**Background:** Alzheimer's disease (AD) is a complex neurodegenerative disorder with well-characterized pathological hallmarks—extracellular amyloid-β (Aβ) plaques, intraneuronal tau tangles, glial activation, and synaptic loss—but a unifying initiating mechanism has remained elusive. The apolipoprotein E (APOE) ε4 allele is the strongest genetic risk factor for sporadic AD, conferring a 2- to 4-fold increased risk in heterozygotes and a 14-fold increased risk in ε4 homozygotes, while the ε2 allele is protective (87% lower odds ratio than ε3 homozygotes). APOE is the principal cholesterol carrier in the brain, transporting glia-derived cholesterol to neurons via lipoprotein particles (11–20 nm in diameter) that navigate the extracellular space (ECS), which measures only ~40 nm between adjacent cells. Neurons cannot synthesize cholesterol and depend entirely on this delivery system, particularly for synaptic plasticity underlying learning and memory.
**Methods:** This is a narrative hypothesis paper synthesizing evidence from genetic studies, neuropathology, cell biology, animal models, and clinical trial outcomes. The authors integrate findings from genome-wide association studies (highlighting cholesterol-handling genes CLU, PICALM, BIN1, ABCA7), experimental models (APP transgenic mice crossed with APOE knockout mice, APOE ε4 transgenic mice, cholesterol depletion in neuronal cultures), human neuropathology (immunohistochemistry, electron microscopy), and clinical observations (traumatic brain injury outcomes, chronic traumatic encephalopathy, Niemann-Pick type C disease).
**Key Results:** The hypothesis proposes that age-related changes to the extracellular matrix—exacerbated by hypertension, diabetes, obesity, inflammation, and physical inactivity—cause 'fibrosis' of the narrow ECS, trapping lipoprotein particles. This entrapment has four key consequences: (1) Aβ deposition: Aβ is transported in the hydrophobic core of lipoprotein particles; when trapped particles degrade, Aβ is released into the aqueous ECS where it aggregates into plaques. APOE is essential for plaque formation, as APP transgenic mice lacking APOE do not accumulate plaques. Cholesterol deficiency shifts APP cleavage from Aβ40 to the more aggregation-prone Aβ42. (2) Impaired synaptic function: Cholesterol deficiency impairs synaptic plasticity, with effects most pronounced in hippocampus and association cortex (regions of highest plasticity) and least in primary motor/sensory cortex, cerebellum, and spinal cord—consistent with the hierarchical loss of function in AD. (3) Tau accumulation: Neuronal cholesterol deficiency induces tau hyperphosphorylation, supported by cholesterol depletion in neuronal cultures (preventable by lipoprotein treatment) and the occurrence of tangles in Niemann-Pick type C disease. (4) Glial dysfunction: In APOE ε4 carriers, lipoprotein particles are smaller, carry less cholesterol, and are more prone to aggregation. Microglial activation occurs in response to extracellular Aβ from degraded particles, generating a neurotoxic pro-inflammatory state.
**Clinical Implications:** The hypothesis explains the failure of clinical trials targeting Aβ or tau, arguing these are consequences rather than causes of the primary abnormality—lipoprotein particle entrapment. The AN1792 immunization trial, which cleared Aβ plaques but did not halt cognitive decline, is cited as supporting evidence. The authors propose that future therapies must target the health of the extracellular space, cholesterol homeostasis, and APOE function rather than amyloid or tau alone. Potential therapeutic directions include improving lipoprotein particle mobility through the ECS, enhancing cholesterol delivery to neurons, and targeting age-related ECM changes. The hypothesis also reconciles the apparent paradox that statins (which reduce brain cholesterol) may reduce AD risk, suggesting that the relative balance of cholesterol, Aβ, and APOE—rather than absolute cholesterol levels—is critical. Key limitations acknowledged include the lack of direct demonstration of Aβ within APOE-containing lipoprotein particles in human brain, absence of direct evidence for particle trapping and disintegration in the ECS, and the observation that APOE is present in some but not all plaques (particularly cored/late-stage plaques).