Comparison of objectively measured and self-reported time spent sitting

Abstract

Until recently, methods for objective quantification of sitting time have been lacking. The aim of this study was to validate self-reported measures against objectively measured total sitting time and longest continuous time with uninterrupted sitting during working hours, leisure time on workdays and leisuredays. Objective diurnal measurement of sitting time was obtained among 26 office workers with 2 accelerometers (ActiGraph GT3X+) for a 7-day period. Customized software (Acti4) was used to identify sitting time separated from other sedentary behaviours. Self-reported sitting time was obtained from a retrospective 7-day questionnaire. A generalized linear model showed the difference between the methods. No significant correlations were found between objective and self-reported sitting time (r<0.315). Total sitting time was significantly underestimated (2.4 h) on a leisureday (p<0.001) and uninterrupted sitting time was in all 3 time settings significantly overestimated (0.4-0.5 h) (p<0.045). Poor agreement (mean difference between 0.5 to -2.4 h) between objectively measured and self-reported sitting time was shown in Bland-Altman plots with wide (3.3-10.8 h) limits of agreement. This study showed a great individual variation and a general lack of agreement between self-reported vs. objectively measured total and uninterrupted sitting time. Objective measures are recommended for determining sitting time.