140: Why Alzheimer’s May Be a Metabolic Disease

Bikman emphasizes that the metabolic underpinnings of Alzheimer's are likely established decades before symptoms appear, citing Cunnane's finding that even young women with PCOS (a state of insulin resistance) already show reduced brain glucose uptake. He argues that “anything that improves insulin sensitivity … may represent the most meaningful Alzheimer's prevention strategy available.” He singles out low‑carbohydrate eating because it directly lowers insulin and enables the brain to access ketones, and he notes that for APOE4 carriers whose insulin receptors are already compromised, carbohydrate restriction may be especially important. The protocol is not framed as a treatment for existing dementia but as a long‑term metabolic defense.

Bikman cites a trial where patients taking an MCT‑based ketogenic compound for 90 days showed significant cognitive improvements, particularly those without APOE4, though benefits were also seen in carriers. He notes that multiple studies on exogenous ketones—especially BHB—are being published every year and expresses optimism. He explicitly mentions ‘go BHB’ as one commercial form. The logic is that if the brain is stranded without fuel, simply providing the missing fuel via supplementation could mitigate the energy crisis.

Bikman walks through the mechanistic chain: insulin in the brain not only facilitates glucose uptake via GLUT4 but also promotes neuronal survival, regulates tau phosphorylation (inhibiting tangle formation), clears amyloid‑beta through insulin‑degrading enzyme, and modulates acetylcholine. When insulin signaling fails—whether from systemic insulin resistance that spills into the brain or a primary brain‑specific defect—neurons effectively starve for energy, tau becomes hyperphosphorylated leading to tangles, and amyloid clearance falters, causing plaques to accumulate as a downstream consequence rather than a cause. He stresses that this theory is anchored in mechanism and supported by epidemiology (a 56‑59% increased dementia risk in diabetes, with hazard ratios up to 3.88 for elevated HbA1c) and by genetic evidence from APOE4. Unlike the amyloid hypothesis, which struggled to explain why plaques don't track with cognition, the metabolic theory offers a coherent story from cell biology to population data.

Bikman presents the two fatal problems: (1) up to 60 % of cognitively normal 80‑year‑olds have abundant brain amyloid, and some Alzheimer's patients have modest plaque loads—the correlation is weak; (2) from 2003 to 2021 no new drug was approved after repeated phase‑3 failures of anti‑amyloid antibodies (senisumab, gentab, krenisunab, etc.). Adukanumab (2021) gained accelerated approval despite conflicting trials, was pulled in 2024 after causing brain swelling and microbleeds in over a third of patients. He then recounts the 2006 Nature paper from the University of Minnesota that reportedly showed direct causality between a specific amyloid species and memory loss—widely cited as the seminal proof. An investigation in 2022 uncovered digitally manipulated images across more than 20 papers; the 2006 paper was retracted in 2024, becoming the second‑most‑cited retracted paper ever. Bikman argues this paper buoyed a weakening hypothesis at a moment of crisis and reinforced tunnel vision, and that the fraud extended to the NIH's neuroscience division director, who had 132 papers with doctored data.

Bikman describes how researchers found that APOE4 physically binds to the insulin receptor and traps it inside endosomes, preventing the receptor from reaching the cell surface. Consequently, the receptor is sequestered inside neurons where insulin cannot bind it. This impairs downstream insulin signaling—glucose uptake, ATP generation, tau regulation—even if systemic insulin levels are normal. The effect was age‑dependent and worsened by diet‑induced obesity in animal models. In human APOE‑targeted replacement mice, APOE4 expression led to lower brain insulin signaling, lower brain glucose content, and peripheral hyperinsulinemia. Crucially, when investigators attempted to infuse insulin to boost cognition, APOE4 carriers showed little to no benefit because their receptors were irreversibly trapped. Post‑mortem studies by De la Monte's group showed that homozygous APOE4 carriers had reduced brain insulin expression before the pathology reached advanced stages, indicating that the metabolic failure precedes clinical symptoms. Bikman argues this means APOE4 doesn't just fail to clear amyloid—it breaks the brain's energy machinery directly.

Cunnane's group found that in mild‑to‑moderate Alzheimer's patients, brain glucose uptake was reduced by up to 32 % in cognitive regions compared to healthy controls, yet ketone uptake and metabolism were completely normal—not merely preserved, but matching controls exactly. Three separate studies confirmed this. Bikman then describes his own lab's contribution (2021, Alzheimer's and Dementia): they analyzed RNA‑seq from over 240 post‑mortem brains across four cell types. Genes governing glycolysis were significantly down‑regulated in every neural cell type in Alzheimer's brains, confirming a broad collapse of glucose‑burning machinery. In contrast, genes involved in ketolysis (ketone burning) were not significantly altered in neurons, astrocytes, or microglia. This molecular signature shows that the Alzheimer's brain is not uniformly dysfunctional; the problem is specifically in the insulin‑dependent glucose pathway, while the ketone pathway remains fully operational. Bikman frames this as a tragedy: the brain is “stranded” because, due to modern high‑carbohydrate diets that keep insulin chronically elevated, endogenous ketone production is near zero most of the time, leaving the brain without the fuel it can still use.