The rapid growth of chronic diseases and health conditions (e.g. obesity, depression, diabetes, respiratory and musculoskeletal diseases) in many OECD countries has become one of the most significant well-being problems, which pressure the sustainability of healthcare and economies. Thus, promoting early diagnosis, intervention, and healthier lifestyles is essential. One partial solution is extending long-term health monitoring from hospitals to natural living environments. It has been shown in laboratory settings and experimental trials that sensor data, such as camera images, radio samples, acoustics signals, infrared, etc., can be used to accurately model activity patterns related to different health conditions. However, due to the rising concern related to private data leaks and, consequently, stricter personal data regulations, the growth of pervasive residential sensing for healthcare applications has been slow. The topic in this talk aims to combine pervasive sensing technology with secured and privacy-preserving distributed frameworks for healthcare applications to mitigate public concerns and meet regulatory requirements. The talk will elaborate on the measures and collaborative framework for preserving and noising personal information, also decoupling the data/model and identification. The basic idea is to leverage local differential privacy federated learning (LDP-FL) to achieve resilience against active and passive attacks and edge computing to avoid transmitting sensitive data over networks. A machine learning architecture explores combinations of sensor data modalities and security architectures for finding the most viable technology combinations, relying on metrics that allow balancing between computational cost and accuracy for a desired level of privacy.