Abstract
OBJECTIVES: It is unclear whether sedentary behaviour, and the domain in which it occurs, is related to body mass index (BMI) change. We aim to elucidate whether sedentary behaviour is prospectively related to BMI change using markers from three domains (leisure, work and commuting). METHODS: Among employed 1958 British birth cohort members (n = 6,562), we analysed whether TV-viewing, work sitting (six categories: 0 h/d to >4 h/d) and motorised commuting (at 45 y) were related to BMI (at 45 y and 50 y) and BMI change 45-50 y, after adjusting for lifestyle and socioeconomic factors. RESULTS: Per category higher TV-viewing, 45 y and 50 y BMI were higher by 0.69 kg/m(2) (95% CI: 0.59,0.80) and 0.75 kg/m(2) (0.64,0.86) respectively. A category higher TV-viewing was associated with 0.11 kg/m(2) (0.06,0.17) increased BMI 45-50 y, attenuating to 0.06 kg/m(2) (0.01,0.12) after adjustment. There was no trend for work sitting with 45 y or 50 y BMI, nor, after adjustment, for BMI change. However, those sitting 2-3 h/d had greater BMI gain by 0.33 kg/m(2) (0.10,0.56) compared to those sitting 0-1 h/d. Associations between TV-viewing and BMI change were independent of work sitting. Motorised commuting was associated with 45 y, but not 50 y BMI or change. CONCLUSIONS: TV-viewing is associated with BMI gain in mid-adulthood; evidence is weaker for other sedentary behaviours.