Abstract

Rising chronic conditions and healthcare costs demand smarter, personalised approaches to preventive care. Traditional predictive models rely heavily on clinical or demographic data, overlooking the psychological drivers of health behaviour. This study introduces the Motivation-Aware Predictive Modelling Framework (MPMF), a novel system that blends Self-Determination Theory (SDT) with machine learning to decode motivation from wearable-derived behavioural data. Behavioural clustering revealed three distinct motivational profiles - amotivation, controlled, and autonomous - which were integrated with clinical and demographic features to predict healthcare claims. Proof-of-concept models achieved strong sensitivity (TPR 8.4%) and highlighted motivational features as key drivers, offering both predictive power and interpretive insight. By uncovering latent motivational signals, the MPMF enables personalised, scalable interventions that go beyond conventional analytics, transforming how health systems identify at-risk individuals and tailor support. This work demonstrates the promise of motivation-informed predictive modelling as a bridge between behavioural science and cutting-edge health technology.

Download

Share

COinS