A new study in npj Parkinson’s Disease describes a machine learning approach for identifying individuals with Parkinson’s disease who are at heightened risk of falling. The work presents a methodology designed to classify fall risk rather than merely tracking clinical progression, aiming to support earlier intervention and better patient safety. As digital biomarkers and risk prediction models become more common in neurodegeneration research, the study adds evidence that falling risk may be computable from structured inputs—though validation across cohorts remains critical for clinical translation.