With the technological advances, the treatment and follow-up of diseases such as Diabetes have now become more comfortables for the physician and the patient, being possible to control glycemic rate information and follow the history of the evolution of the treatment of the disease by various systems available. However, the human-computer interfaces developed for these systems are often projected on the assumption that they will be used by healthy people, who have normal perceptual and cognitive abilities, and who are in a comfortable and stable environment. Therefore, any deviation from these "standards" may hinder the use of these systems. Another complicating factor is the fact that, in general, users are not very familiar with the technologies, which makes it difficult to understand and navigate these systems. Therefore, the creation of an adaptive interface becomes feasible to soften this problem. In view of the above, the present study aimed to implement an adaptive interface for MobiLEHealth, allowing it to adapt dynamically to the needs of users with chronic disease, specifically users with Diabetes who use the system, focusing on their difficulties/limitations. For this, a mapping of the characteristics/difficulties associated to these patients was carried out. The results of this mapping were represented in an ontology. Based on the ontology created on the characteristics/difficulties of users with Diabetes and the use of a multi-agent system, the developed solution allows to automatically choose and apply adaptation actions in the interface, in order to meet the needs of the users. For validation, a test script was applied with specialists related to the health area at the interface. The information was obtained through questionnaires developed following TAM methodology. The results show that professionals indicate that the adaptive interface has the potential to achieve the goals that were proposed, not only for patients with Diabetes, but also for people with some type of difficulty