Temporal trends in non-occupational sedentary behaviours from Australian Time Use Surveys 1992, 1997 and 2006

BACKGROUND Current epidemiological data highlight the potential detrimental associations between sedentary behaviours and health outcomes, yet little is known about temporal trends in adult sedentary time. This study used time use data to examine population trends in sedentary behaviours in non-occupational domains and more specifically during leisure time. METHODS We conducted secondary analysis of population representative data from the Australian Time Use Surveys 1992, 1997 and 2006 involving respondents aged 20 years and over with completed time use diaries for two days. Weighted samples for each survey year were: n = 5851 (1992), n = 6419 (1997) and n = 5505 (2006). We recoded all primary activities by domain (sleep, occupational, transport, leisure, household, education) and intensity (sedentary, light, moderate). Adjusted multiple linear regressions tested for differences in time spent in non-occupational sedentary behaviours in 1992 and 1997 with 2006 as the reference year. RESULTS Total non-occupational sedentary time was slightly lower in 1997 than in 2006 (mean = 894 min/2d and 906 min/2d, respectively; B = -11.2; 95%CI: -21.5, -0.9). Compared with 2006, less time was spent in 1997 in sedentary transport (B-6.7; 95%CI: -10.4, -3.0) and sedentary education (B = -6.3; 95%CI: -10.5, -2.2) while household and leisure sedentary time remained stable. Time engaged in different types of leisure-time sedentary activities changed between 1997 and 2006: leisure-time computer use increased (B = -26.7; 95%CI: -29.5, -23.8), while other leisure-time sedentary behaviours (e.g., reading, listening to music, hobbies and crafts) showed small concurrent reductions. In 1992, leisure screen time was lower than in 2006: TV-viewing (B = -24.2; 95%CI: -31.2, -17.2), computer use (B = -35.3; 95%CI: -37.7, -32.8). In 2006, 90 % of leisure time was spent sedentary, of which 53 % was screen time. CONCLUSIONS Non-occupational sedentary time has increased slightly from 1997 to 2006 in the Australian adult population. This seems to be the result of small increases in sedentary transport and education time while sedentary household and leisure time were stable over this time period. However, almost all leisure time is spent sedentary and the composition of sedentary leisure time changed between 1992 and 2006 towards a larger proportion being screen-based activities. This could be an important observation for public health, as most of the evidence on the detrimental effects of sedentary behaviour is around watching television and health.