Hypothesis

The aging brain does not actively prune neurons; instead, gut-derived indole-3-propionic acid (IPA) buffers metabolically stressed neurons, delaying failure until protective reserves are depleted, creating a pattern that mimics selective eviction.

Mechanistic basis

• Neuronal insulin resistance reduces GLUT4-mediated glucose uptake, lowering ATP and impairing ion gradients [https://www.pnas.org/doi/10.1073/pnas.2416433122].
• High‑energy neurons (e.g., substantia nigra dopaminergic cells) experience calcium‑driven mitochondrial stress earlier [https://pubmed.ncbi.nlm.nih.gov/20053445/]
• IPA, a tryptophan metabolite, crosses the blood‑brain barrier and activates the pregnane X receptor (PXR), boosting antioxidant enzymes and improving mitochondrial resilience [https://pmc.ncbi.nlm.nih.gov/articles/PMC12632128/]
• By raising the threshold for oxidative‑induced apoptosis, IPA allows neurons to sustain higher workload before ATP deficit triggers failure.
• As microbiome‑derived IPA declines with age, the protective buffer wanes, and neurons fail in order of their intrinsic energy demand, producing the observed regional volume loss (frontal ~12%, temporal ~9% over 34‑97 yr) [https://pmc.ncbi.nlm.nih.gov/articles/PMC9281621/]

Testable predictions

1. Supplementation with IPA in middle‑aged mice will delay the onset of insulin‑resistance markers in hippocampal neurons and preserve synaptic density compared with controls.
2. Metabolomic profiling of human cerebrospinal fluid will show a negative correlation between IPA levels and neuronal insulin resistance markers (e.g., p‑AKT/AKT ratio) across the adult lifespan.
3. Genetic knock‑down of intestinal tryptophanase (the enzyme producing IPA) will accelerate regional atrophy in MRI scans of aged rats, specifically affecting high‑metabolic regions first.
4. Rescue experiments where IPA is administered after the onset of insulin resistance will restore potassium gradient stability and axonal conduction velocity, reversible only before the irreversible damage window (≈60‑79 yr in humans).

Potential falsification

If longitudinal studies reveal that neuronal loss patterns remain unchanged despite sustained elevation of gut‑derived IPA (via probiotics or dietary tryptophan enrichment), the hypothesis that IPA modulates the threshold of metabolic failure would be falsified. Similarly, if IPA supplementation does not affect insulin‑resistance markers or neuronal survival in models where metabolic stress is induced independently of gut microbiota, the proposed buffering mechanism would lack support.