Background: Little is known about associations of temporal patterns of sitting (i.e., distribution of sitting across time) with obesity. We aimed investigating the association between temporal patterns of sitting (long, moderate and brief uninterrupted bouts) and obesity indicators (body mass index (BMI), waist circumference and fat percentage), independently from moderate-vigorous physical activity (MVPA) and total sitting time among blue-collar workers. 

Methods: Workers (n = 205) wore Actigraph GT3X+ accelerometers on the thigh and trunk for 1-4 working days. Using the validated Acti4 software, the total sitting time and time spent sitting in brief (≤5 mins), moderate (>5 and ≤30 mins), and long (>30mins) bouts on working days were determined for the whole day, and for leisure and work separately. BMI (kg/m2), waist circumference (cm) and fat percentage were objectively measured. 

Results: Results of linear regression analysis adjusted for multiple confounders indicated that brief bouts of sitting was negatively associated with obesity for the whole day (BMI, P < 0.01; fat percentage, P < 0.01; waist circumference, P < 0.01) and work (BMI, P < 0.01; fat percentage, P < 0.01; waist circumference, P < 0.01), but not for leisure. Sitting time in long bouts was positively associated with obesity indicators for the whole day (waist circumference, P= 0.05) and work (waist circumference, P =0.01; BMI, P = 0.04), but not leisure. 

Conclusions: For the whole day as well as for work, brief bouts and long bouts of sitting showed opposite associations with obesity even after adjusting for MVPA and total sitting time, while sitting during leisure did not show these associations. Thus, the temporal distribution of sitting seems to influence the relationship between sitting and obesity.