Abstract
Abstract
High prevalence of HIV-associated neurocognitive impairment persists despite effective antiretroviral therapy. While conventional neuroimaging has been previously used to explain cognitive deficits in virally suppressed people with HIV (PWH), magnetic resonance elastography (MRE) offers a novel approach to quantify microstructural tissue coherence through brain biomechanics. This study investigated whether MRE-derived biomechanical metrics exhibit a pathological, group-specific relationship with cognitive performance in PWH, and compared their sensitivity against advanced diffusion MRI and peripheral biomarkers of neuronal injury and inflammation. In a cross-sectional cohort of 27 virally suppressed PWH and 30 matched healthy controls, participants underwent comprehensive cognitive testing, multi-frequency MRE (measuring stiffness, viscosity, and strain), multi-shell diffusion MRI (including NODDI metrics), and blood sampling for plasma neurofilament light chain and glial fibrillary acidic protein. PWH exhibited significantly lower cognitive scores than controls, particularly in learning and memory. Although baseline mean values across imaging and peripheral biomarkers did not notably differ between the groups, a significant group-by-cognition interaction emerged exclusively for MRE parameters. Specifically, in PWH, elevated viscosity and reduced strain correlated with better attention, whereas lower viscosity and strain correlated with poorer memory. These biomechanical-behavioral relationships were absent in healthy controls, and no notable interactions were observed for advanced diffusion metrics or blood biomarkers. Ultimately, brain biomechanics measured by MRE demonstrate greater sensitivity to altered cognitive function in PWH than advanced diffusion MRI and peripheral biomarkers, highlighting their potential as functional correlates of neurocognitive impairment that warrant further longitudinal validation.