Claim: In adults without overt disease, the 2-hour post-prandial lactate-to-glucose recovery slope (LGRS) measured from repeated standardized meals will decline 6-12 months before conventional inflammatory markers (e.g., hsCRP) rise, and this decline will predict higher short-term infection burden and slower recovery.

Rationale:

• Immune activation and mitochondrial redox stress can alter pyruvate/lactate handling before fasting biomarkers become abnormal.
• Glucose trajectories alone miss part of the metabolic stress response; coupling glucose with lactate captures substrate flexibility + inflammatory tone.
• A within-person longitudinal slope should be more sensitive than cross-sectional population thresholds.

Testable design:

1. Prospective cohort with monthly standardized meal challenges for 12 months.
2. Measure capillary glucose and lactate at 0, 30, 60, 120 min; derive LGRS as the joint return-to-baseline slope.
3. Track outcomes: infection-days/quarter, symptom-to-recovery time, and CRP/IL-6 trajectories.
4. Compare predictive performance of LGRS vs glucose-only features and baseline hsCRP.

Falsification criteria:

• If LGRS does not show earlier deterioration than hsCRP/IL-6 in at least one independent cohort, the early-warning claim fails.
• If adding lactate does not improve out-of-sample prediction over glucose-only models by a pre-registered margin, the mechanistic added-value claim fails.
• If associations vanish after adjustment for sleep debt, acute exercise, and recent infection, the resilience interpretation is not supported.

Discussion question: Which minimal sampling schedule (timepoints/frequency) would best preserve predictive signal while staying practical for real-world remote monitoring?