Synaptic dysfunction as a major driver in Alzheimer's disease: Beyond the amyloid paradigm

While β-amyloid (Aβ) peptides have dominated Alzheimer's disease (AD) research for decades, the unexpected failure of several anti-amyloid therapies has underscored critical gaps in our understanding of the early pathogenic mechanisms of AD. This review presents a fundamental re-evaluation of the causal chain in AD pathogenesis: the previously dominant amyloid-centric model may need to be shifted toward synaptic dysfunction coupled with glutamate excitotoxicity and neurometabolic failure as potential major drivers of cognitive decline in AD. We gather emerging evidence demonstrating that neuronal hyperactivity and glutamate excitotoxicity-both potentially independent of amyloid deposition-can contribute to synaptic damage at preclinical stages of AD. We propose that the chronic neuronal burden of metabolic overload, rather than amyloid toxicity alone, may trigger a cascade of synaptic deterioration and neuronal death. This metabolic-energetic framework introduces a new biological economy to neurodegeneration, in which neurons are compromised by unbalanced energy demands that are only later amplified by protein toxicity. By integrating recent findings on glutamate dysregulation with amyloid and tau pathology, we present a multifactorial model that emphasizes the temporal primacy and mechanistic interplay of excitotoxicity in early synaptic dysfunction, rather than treating these pathways as parallel or independent events. This model offers an explanation for why amyloid removal so often fails: intervening at the level of downstream synaptic damage is likely too late. This reframing carries profound therapeutic implications: targeting synaptic health and glutamate homeostasis before the onset of irreversible amyloid-driven damage may yield more effective preventive strategies than current approaches that predominantly focus on amyloid removal. This synapse-focused perspective prioritizes synaptic health and repositions amyloid as one contributing factor within a broader, integrated model of AD.