Like previously17 for the purposes of this study standing will be considered a measure of sedentary behaviour. Physical activity questions included frequency (number of days in the past 4 weeks) and duration (min/day) of participation in walking for any purpose, domestic physical activity,12 23 and any recreational sports and likewise exercise including cycling for any purpose.24 Both the physical activity and the ST questions have been validated against accelerometry.25 Data handling Regrouping the SEP variables Owing to small numbers of observations, the top and bottom two categories of social class were

collapsed, resulting in four categories: unskilled/semiskilled manual; skilled manual; skilled non-manual; and managerial/technical/professional. Using existing methods,12 we derived a composite SEP score using household income, individual education and occupational social class of the head of household. The lowest category of each component variable was assigned a SEP score of 0, with the highest category given a SEP score of 4. The scores for each individual SEP indicator were then aggregated, resulting in a SEP score ranging from 0 to 12. Owing to small numbers of observations

in the high end of the score, the top SEP score was collapsed into five categories of comparable sample size: SEP1 consisted of the lowest two SEP categories (0 and 1), SEP2 comprised categories 2 and 3, SEP3 comprised categories 4 and 5, SEP4 comprised categories 6 and 7 and SEP5 comprised of categories 8 and 9 (the highest observed SEP category). Deriving ST and physical activity variables Weekday and weekend day-specific TV and non-TV leisure time sitting were converted to all-week time (minutes) using the following formula:

(weekday time×5)+(weekend day time×2)/7. Occupational sitting/standing time (minutes) per day was calculated by multiplying the number of days worked per week by the average time spent sitting/standing at work on a work day, and dividing by 7. Weekly self-reported moderate to vigorous physical activity (MVPA) hours/week GSK-3 were calculated as number of days of participation multiplied by time per day in each activity type7 8 Owing to the large number of participants and the very skewed distribution, self-reported MVPA was categorised into none, less than 30 min, 30 min–1 h, 1–2 h, and more than 2 h of MVPA per day. For the accelerometry data we used 0–99 counts/min to denote sedentary (<1.5 MET)3 and ≥2020 counts/min to denote MVPA (>3 MET).26 Accelerometry-measured variables were converted to time (minutes) per valid day and daily ST time was calculated as the sum of the average ST minutes per valid day divided by the number of valid days.