Changes in the cerebral glymphatic system after rehabilitation of COVID-19 patients: A five-year follow-up study

Abstract

Background COVID-19 patients often experience cognitive deficits and sleep disorders, potentially linked to cerebral glymphatic system (GS) dysfunction. Long-term progression remains unclear. Methods A prospective follow-up study was conducted on 17 COVID-19 patients at 1(V1), 2 (V2), and 5 years (V5) post-discharge. Multimodal MRI was used to assess the Analysis Along the Perivascular Space (ALPS) index, White Matter Free Water (WM-FW), Choroid Plexus Volume (CPV), and Global Blood-Oxygen-Level-Dependent–Cerebrospinal Fluid (gBOLD–CSF) Coupling, alongside comprehensive neuropsychological tests. Data were analyzed using Linear Mixed-Effects Models (LME), correlation, and regression analyses (P ' 0.05), with Minimal Clinically Important Difference (MCID) and Receiver Operating Characteristic (ROC) analyses applied to evaluate clinical significance and discriminative performance. Results COVID-19 patients exhibited persistent declines in cognitive and sleep functions, as measured by the Logical Memory Test-B (V5: P ' 0.001) and the Pittsburgh Sleep Quality (V1: P ' 0.001, V5: P = 0.004). The ALPS index remained reduced (V1L: P = 0.026; V2L: P = 0.011; V5R: P = 0.025), while WM-FW fluctuated nonlinearly (V1: P = 0.028; V5: P = 0.045) and CPV was stable (L: χ2 = 0.36, P = 0.83; R: χ2 = 0.27, P = 0.87). gBOLD–CSF showed significant recovery (χ2 = 24.95, P ' 0.001), exceeding the MCID and correlating with word fluency (β = –0.754, P ' 0.001) and sleep quality; ROC analysis confirmed high discriminative power (AUC = 0.853). Conclusion Structural GS impairment may persist up to 5 years, while gBOLD–CSF shows reversible, clinically meaningful recovery. Integrating four imaging markers, this study outlines a “persistent damage–partial recovery–nonlinear fluctuation” trajectory, offering mechanistic insights into SARS-CoV-2 infection and a potential biomarker for long-term monitoring.