White-matter functional network dysfunction associated with cognitive deficits and clinical phenotypes in patients with end-stage renal disease

Abstract Among patients with end-stage renal disease (ESRD), the white matter (WM) is a particularly vulnerable area that is susceptible to various clinical risk factors. However, whether WM function is disrupted in ESRD patients and how this disruption provides valuable information for cognitive deficits and potential clinical phenotypes remain unknown. We prospectively enrolled 78 ESRD patients and 50 healthy controls. Using resting-state functional magnetic resonance imaging, we studied ESRD-related WM functional networks alterations. Functional connectivity, functional covariance connectivity, and coefficient Granger causality analysis were probed interactions among WM functional networks. The machine-learning models with leave-one-out cross-validation were applied. ESRD patients exhibited extensively disrupted interactions among WM functional networks, which correlated with cognitive deficits and ESRD-specific clinical risk factors, including uremic toxin accumulation, dysregulation of calcium-phosphorus homeostasis, and anemia. A random forest classifier achieved a maximum performance of 95.31% accuracy and 0.982 area under the ROC curve (AUC). Our results emphasized the imbalances of WM functional networks in ESRD patients, which might be used as potential neuroimaging markers for cognitive deficits and potential clinical phenotypes.