ABSTRACT
The purpose of the present study was to examine the associations of replacing accelerometer-measured sedentary behavior (SB) with physical activity (PA) and depression in a sample of Japanese frail older adults using an Isotemporal Substitution (IS) model. Among 139 frail older adults, accelerometer was utilized to identify the daily average time spent in SB, light-intensity PA (LPA), and moderate- to vigorous-intensity PA (MVPA). Depression was assessed using the Geriatric Depression Scale-Short Version-Japanese (GDS-S-J). The relationship between SB, LPA, and MVPA with depression was examined with three models of multiple regression analysis: single factor model, partition model, and IS model. As results, LPA times was negatively associated with GDS-S-J scores in the single factor model (B = -0.09, p = 0.007) and the partition model (B = -0.13, p < 0.001). In the IS model, replacement of 10 minutes of SB with an equal amount of LPA time showed a significant negative association with GDS-S-J scores (B = -0.09, p=0.003). These results suggest that replacing a small amount of SB with LPA may be associated with an improvement in depression in frail older adults.