Amyloid-Induced Network Resilience and Collapse in Alzheimer's Disease: Insights from Computational Modeling

Abstract

Alzheimers disease (AD) is marked by progressive synaptic loss and cognitive decline, with amyloid-beta ( A {beta} ) accumulation playing a central pathological role. While the molecular effects of A {beta} are well-characterized, the cascade from local synaptic dysfunction to largescale network collapse remains poorly understood. We present a biophysically grounded computational model that integrates small-world network topology, Izhikevich neuron dynamics, short-term plasticity (STP), and A {beta}-induced synaptic pruning to investigate how amyloid burden degrades functional brain connectivity over time. Crucially, we simulate the evolving network dynamics across disease stages, revealing how firing rate, synchrony, variability (CV), and Fano factor evolve in response to progressive structural degradation. Our results demonstrate a two-phase deterioration: initial compensatory dynamics followed by abrupt disintegration of coordinated activity. This temporal dissociation mirrors clinical observations and offers mechanistic insight into AD progression. We further identify a critical amyloid threshold beyond which network collapse becomes inevitable, preceded by early-warning indicators such as declining global efficiency. Network topology, synaptic time constants, and repair mechanisms strongly influence resilience, with small-world networks showing delayed collapse and greater functional compensation. Finite-size scaling reveals that larger networks experience earlier and sharper phase transitions.These findings unify molecular, structural, and dynamical perspectives, suggesting that fluctuations in global efficiency and synchrony could serve as sensitive biomarkers for early detection. Our model offers a predictive, multiscale framework for understanding AD progression and evaluating therapeutic strategies.