Cerebellar Degeneration Induces Timeliness Disruption, Paradoxical Information Flow and Network Overdrive in a Digital Model of Schizophrenia

Among the main etiopathological hypotheses of schizophrenia (SZ), a prominent view attributes a central role to cerebellar dysfunction, driven by atrophy and altered cellular communication predominantly at posterior zones. Within this framework, the glutamatergic hypothesis is particularly relevant: dysregulation of excitatory transmission may, over time, induce secondary and bidirectional alterations in the dopaminergic system, contributing to symptom onset and progression. However, how microcircuit-level alterations propagate to systems-level dysfunction remains an open question. Here, we develop a multiscale computational framework of the cerebellum that integrates a continuum of morphometric degeneration and synaptic alterations grounded in SZ experimental literature. The model systematically explores progressive atrophy and excitatory imbalance, while incorporating a circuit-level resilience parameter reflecting evidence that cerebellar resilience modulates vulnerability to symptom emergence. Under baseline stimulation, cerebellar cortical atrophy induces a modest early hyper-responsiveness in Purkinje cells, whereas the deep cerebellar nuclei exhibit pronounced, resilience-modulated overactivity (from 12Hz up 27Hz). This activity pattern is consistent with reported alterations in cerebello–cortical functional connectivity and may mechanistically contribute to the overproduction of striatal dopamine observed in SZ. Under saccade-like mossy fiber stimulation, the model reveals disrupted mossy fiber–Purkinje cell interactions, characterized by impaired temporal alignment consistent with neural correlates of cognitive and sensorimotor dysmetria (from 7ms up to 13ms delay). Notably, we observe a paradoxical, non-monotonic pattern of transfer entropy, with peak transmission values emerging at low levels of degeneration and shifting under preserved resilience. This transmission profile resembles dynamics akin to the optimal brain-damage effect, in which moderate pruning transiently enhances computational efficiency before eventual functional collapse at higher levels of degeneration. Importantly, these results delineate pathological scenarios along a virtual continuum of degeneration and excitatory imbalance, offering a mechanistic framework that may support experimental and clinical researchers in interpreting heterogeneous findings and in understanding the trajectories of SZ progression. This framework provides a unified computational account linking cerebellar microcircuit alterations to impaired timing, information flow, and network overdrive across SZ-like stages.