Epidemiology of Sedentary Behaviour: Novel Findings in Health and Measurement

The overall aim of this PhD was to further examine the associations between physical activity, sedentary behaviour and health (cardiometabolic health, all‐cause mortality, cognitive function), and explore novel approaches for analysing physical activity and sedentary behaviour data. Key Findings: •In a national survey sample of adults (Health Survey for England), being physically active was associated with better cardiometabolic health, even in those with high sedentary time. •In a regional sample of high risk of T2DM adults (Walking Away from Type 2 Diabetes), MVPA time was associated with a lower risk of mortality. Conversely, sedentary time showed no association with mortality. •In a large sample of UK adults (UK Biobank), TV viewing and driving time were inversely associated with cognition. Conversely, computer use time was positively associated with cognition. Further analyses demonstrated that fitness did not modify these associations, and that the number of healthy lifestyle factors was positively associated with cognition. •Sedentary behaviours can be separated from light activities (except standing still) using intensity‐based thresholds derived on experimental raw acceleration data. In conclusion, this project has helped fill several epidemiological gaps in knowledge via exploiting multifaceted databases, and evaluated innovative measurement techniques. The observational analyses demonstrated the importance of physical activity as a determinant of cardiometabolic health and mortality, but found the role of sedentary behaviour to be relatively equivocal. Additional work with cognitive outcomes showed that some sedentary behaviours, but not all, are associated with poor cognition. These results provide robust data supporting public health policies designed to reduce TV viewing and driving time in adults, and increase healthy behaviours for cognitive wellbeing. Intervention studies are required to confirm these findings. The experimental analyses showed that researchers can accurately separate sedentary behaviours from light activities using thresholds on raw acceleration data; thus, providing a useful resource for future studies