Hypothesis

Individuals with higher wearable‑derived circadian glucose resilience (CGR), defined as the amplitude‑adjusted ratio of nocturnal glucose stability to post‑prandial variability, will show a greater reduction in HbA1c after a 12‑week timed‑carbohydrate‑restriction (TCR) protocol compared with low‑CGR peers.

Mechanistic Rationale

CGM data reveal that nocturnal glucose stability reflects hepatic insulin sensitivity and mitochondrial oxidative capacity, while post‑prandial spikes indicate peripheral muscle GLUT4 translocation efficiency [2]. The CGR metric captures the balance between these two compartments. In high‑CGR subjects, the liver can suppress glycogenolysis during sleep, preserving NAD+ levels that activate SIRT1‑PGC1α signaling, enhancing mitochondrial biogenesis [3]. When carbohydrates are confined to the early‑day window, muscle contraction‑induced AMPK activation synergizes with this primed oxidative state, driving greater glucose disposal and lowering HbA1c. Low‑CGR individuals, by contrast, exhibit nocturnal hepatic glucose output that blunts SIRT1 signaling, limiting the benefit of timed restriction.

Testable Predictions

1. Baseline CGR will correlate positively with change in HbA1c after TCR (r > 0.4, p < 0.01).
2. The intervention will reduce fasting glucose more in high‑CGR than low‑CGR subgroups (interaction p < 0.05).
3. CGR will mediate the relationship between baseline nocturnal glucose variability and HbA1c change (bootstrap indirect effect CI excludes zero).
4. No significant HbA1c change will occur in a control group receiving uniform carbohydrate distribution.

Experimental Design

• Recruit 120 prediabetic adults (HbA1c 5.7‑6.4%) from a community cohort.
• Collect 14 days of baseline CGM + actigraphy to compute CGR: (mean nocturnal glucose SD / mean post‑prandial glucose AUC) inverted and z‑scored.
• Randomize participants 1:1 to TCR (carbohydrates 0‑10 g before 12:00 h, ad libitum after) or control (isocaloric meals spread evenly).
• Blind outcome assessors; participants receive weekly CGM summaries but not group assignment.
• Primary outcome: ΔHbA1c at week 12. Secondary: fasting glucose, weight, sleep‑HRV.
• Statistical plan: ANCOVA adjusting for baseline HbA1c, age, BMI; test interaction CGR×group; mediation analysis using structural equation modeling.

Potential Confounds and Mitigation

• Sleep disruption could affect nocturnal glucose independent of metabolic resilience; include actigraphy‑derived sleep efficiency as covariate.
• Dietary adherence monitored via photo food logs and weekly dietitian review; non‑adherers (>20% deviation) excluded in per‑protocol analysis.
• Genetic variability in TCF7L2 may influence CGR; genotype participants and explore as moderator in exploratory analysis.

If the hypothesis fails—i.e., CGR does not predict TCR response—it would challenge the notion that wearable‑derived nocturnal glucose metrics reflect actionable mitochondrial resilience and suggest that alternative biomarkers (e.g., circulating metabolites) are needed for precision nutrition.