A novel mathematical model for predicting the benefits of physical activity on type 2 diabetes progression

TL;DR

A novel mathematical model is developed that formalizes the link between exercise and short- and long-term glucose-insulin dynamics to predict the benefits of regular exercise on T2D progression and can be the basis for future development of decision support tools able to assist patients and clinicians in tailoring preventive lifestyle interventions.

Abstract

Despite the well-acknowledged benefits of physical activity for type 2 diabetes (T2D) prevention, the literature surprisingly lacks validated models able to predict the long-term benefits of exercise on T2D progression and support personalized risk prediction and prevention. To bridge this gap, we developed a novel mathematical model that formalizes the link between exercise and short- and long-term glucose-insulin dynamics to predict the benefits of regular exercise on T2D progression. The model quantitatively captured the dose-response relationship (larger benefits with increasing intensity and/or duration of exercise), it consistently reproduced the benefits of clinical guidelines for diabetes prevention, and it accurately predicted persistent benefits following interruption of physical activity, in line with real-world evidence from the literature. These results are encouraging and can be the basis for future development of decision support tools able to assist patients and clinicians in tailoring preventive lifestyle interventions.

Figures (4)

• Figure 1: Basal glucose concentration (top panel) and insulin concentration (bottom panel) as a function of $u$ over a five-year simulation horizon (simulations: 60 minutes/session, 3 sessions/week, $\tau_{SI}=150$ days).
• Figure 2: Basal glucose concentration of vigorous exercise ($u=75\%$) for 75 minutes/week and moderate exercise ($u=50\%$) as a function of exercise duration, from 150 to 500 minutes/week, using $\tau_{SI}=180$.
• Figure 3: Time trend of insulin sensitivity over a five-year horizon (simulation: 60 minutes/session, 3 sessions/week, $u = 50\%$, ${S_{I\!,\text{\it target}}=0.18}$, $\tau_{SI}=150$ days). The green curve shows the increase produced by the exercise with respect to the trend observed when no intervention is considered (magenta curve).
• Figure 4: Glucose concentration (top panel) and beta-cell mass (bottom panel) as a function of time obtained using the preliminary version of the model DePaolaEtAl (dotted lines) and the model here proposed (Equations (1), continuous lines) (simulation: 60 minutes/session, 3 sessions/week, $u = 20\%$ and $50\%$, ${S_{I\!,\text{\it target}}=0.18}$, $\tau_{SI}=150$ days).