Exercise-induced improvement of glycemic fluctuation and its relationship with fat and muscle distribution in type 2 diabetes

Abstract

Aims: Management of blood glucose fluctuation is essential for diabetes. Exercise is a key therapeutic strategy for diabetes patients, although little is known about determinants of glycemic response to exercise training. We aimed to investigate the effect of combined aerobic and resistance exercise training on blood glucose fluctuation in type 2 diabetes patients and explore the predictors of exercise-induced glycemic response.

Materials and methods: Fifty sedentary diabetes patients were randomly assigned to control or exercise group. Participants in the control group maintained sedentary lifestyle for 2 weeks, and those in the exercise group specifically performed combined exercise training for 1 week. All participants received dietary guidance based on a recommended diet chart. Glycemic fluctuation was measured by flash continuous glucose monitoring. Baseline fat and muscle distribution were accurately quantified through magnetic resonance imaging (MRI).

Results: Combined exercise training decreased SD of sensor glucose (SDSG, exercise-pre vs exercise-post, mean 1.35 vs 1.10 mmol/L, p = .006) and coefficient of variation (CV, mean 20.25 vs 17.20%, p = .027). No significant change was observed in the control group. Stepwise multiple linear regression showed that baseline MRI-quantified fat and muscle distribution, including visceral fat area (β = -0.761, p = .001) and mid-thigh muscle area (β = 0.450, p = .027), were significantly independent predictors of SDSG change in the exercise group, as well as CV change.

Conclusions: Combined exercise training improved blood glucose fluctuation in diabetes patients. Baseline fat and muscle distribution were significant factors that influence glycemic response to exercise, providing new insights into personalized exercise intervention for diabetes.

Keywords: body composition; continuous glucose monitoring; diabetes; exercise; glycemic variability; magnetic resonance imaging.

FIGURE 1

Study design. ASFA, abdominal subcutaneous fat area; HMA, hip muscle area; HSFA, hip subcutaneous fat area; LMA, lower leg muscle area; LSFA, lower leg subcutaneous fat area; M‐TMA, mid‐thigh muscle area; M‐TSFA, mid‐thigh subcutaneous fat area; MRI, magnetic resonance imaging; U‐TSFA, upper thigh subcutaneous fat area; U‐TMA, upper thigh muscle area; VFA, visceral fat area.

FIGURE 2

FGM‐based glucose levels in response to study intervention. FGM‐based glucose levels were assessed before and during the intervention. A, Blue lines and symbols, control group (n = 25); B, red lines and symbols, exercise group (n = 25). Using two‐way repeated‐measures analysis of variance: ^§ p < .05 for the effect of time; ^# p < .05 for the interaction between study intervention and time. Using Bonferroni's multiple comparison test: *p < .05 for the difference between pre and post period in each group. Values are mean ± SD. FGM, flash continuous glucose monitor.

FIGURE 3

Comparison of glycemic variability indexes between control and exercise group. (A) MSG, (B) SDSG, (C) CV, and (D) TIR levels between pre and post period in the control and exercise groups. Blue lines and symbols, control group (n = 25); red lines and symbols, exercise group (n = 25). *p < .05 for the difference between pre and post period in each group. CV, coefficient of variation; MSG, mean sensor glucose; SDSG, SD of sensor glucose; TIR, time in range.