Hypothesis

The rate of increase in plasma β‑hydroxybutyrate (BHB) during the early phase of an alternate‑day fast predicts the magnitude of insulin‑sensitivity improvement after 12 weeks of ADF, independent of total ketone concentration or weight loss.

Mechanistic rationale

1. Early BHB surge reflects rapid hepatic acetyl‑CoA flux – Within the first 4 h of fasting, hepatocytes shift from glucose to fatty‑acid oxidation. A steep rise in BHB indicates that mitochondrial HMG‑CoA synthase activity is upregulated and that cytosolic acetyl‑CoA is being diverted to ketogenesis rather than to de novo lipogenesis. This metabolic state activates the nuclear receptor PPARα, which drives expression of fatty‑acid oxidation genes and fibroblast growth factor‑21 (FGF21) secretion.
2. PPARα‑FGF21 axis enhances adipose insulin sensitivity – FGF21 acts on adipose tissue to increase adiponectin release and suppress inflammation via GPR109A signaling on macrophages. In insulin‑resistant individuals (HOMA‑IR > 3.7), this pathway is primed but blunted; a brisk ketone signal can overcome the threshold needed for PPARα co‑activator recruitment.
3. Kinetic marker vs. static level – Measuring only the fasting BHB concentration at a single time point conflates total ketone production with clearance rates. The slope (ΔBHB/Δt) captures the dynamic capacity of the liver to switch substrates, a proxy for mitochondrial flexibility that is lost in obese/T2DM phenotypes.

Testable predictions

• Prediction 1: In a cohort of overweight/obese adults stratified by baseline HOMA‑IR, participants whose BHB rises >0.5 mM·h⁻¹ during the first 4 h of a 30‑h fast will show ≥20 % greater reduction in fasting insulin and HOMA‑IR after 12 weeks of ADF compared with those with slower BHB kinetics, even when total weight loss is matched.
• Prediction 2: Pharmacologic blunting of the early BHB rise (e.g., acute administration of a reversible HMG‑CoA synthase inhibitor) will attenuate the ADF‑induced improvement in insulin sensitivity without affecting overall caloric deficit.
• Prediction 3: Elevated early BHB kinetics will correlate with increased hepatic PPARα target gene expression (Cpt1a, Acox1) and circulating FGF21 levels measured at the end of the fast.

Experimental design

• Recruit 120 participants with BMI 28‑35; measure baseline HOMA‑IR.
• Randomize to ADF (20‑h fast/4‑h feed) or matched daily calorie restriction for 12 weeks.
• On day 1 and day 84, perform a 30‑h fast with blood draws every 30 min for the first 6 h to compute BHB slope.
• Primary outcome: change in HOMA‑IR; secondary: weight loss, hepatic PPARα mRNA (via peripheral blood mononuclear cell proxy), plasma FGF21.
• Statistical test: interaction between BHB slope (high vs. low) and diet arm on HOMA‑IR change using ANCOVA adjusting for baseline HOMA‑IR and weight loss.

If the hypothesis holds, early ketone kinetics become a actionable biomarker to personalize ADF, guiding who benefits most from fasting alone versus needing added caloric restriction or pharmacologic PPARα activation.

Key references – ADF benefits in insulin‑resistant individuals [1]; ketone measurements in ADF [2]; non‑essentiality of Hmgcs2 for metabolic adaptations [3]; inflexibility in obese models [4].