P-242: Comparing complex systems approaches to physical resilience measurement in older adults

摘要

Introduction: In geriatric medicine, the concept of physical resilience—the dynamic ability to respond to and recover from health stressors—may help objectively quantify the individual's recovery capacity. However, it is unclear whether resilience indicators derived from different theoretical approaches provide unique or complementary information. This study compares resilience indicators derived from two prominent conceptualizations: Critical Slowing Down and Loss of Complexity. Methods: Participants were 121 geriatric patients (mean age = 84 years; 60% female) admitted to Radboud university medical center. During hospitalization, physical activity was continuously monitored with a chest-worn accelerometer. Critical slowing down (variance, temporal autocorrelation) and multiscale entropy (MSE) indicators were estimated and compared to patients' physical functioning and well-being. Indicators for active and rest periods were estimated separately to examine whether the associations were context dependent. Results: There were weak-to-moderate correlations among the two types of resilience indicators (range: ρ = 0.19-0.60, p < 0.05). In regression analyses, the strongest associations during rest were found between MSE and frailty index score (β = - 0.314, p = 0.003) and activities of daily living (ADL) limitations (β = - 0.302, p = 0.008). During active hours, variance during activity showed the strongest association with ADL (β = - 0.481, p <0.001) and frailty (β = - 0.368, p < 0.001). Conclusions: Indicators for critical slowing down (variance) and loss of complexity (MSE) in activity time-series were related to physical functioning. Activity variance and complexity showed the strongest associations during active and rest periods, respectively, suggesting complementary information. Considering the underlying theory, the inverse association with variance, while expected intuitively, warrants further exploration.