Device-measured baseline physical activity is associated with 24-h CGM glycemic responses and 12-week time-in-range after mixed exercise training in type 2 diabetes: a randomised crossover trial

To evaluate whether baseline habitual physical activity (PA), modeled as continuous moderate-to-vigorous PA (MVPA), is associated with acute glycemic responses to different exercise modalities, and whether acute response is associated with 12-week glycemic status in older adults with type 2 diabetes. A total of 23 participants completed the acute crossover phase of the study, while the final longitudinal analysis for the 12-week intervention included 22 participants due to incomplete data from one individual. Continuous glucose monitoring quantified time in range (TIR, 3.9–10.0 mmol/L). Acute responses were analyzed using linear mixed models including modality, baseline MVPA, and modality × MVPA interaction terms, adjusted for age and sex. Associations between acute 24-h TIR after the mixed session and 12-week mean TIR were tested using Pearson correlation and HbA1c-adjusted partial correlation; stability was evaluated with leave-one-out cross-validation (LOOCV). Adherence interaction analyses were exploratory. Evidence for effect modification by baseline MVPA in the acute phase was limited (Mixed × MVPA: β = -0.263, 95% CI -0.585 to 0.058; P = 0.109). Acute 24-h TIR after mixed exercise was associated with 12-week mean TIR (r = 0.887; partial r = 0.815; both P < 0.001), with stable LOOCV performance (R² = 0.729). Exploratory analyses suggested a baseline MVPA × adherence interaction for long-term TIR (β = 0.69; P = 0.031). Acute CGM response was strongly associated with longer-term glycemic status, whereas acute interaction evidence by baseline MVPA was not definitive. A short CGM-based exercise challenge may help characterize individual response patterns, although external validation is needed. Chinese Clinical Trial Registry, ChiCTR2400085258. Registered 04 June 2024. Retrospectively registered.