Synapse loss is the strongest structural correlate of cognitive decline in Alzheimer's disease (AD), yet the prevailing narrative treats it as a purely degenerative process. Here we propose that a substantial fraction of early synaptic change in AD represents an adaptive, energy-conserving response by the metabolically prioritising brain — one that becomes self-defeating as the underlying energy crisis deepens. We integrate three recent lines of evidence: (i) Peters' Selfish Brain theory, which posits that the brain prioritises its own glucose supply at the expense of peripheral organs; (ii) the discovery that approximately 25% of synapses in adult mouse neocortex are functionally silent, residing on NMDA-receptor-only filopodia that lack AMPA receptors (Vardalaki et al., Nature, 2022); and (iii) the finding that memantine, the standard NMDA-receptor antagonist used in AD, also partially inhibits calcium-permeable AMPA receptors (Carrillo et al., Nature Communications, 2025). We propose a four-stage model: (1) preclinical amyloid accumulation triggers cerebral glucose hypometabolism, prompting adaptive synaptic silencing through AMPA-receptor internalisation and morphological regression from mature spines to filopodia; (2) the silent synapse reservoir expands while the glucose-dependent molecular machinery for reactivation — CaMKII signalling, AMPA-receptor trafficking, Arp2/3-mediated actin remodelling — becomes progressively unavailable; (3) brain-pull incompetence triggers systemic metabolic compensation, creating a feed-forward loop; (4) the structural scaffold itself degrades through complement-mediated pruning of inactive synapses, rendering connectivity collapse irreversible. The framework generates several testable predictions, including a non-monotonic filopodial density curve across AD stages, and identifies the Stage 2–3 transition as a critical therapeutic window. We discuss pharmacological implications, arguing that memantine may inadvertently narrow the plasticity window, and propose that combination strategies pairing NMDA-receptor modulation with AMPA-receptor positive allosteric modulators could restore the conditions under which the silent synapse reservoir can be recruited.
Topcular et al. (Mon,) studied this question.