Sedentary behavior type matters: compositional analysis of 24-hour movement behaviors and cognitive function in older adults

BACKGROUND: Sleep, sedentary behaviors (SB) and physical activity are independently associated with cognitive function in older adults, yet the joint relationship of these 24-hour movement behaviors with cognitive function is less well studied. Additionally, the association between SB and cognitive function may differ depending on whether SB is mentally active or inactive. This study aimed to examine the associations between 24-hour movement behavior compositions and cognitive function in older adults, explore the differential associations of different sedentary behavior (SB) types with cognitive function, and assess the predicted cognitive differences associated with time reallocation among these behaviors. METHODS: Data were drawn from 2516 US adults aged ≥ 60 years in the 2011-2014 National Health and Nutrition Examination Survey. Sleep, total SB, light physical activity (LPA), and moderate-to-vigorous physical activity (MVPA) were all assessed via wrist-worn accelerometry. Total SB was disaggregated into TV watching (inactive SB), computer use (active SB), and other SB using individual proportional weights derived from the Global Physical Activity Questionnaire (GPAQ). Cognitive function was evaluated using the Consortium to Establish a Registry for Alzheimer's Disease Word Learning (memory), Animal Fluency (language), and Digit Symbol Substitution (executive function) tests. Standardized scores were combined into a global cognition score. Weighted compositional linear regression and isotemporal substitution models were applied. RESULTS: In the 4-component model, greater relative time in MVPA and total SB were positively associated with global cognition (β = 0.191 and β = 0.260, respectively; both P ≤ 0.001), whereas relative sleep and LPA were negatively associated with global cognition. In the 6-component model, disaggregating total SB revealed divergent associations: computer use was positively associated with all cognitive domains (global: β = 0.045, P < 0.001), while TV watching was negatively associated with global cognition (β=-0.050, P = 0.047) and language. Notably, neither "other SB" nor sleep remained significantly associated with cognition in this expanded model. Isotemporal substitution analyses indicated that reallocating time to MVPA or computer use was associated with predicted higher global cognition scores, revealing a pronounced asymmetric trajectory where reducing these behaviors predicted steep cognitive declines. CONCLUSIONS: Engaging in MVPA or mentally active SB may help preserve cognitive function in older adults, whereas mentally inactive SB is associated with poorer cognition. Optimizing 24-hour time use by replacing passive sedentary behaviors with MVPA or cognitively engaging activities represents a promising strategy.