Bidirectional Associations of Recreational Sedentary Screen Time and 24-hour Behaviors: a dynamic cross-sectional multilevel model analysis

Background Recreational sedentary screen time (rSST) is the most prevalent discretionary sedentary behavior and is strongly linked to poor health outcomes, but how time spent in rSST relates to other 24-hr behaviors is not well understood. The purpose of this study was to examine within and between day associations between rSST and other 24-hr behaviors– non-rSST or other sedentary time (other-SED), standing (STAND), light-to-moderate-vigorous physical activity (LPA), moderate-to-vigorous physical activity (MVPA), and total sleep (SLEEP). Methods Baseline data from participants randomized in the StandUPTV study, an intervention that aimed to reduce rSST in adults, were included. All 24-hour behaviors were assessed continuously for 7-days. The activPAL was used to assess rSST, other-SED, STAND, LPA, and MPVA; SLEEP was assessed using a GENEactiv accelerometer. rSST was collected using WiFi plugs to capture TV time and tablet app usage. A multilevel modelling approach was used to assess between(across persons)- and within(across days)-person level associations bidirectionally between rSST (total, daytime, evening) and 24-hour behaviors, and adjusted for age, sex, chronotype, education level, and week vs. weekend day. The results were scaled hourly for interpretation. Results On average, 8.0 ± 1.6 days of continuous daily 24-hour behavior data were included from 94 participants (aged 23–64 years [ M  ±  SD : 42.3 ± 11.5]; 82% female; 78% White; BMI 20.5–67.5 kg/m ² [ M  ±  SD : 29.8 ± 7.8]). Greater total rSST was significantly associated with less other-SED (between-person b=-45.0, SE = 4.4, p < 0.01; within-person b=-44.5, SE = 2.0, p < 0.01). Similar results were observed when examining both daytime and evening rSST with other-SED. Negative associations were also observed between other-SED, STAND, LPA, and MVPA with rSST variables. No significant associations were observed between rSST variables and SLEEP. Conclusions This is the first known analysis of the bidirectional relationship between rSST and 24-hour behaviors. More rSST was associated with less other-SED, suggesting rSST may displace rather than contribute to more cumulative sedentary time. These findings suggest that contexts of sedentary behavior should be considered as distinct behavioral targets in intervention development. Future interventions that target rSST reduction should also include strategies that are designed to reduce total sedentary time.