Objectives: Limited understanding of the causes of variability in individual response to similar lifestyle changes makes it challenging to design individualized diet plans for diabetes prevention. We previously reported a simulation model of lifestyle disease and tested it using data from DPP and a low carb-low fat diet crossover study. Here we used the model to analyze data from J-DOIT1 to understand the causes of variability and to optimize diet therapy.
Methods: J-DOIT1 is a large-scale intervention study of 2,607 subjects with a 5.5 year follow-up. We selected 60 baseline-matched individuals from placebo and intervention arms. Model was calibrated to each subject’s body weight, FPG, HbA1c time courses to estimate physiology and lifestyle parameters (e.g., RMR, diet intake). Simulation analysis was used to optimize individual diet.
Results: Individuals with the most and least improved biomarkers showed no significant difference in model-estimated energy balance or in self-reported lifestyle at baseline. Neither caloric deficit nor linear combinations of model-estimated parameters were predictive of outcomes, showing the value of our complex physiological model for prediction. The model suggests that unique sets of optimal diets exist for achieving individual health goals (Fig. 1).
Optimized changes in fat and carbohydrate intake for two individuals to reduce body weight by 5-7%.
Optimized changes in fat and carbohydrate intake for two individuals to reduce body weight by 5-7%.
J.H. Chen: None. M. Fukasawa: None. Q. Chen: None. S.P. Burns: None. K. Kumar: None. S. Nirengi: None. K. Takahashi: None. A. Suganuma: None. H. Kuzuya: None. N. Sakane: Research Support; Spouse/Partner; Pwc. G. Dwivedi: None.
