IF adults with prediabetes and mild executive-function deficits are randomized for 6 weeks to an intervention that specifically reduces post-prandial glucose variability (CGM-guided meal sequencing + 10-15 min post-meal walking + early eating window), while matched controls receive standard calorie advice only,\n\nTHEN the intervention arm will show larger improvements in executive function (Stroop interference, task-switch cost, N-back accuracy) and psychomotor vigilance than controls, even when between-group HbA1c change is small (<0.2%),\n\nBECAUSE glucose volatility (high MAGE/SD and repeated rapid excursions) is expected to drive transient endothelial stress and cytokine pulses (IL-6, TNF-alpha, hsCRP), which in turn perturb fronto-striatal network efficiency more directly than average glycemia.\n\n## Testable design\n- Population: age 40-65, prediabetes (HbA1c 5.7-6.4), no dementia\n- Duration: 6 weeks\n- Primary exposure: change in MAGE + time-above-range (CGM)\n- Primary outcomes: change in executive composite z-score\n- Secondary outcomes: IL-6, TNF-alpha, hsCRP, resting-state fronto-parietal connectivity\n\n## Falsification criteria\nThis hypothesis is weakened if:\n1. Executive function improves without any measurable reduction in glucose variability, or\n2. Variability decreases but cognitive performance and inflammatory markers do not improve versus control.\n\n## Why this matters\nIf confirmed, metabolic-cognitive prevention should target variability-first phenotypes (dynamic glycemic instability), not just static HbA1c thresholds.