Metabolic endotyping for mechanism-informed Alzheimer's and Parkinson's disease trials

Metabolically targeted therapies for Alzheimer's disease and Parkinson's disease have yielded mixed results despite convergent evidence implicating three interconnected axes of metabolic dysfunction: insulin signaling, cellular bioenergetics, and lipid-lysosomal biology. We argue that this pattern partly reflects a trial-design problem: therapeutics such as glucagon-like peptide 1 receptor agonists, terazosin, and stearoyl-CoA desaturase inhibitors generally have been tested in biologically unselected populations, with endpoints often distal to the targeted mechanism. We propose as a testable hypothesis that a metabolic endotype-a measurable, biomarker-supported pattern of metabolic dysfunction-spans subsets of both diseases and that failure to enrich for this endotype could dilute therapeutic signals in all-comer trials. We outline a modular four-tier framework integrating prodromal risk, metabolic vulnerability, pathway-specific confirmation, and mechanistic deep phenotyping to make Alzheimer's and Parkinson's disease trials more interpretable and therapeutically informative.