Nowadays, the world is facing two major issues: Non-Communicable Diseases and ageing population. Although committing in healthy behaviors has been shown to be highly beneficial for individual's health and well-being, the challenge remains in motivating the adoption and the long-term engagement in such behaviors. This thesis focuses on the efficiency of self-monitoring technologies to promote positive change in the long-term on modifiable behaviors, mainly regarding physical activity and nutrition. It sheds light on the opportunities and the limitations of self-monitoring, gamified, social and conversational applications and intends to provide guidelines for designing these technologies for specific population, namely: chronically ill and elderly patients. Overall, the work conducted within this dissertation offers new perspectives on the design of self-monitoring technologies for elderly and chronically-ill patients. It makes several research contributions that are of interest to the greater Human Computer Interaction, Digital Health, Conversational AI, Behavioral sciences, Nutrition and Physical Rehabilitation communities.