Hypothesis

Rapid, high-amplitude declines in interstitial glucose preferentially activate AMPK‑driven lipophagy, whereas prolonged, modest glucose reductions favor mTORC1‑dependent mitophagy, such that CGM-derived kinetic parameters (rate of decline, variability, time‑in‑range) can forecast which autophagic substrate is prioritized in peripheral blood mononuclear cells.

Mechanistic Rationale

Autophagy operates as a nutrient‑rationing system where substrate choice hinges on the energetic yield and signaling context of the stressor. Acute ATP depletion sensed by AMPK rapidly phosphorylates ULK1 and ACC, stimulating lipophagy to mobilize fatty acids that yield >100 ATP per palmitate[1](https://pmc.ncbi.nlm.nih.gov/articles/PMC3428762/). In contrast, sustained amino‑acid or mild glucose restriction inhibits mTORC1, shifting ULK1 activity toward selective organelle quality control, particularly mitophagy via PINK1/Parkin recruitment[2](https://doi.org/10.1080/15548627.2019.1586258). CGM captures the temporal shape of glucose excursions: steep drops generate a sharp ATP/AMP surge (AMPK‑dominant), while gradual declines produce a prolonged but modest energy deficit that permits mTORC1 sensing of amino‑acid sufficiency. Thus, the glucose kinetic profile acts as a upstream “nutrient‑stress barcode” that biases the autophagy machinery toward either lipid or mitochondrial catabolism.

Testable Predictions

1. Individuals exhibiting a >30 mg/dL drop in glucose within 15 min (high rate of decline) will show increased plasma long‑chain acylcarnitines (C16:0, C18:1) and elevated LC3‑II flux in PBMCs after a 6‑h fast, indicating lipophagy activation.
2. Those with a sustained glucose reduction of 10‑20 mg/dL lasting >2 h (low variability, extended time‑below‑baseline) will display elevated mitochondrial‑derived phosphatidylethanolamine species (e.g., PE 18:0/22:6) and higher PINK1/Parkin‑dependent mitophagy markers without a proportional rise in acylcarnitines.
3. Pharmacological AMPK inhibition (Compound C) will abolish the lipophagy signal in the rapid‑decline group but not affect mitophagy in the gradual‑reduction group, whereas mTORC1 inhibition (rapamycin) will enhance mitophagy markers only in the latter.

Experimental Approach

• Recruit 30 healthy adults; equip them with blinded CGM devices for 48 h while maintaining a standardized diet.
• Extract glucose kinetic metrics: maximum rate of decline (mg/dL/min), glucose variability (coefficient of variation), and area under the curve below 70 mg/dL.
• After the monitoring period, collect fasting blood; isolate PBMCs and measure autophagic flux via LC3‑II turnover with/without bafilomycin A1, plasma acylcarnitines (targeted LC‑MS), and phosphatidylethanolamine species.
• Validate pathway specificity using siRNA knockdown of ULK1 (AMPK arm) or PINK1 (mitophagy arm) in cultured PBMCs stimulated with autologous serum mimicking the observed glucose conditions.
• Statistical analysis: multivariate regression linking each glucose kinetic predictor to lipid‑ vs. mitochondrion‑derived autophagy biomarkers, controlling for age, sex, and baseline insulin sensitivity.

Potential Implications

Confirming this relationship would transform CGM from a glycometric tool into a dynamic readout of intracellular catabolic programming, enabling personalized timing of fasting‑mimicking diets, exercise, or pharmacologic agents to preferentially mobilize lipids or renew mitochondria based on an individual’s real‑time glucose stress signature. Failure to observe the predicted links would falsify the notion that glucose kinetics directly gate autophagy substrate selection, prompting a revision of the siege model toward more generic energy‑sensing mechanisms.