Moving After Meals: Timing of Free-Living Activity and Postprandial Glycemia in CGMacros

Background. Controlled studies show that walking after meals can lower glucose, but free-living wearable records do not represent assigned exercise. Objective. To test whether meal-to-meal variation in wrist-derived metabolic equivalents (METs) was associated with two-hour glucose exposure and whether timing changed the association. Methods. I analyzed CGMacros version 1.0.0. The glucose sample contained 1,329 isolated meals from 42 adults; activity analyses included 1,047 meals from 32 adults. Positive incremental area under the glucose curve (iAUC) and peak rise were reconstructed over 120 minutes. Participant fixed-effects models estimated within-person associations adjusted for meal composition, baseline glucose, and meal type, with standard errors clustered by participant. Eight 15-minute windows were tested, and false-discovery-rate correction was applied within each outcome. Results. For meals with at least 80% activity coverage, the mean METs across 0 to 120 minutes were not associated with iAUC (β = -223.33 mg/dL × min per MET, 95% CI [-482.83, 36.17], p = .089) or peak rise (β = -2.60 mg/dL per MET, 95% CI [-5.66, 0.47], p = .094). Most timing coefficients were negative, but no post meal window remained significant after correction. Carbohydrates did not modify the early association, and slopes did not differ by glycemic group. One exploratory premeal window was associated with lower peak rise after correction (β = -3.11 mg/dL per MET, q = .024). Conclusions. Greater wrist-derived activity was not reliably associated with lower postprandial glucose after multiplicity correction. The mostly inverse estimates are compatible with benefit but too imprecise for a causal or prescriptive conclusion.