Multimodal neuroimaging in Long-COVID and its correlates with cognition 1.8 years after SARS-CoV-2 infection: a cross-sectional study of the Aliança ProHEpiC-19 Cognitiu

Abstract

Introduction: There is a growing interest in the effect of Long-COVID (LC) on cognition, and neuroimaging allows us to gain insight into the structural and functional changes underlying cognitive impairment in LC. We used multimodal neuroimaging data in combination with neuropsychological evaluations to study cognitive complaints in a cohort of LC patients with mild to moderate severity symptoms.

Methods: We conducted a 3T brain magnetic resonance imaging (MRI) study with diffusion tensor imaging (DTI) and functional MRI (fMRI) sequences on 53 LC patients 1.8 years after acute COVID-19 onset. We administered neuropsychological tests to evaluate cognitive domains and examined correlations with Tract-Based Spatial Statistics (TBSS) and resting state.

Results: We included 53 participants with LC (mean age, 48.23 years; 88.7% females). According to the Frascati criteria, more than half of the participants had deficits in the executive (59%) and attentional (55%) domains, while 40% had impairments in the memory domain. Only one participant (1.89%) showed problems in the visuospatial and visuoconstructive domain. We observed that increased radial diffusivity in different white matter tracts was negatively correlated with the memory domain. Our results showed that higher resting state activity in the fronto-parietal network was associated with lower memory performance. Moreover, we detected increased functional connectivity among the bilateral hippocampus, the right hippocampus and the left amygdala, and the right hippocampus and the left middle temporal gyrus. These connectivity patterns were inversely related to memory and did not survive false discovery rate (FDR) correction.

Discussion: People with LC exhibit cognitive impairments linked to long-lasting changes in brain structure and function, which justify the cognitive alterations detected.

Keywords: Long-COVID; cognition; connectivity; diffusion tensor imaging; multimodal neuroimaging; resting state.

Figure 1

Increased radial diffusivity (RD) (red-yellow) superimposed on the tract skeleton (green).

Figure 2

Significant regression with a negative slope between the memory performance and the resting-state fMRI activity in the right fronto-parietal network. MTG, middle temporal gyrus; PCC, posterior cingulate cortex right and left. The color bar shows the T-scores.

Figure 3

Negative regression between memory performance and resting-state activity in the fronto-parietal network. Abbreviations: act, resting-state fMRI activity; MEM, memory domain; MTG_R, middle temporal gyrus, right; PCC_R, posterior cingulate cortex, right; PCC_L, posterior cingulate cortex, left.

Figure 4

Resting-state functional connectivity (rs-FC) between the left and the right hippocampus, the right hippocampus and the left amygdala, and the left middle temporal gyrus. The color bar shows the T-values. The p values did not survive FDR.

Figure 5

Negative regression between memory performance and the right and left hippocampal functional connectivity with the aforementioned anatomical regions.