Complexity analysis of physiological signals in the wearable era: New insights into chronic diseases and aging

The human body is a complex and irregular system that generates diverse physiological signals. In the era of wearable technology, an increasing number of physiological signals can be continuously monitored, generating rich time series data that capture the dynamic physiological fluctuations and may reveal subtle signs of dysregulation. For example, continuous glucose monitoring (CGM) systems generate glucose time series data. As such, new metrics harnessing the time series nature of physiological signals may serve as sensitive markers of health. Of them, complexity focuses more on the dynamical characteristics of physiological signals and can reveal hidden information that linear methods may miss, allowing for the handling of nonstationary data and providing a more precise capture of individual differences. Importantly, the complexity of physiological signals may reflect the regulatory capacity, adaptability, and functional reserves of the body, which are often impaired in chronic diseases and during aging. Using CGM-derived glucose complexity as a representative example, this review discusses the clinical significance and applications of complexity in chronic diseases and aging. Finally, we outline future directions for complexity-related research, particularly regarding the roles of artificial intelligence (AI), wearable devices, and clinical translation.