Sustainability in health services has become a crucial issue in the 21st century, given this sector’s environmental, social, and economic impact globally. Sustainable health policies aim to balance providing high-quality medical services with reducing environmental impact, promoting social equity, and achieving economic efficiency. One key aspect in defining sustainable health services policies is providing specialized care to people, especially vulnerable groups and elderly adults, when symptoms may indicate a decline in their health. Traveling to medical centers demands time and effort. Monitoring using smart devices and virtual tools, such as telemedicine, enables medical specialists to provide advice and treatments for managing diseases. One health problem that has increased recently is related to gait or walking disorders. This paper presents a methodology for conducting ten gait tests to classify data collected using four mobile device sensors: acceleration, user acceleration, gyroscope, and magnetometer. Machine learning techniques were applied to a training dataset. Testing on a separate dataset achieved high accuracy in detecting gait alterations. With this approach, gait tests can be performed at patients’ homes to detect gait alterations early and monitor the progression of diseases.