Elderly people who show more random changes in daily movement tend to be at greater risk of frailty, disability and death, according to a large study involving 1,275 individuals over the course of 13 years. The findings - based on a technique called fractal physiology - imply that analyzing daily motor activity could help predict wellness and health in the elderly population and spot people who might benefit from earlier interventions. Fractals are repeating, self-similar patterns that are often encountered in mathematics and the physical sciences. They are also relevant in biology and medicine; fractals can be used to understand the structure of biological tissues, and changes in fractal patterns in physiological measures (such as movement) can indicate the presence of diseases such as dementia. Peng Li and colleagues investigated whether a fractal physiology approach could be used to gauge long-term health outcomes in older individuals. They enrolled 1,275 elderly subjects (56 to 100 years old) and monitored their daily movement activity using activity monitors worn on the wrist. Using statistical models, the team made a striking observation: participants who showed greater random fluctuations in their motor activity over two time scales (between one and 90 minutes, or greater than two hours) had a higher risk of becoming frail or developing a disability that impaired daily activities over the next 13 years. The subjects with perturbed fractal motor activity also showed higher mortality: those in the lower tenth percentile of "normal" activity - who showed more random fluctuations - had an overall 1.6-fold higher risk of death compared to participants in the 90th percentile. The authors say fractal physiology could become a useful tool for health monitoring and spotting early signs of disability and disease before they become visible.