The association between well-being and a large variation of accelerometer-assessed physical activity and sedentary behavior measures

Higher well-being has been associated with more physical activity (PA) and less sedentary behavior (SB), both when assessed by self-report or accelerometers. Most studies using accelerometer data only examined estimates of total volume or daily average of PA/SB in relation to well-being. Taking into account the richness of accelerometer data, we investigated the association of different measures of SB, light PA (LPA) and moderate-to-vigorous PA (MVPA) and well-being including the combined effect and the PA/SB timing and patterns. We explored whether results differed between occupational and non-occupational time. In an adult sample (n = 660, M_age: 30.4, SD = 8.1, 74.5% female), we applied pre-registered analyses. First, we created different global scores of SB, LPA and MVPA based on 4 to 7-days of Actigraph data and investigated associations with well-being, i.e., defined as life satisfaction. These analyses were done using raw scores and transformed scores using compositional data analysis. Next, we applied multilevel models including time of the day and well-being as predictors of PA/SB. Finally, we clustered participants based on PA/SB intensity, timing and accumulation and explored differences in well-being across clusters. In total wear time, there were no associations between different measures of SB/LPA/MVPA and well-being. Restricting to non-occupational wear time, less total SB and more total LPA were associated with higher well-being, both in absolute and relative sense. Well-being was not associated with the PA/SB timing or patterns. In conclusion, beyond the association between total non-occupational SB and LPA and well-being, the PA/SB timing or patterns had no added value in explaining the association between PA/SB and well-being.