without long-term symptoms. We reveal widespread alterations in cerebral microstructure, attributed to a shift in volume from neuronal compartments to free fluid, associated with the severity of the initial infection. Correlating these alterations with cognition, olfaction, and fatigue unveils distinct affected networks, which are in close anatomical-functional relationship with the respective symptoms.

Fig. 1. Gray matter DMI parameters indicate microstructural changes between groups.

Box plots show the distribution of microstructural compartments within the entire gray matter. In patients with Post-COVID-Condition (PCC), a significant decrease in the extraneurite volume fraction (V-extra) was accompanied by a significant increase in the free-fluid fraction (V-CSF) when compared to Unimpaired Post-COVID (UPC) patients and the Healthy Non-COVID (HNC) group. Regarding the intraneurite volume fraction (V-intra), no statistically significant differences were present between groups. Group comparisons were performed using two-tailed ANCOVAs, with age and sex as nuisance covariates. The statistic was based on 46 HNC, 38 UPC, and 89 PCC cases. The center lines represent the median, box bounds show interquartile range (IQR), whisker borders indicate 1.5 x IQR. Dots represent individual subject values.

Fig. 2. Spatial distribution and direction of microstructural changes after COVID-19.

A The standardized regression coefficients of the factors V-extra and V-CSF were extracted from two-tailed linear regression models yielded by voxel-wise comparisons between all participants after COVID-19 infection (i.e., combined PCC and UPC group) and controls without a history of COVID-19 (HNC; with nuisance covariates age and sex), and were superimposed onto a T1w MRI template. Color-coding indicates the coefficient values as a measure of effect size of the factor COVID-19 (hot colors: positive effects vs. cold colors: negative effects). Results of voxel-based two-tailed linear regression models of V-extra after threshold-free cluster enhancement and FWE-correction (top row), and standardized regression coefficients derived from the same model (bottom row) between different groups: B Post-COVID-Condition (PCC) vs. Healthy Non-COVID controls (HNC), C. Unimpaired-Post-COVID participants (UPC) vs. Healthy Non-COVID controls (HNC), and (D). Post-COVID-Condition (PCC) vs. Unimpaired-Post-COVID (UPC). Voxels with significantly different V-extra were indicated by hot shading and superimposed onto a T1w MRI template (top rows). Two-tailed P values were corrected for multiple comparisons across voxels using the family-wise error rate (FWE), with age and sex as nuisance covariates. Color-coding indicates the coefficient values as a measure of effect size of the factor COVID-19 (hot colors: positive effects vs. cold colors: negative effects; bottom rows). Radiological orientation: left side of the image corresponds to the patient’s right; numbers denote the axial (z) position in millimeters. The statistic was based on 46 HNC, 38 UPC, and 89 PCC cases.

Fig. 3. Significant association between V-extra and symptoms associated with Post-COVID-Condition.

To relate clinical outcomes to V-extra in patients with a previous COVID-19 infection (i.e., combined PCC- and UPC-group), voxel-based two-tailed linear regression models were employed, with V-extra as a dependent variable, covariates age and sex, and threshold-free cluster enhancement. Two-tailed P values were corrected for multiple comparisons across voxels using the family-wise error rate (FWE). Voxels with significantly different V-extra are indicated by hot shading and superimposed onto a T1w MRI template. Radiological orientation, i.e., left side of the image corresponds to the right side of the patient’s body; numbers denote the axial (z) position in millimeters. This statistic was based on n = 127 patients with previous COVID-19 infection.