Mathematical modeling of childhood obesity from a social epidemic point of view for the spanish region of Valencia: numerical and analytical solutions
- Autores: Gilberto C. González Parra
- Directores de la Tesis: Lucas Antonio Jódar Sánchez (dir. tes.)
, Rafael Villanueva Micó (dir. tes.) - Lectura: En la Universitat Politècnica de València ( España ) en 2009
- Idioma: inglés
- Tribunal Calificador de la Tesis: Rafael Bru García (presid.) , Juan Carlos Cortés López (secret.) , Fransisco Solis Lozano (voc.) , Benito M. Chen-Charpentier (voc.) , José Antonio Martín Alustiza (voc.)
- Enlaces
- Tesis en acceso abierto en: TESEO
- Resumen
This thesis dissertation deals with the mathematical modeling of childhood obesity from a social epidemic point of view for the Spanish region of Valencia, Three mathematical models based on systems of nonlinear ordinary differential equations of first order were constructed. The first one is constructed for simulating childhood obesity for the 3-5 years old population. For this model a nonstandard scheme based on the techniques developed by Ronald Mickens is constructed. This model is simulated with real data and the results show an increasing trend of obesity for the next years. The second model is an age-structured model developed in order to study the influence of age stages in the obesity population dynamics. This model considers overweight and obese in the groups 6-8 and 9-12 years old. Based on the numerical simulations of different scenarios it is shown that the prevention of children obesity in early years is of paramount importance. Therefore public health strategies should be designed as soon as possible to reduce the worldwide social obesity epidemic. The last model considers seasonal fluctuations of obesity prevalence using a nonautonomuos system of nonlinear of ordinary differential equations and we show that their solutions are periodic using a Jean Mawhin's Theorem of Coincidence. To corroborate the analytical results and perform numerical simulations, multistage Adomian and differential transformation methods are implemented. Numerical solutions using these methods are compared with those produced using Runge-Kutta type schemes. These implemented methods ensure good approximations using larger step sizes.
