For Alzheimer’s disease, the central dogma has started to wobble: amyloid may not be the master switch scientists once believed. The amyloid hypothesis held that sticky beta-amyloid plaques kicked off a cascade of neuronal death, and drug development pipelines were built almost entirely around clearing those deposits from the brain.

What followed was an enormous experiment in real time. Dozens of monoclonal antibodies were engineered to target beta-amyloid, guided by biomarkers such as cerebrospinal fluid assays and positron emission tomography imaging. Many agents removed plaque but failed to deliver meaningful cognitive improvement in phase-three clinical trials, exposing a troubling gap between surrogate endpoints and lived outcomes. The mismatch has forced researchers to revisit core neurobiology, from synaptic plasticity to mitochondrial dysfunction, and to ask whether amyloid is a trigger, a bystander or simply a pathological footprint of entropy increase in ageing neural networks.

Attention is shifting toward tau tangles, neuroinflammation driven by microglia, and disturbances in glucose metabolism that alter basic metabolic rate in brain tissue. Combination therapies, adaptive trial designs and richer longitudinal cohorts are beginning to redefine the marginal effects of targeting any single protein. In the quiet of imaging suites and biobank freezers, a field is learning to live with uncertainty, even as patients and families still look for a definitive culprit.