Vestibular-Visual Reweighting in Persistent Postural-Perceptual Dizziness: A Multilevel Resting-State fMRI Study

Abstract

Persistent postural-perceptual dizziness (PPPD) is a disabling functional vestibular disorder characterized by chronic dizziness and visually and motion-induced unsteadiness that markedly impairs daily activities, yet it lacks objective neurobiological markers. We acquired resting-state functional MRI (rs-fMRI) in 52 patients with PPPD and 50 age- and sex-matched healthy controls (HCs) and analyzed the data using a three-tier approach: (i) functional network connectivity (FNC) of independent component analysis (ICA-FNC), (ii) voxel-wise measures of spontaneous amplitude of low-frequency fluctuations (ALFF) and fractional ALFF (fALFF), and (iii) seed-based connectivity using a priori vestibular and subcortical regions of interest (ROIs; e.g., cerebellar (CB) nodulus, parafascicular thalamus, and caudate). Integrating these analytic tiers, we observed a coherent pattern: broadly increased connectivity of CB and primary visual (VIS) networks together with selective hypoconnectivity between a brainstem–cerebellar (BSC) component and the multimodal vestibular cortex (MVC), oculomotor (frontal eye field [FEF]), and default-mode networks (DMN). Voxel metrics revealed decreased ALFF in parietal and frontal opercular cortices—key vestibular integration regions—contrasting with increased fALFF in mid-cingulate, lateral occipital, and premotor areas. Seed-based mapping identified strengthened thalamo-VIS, striato-limbic, and nodulus-hippocampal connectivity. Importantly, increased BSC-to-VIS coupling correlated positively with depressive symptom severity and state anxiety, but negatively with balance confidence and psychological resilience, linking network imbalance to the biopsychosocial phenotype of PPPD. These findings support a multiscale signature of vestibular cortical disengagement accompanied by maladaptive VIS-CB reinforcement and motivate multicenter validation of network-level markers as adjuncts to symptom-based diagnosis.