Hypothesis

Individuals who exhibit a ≥15 % drop in resting metabolic rate (RMR) within the first 12 hours of a 48‑hour fast, coupled with plasma β‑hydroxybutyrate (BHB) concentrations ≥1.5 mM, will show a statistically significant increase in autophagic flux (measured by LC3‑II/I ratio and p62 degradation in peripheral blood mononuclear cells) compared with participants who fail to meet both criteria, regardless of total caloric intake. Furthermore, a refeed period enriched in nutrient‑dense, low‑calorie foods (≥80 % of calories from vegetables, legumes, and whole grains) will amplify the autophagic response by accelerating mTORC1 reactivation kinetics, whereas an isocaloric, nutrient‑poor refeed will blunt this effect.

Mechanistic Basis

A rapid RMR decline reflects heightened mitochondrial uncoupling and increased AMP/ATP ratio, activating AMPK. AMPK directly phosphorylates ULK1, initiating autophagosome formation, while concurrently inhibiting mTORC1. The rise in BHB serves as both a signaling molecule and a substrate for HDAC inhibition, leading to increased acetylation of autophagy‑related genes and enhanced transcription via FOXO3. NAD⁺ levels rise alongside AMPK activation, stimulating SIRT1 deacetylation of autophagy effectors such as ATG5 and LC3. When these pathways converge, autophagosome formation peaks. Nutrient‑dense refeeds supply micronutrients (e.g., magnesium, zinc) and polyphenols that support AMPK‑SIRT1 signaling and prevent premature mTORC1 rebound, thereby prolonging the autophagic window. In contrast, nutrient‑poor refeeds provoke rapid insulin spikes and mTORC1 reactivation, truncating autophagy despite similar ketosis.

Experimental Design

A randomized, crossover trial with 30 healthy adults (age 20‑40, BMI 22‑27) will undergo two 48‑hour conditions separated by a 4‑week washout:

1. Fast‑Only: Water‑only fast, RMR measured via indirect calorimetry at 0, 6, 12, 24, 36, and 48 h; blood drawn for BHB, lactate, insulin, glucose, NAD⁺/NADH, and PBMC isolation for LC3‑II/I and p62 Western blot.
2. Fast + Nutrient‑Dense Refeed: Identical fast, followed by a 4‑hour refeed providing 500 kcal composed of 80 % nutrient‑dense foods (leafy greens, cruciferous vegetables, legumes, berries) and 20 % lean protein. A parallel arm will receive an isocaloric, nutrient‑poor refeed (refined grains, sucrose, saturated fat) to test the modulation hypothesis. Primary outcome: change in LC3‑II/I ratio from baseline to 48 h. Secondary outcomes: p62 levels, BHB, NAD⁺/NADH, RMR percent change, and inflammatory cytokines (IL‑6, TNF‑α).

Predictions and Falsifiability

• Prediction 1: Participants meeting the RMR‑drop ≥ 15 % and BHB ≥ 1.5 mM threshold will exhibit ≥2‑fold increase in LC3‑II/I ratio versus non‑responders (p < 0.01).
• Prediction 2: Nutrient‑dense refeed will sustain elevated LC3‑II/I ratios 2 hours post‑refeed compared with nutrient‑poor refeed (p < 0.05).
• Prediction 3: NAD⁺/NADH ratio will correlate positively with LC3‑II/I changes (r > 0.5). Falsification occurs if no significant difference in autophagic markers is observed between responders and non‑responders, or if nutrient composition of the refeed fails to alter LC3‑II/I kinetics despite comparable ketosis and RMR changes.

Potential Confounds and Controls

Physical activity will be standardized (light ambulation only) and monitored via accelerometry. Sleep will be standardized to 7‑8 hours per night. Menstrual phase will be tracked in female participants to control for hormonal influences on metabolism. All blood draws will be performed in a fasting state except the designated refeed time points to avoid acute feeding effects on LC3‑II/I.

By linking a readily measurable physiological metric (real‑time RMR) to molecular autophagy readouts, and testing how nutrient quality modulates the duration of the autophagic state, this hypothesis provides a concrete, falsifiable framework to optimize fasting‑based longevity interventions.