Evaluating the effectiveness of organisational-level strategies with or without an activity tracker to reduce office workers' sitting time: A cluster-randomised trial

© 2016 The Author(s). Background: Office workers engage in high levels of sitting time. Effective, context-specific, and scalable strategies are needed to support widespread sitting reduction. This study aimed to evaluate organisational-support strategies alone or in combination with an activity tracker to reduce sitting in office workers. Methods: From one organisation, 153 desk-based office workers were cluster-randomised (by team) to organisational support only (e.g., manager support, emails; 'Group ORG', 9 teams, 87 participants), or organisational support plus LUMOback activity tracker ('Group ORG + Tracker', 9 teams, 66 participants). The waist-worn tracker provided real-time feedback and prompts on sitting and posture. ActivPAL3 monitors were used to ascertain primary outcomes (sitting time during work- and overall hours) and other activity outcomes: prolonged sitting time (=30 min bouts), time between sitting bouts, standing time, stepping time, and number of steps. Health and work outcomes were assessed by questionnaire. Changes within each group (three- and 12 months) and differences between groups were analysed by linear mixed models. Missing data were multiply imputed. Results: At baseline, participants (46 % women, 23-58 years) spent (mean ± SD) 74.3 ± 9.7 % of their workday sitting, 17.5 ± 8.3 % standing and 8.1 ± 2.7 % stepping. Significant (p < 0.05) reductions in sitting time (both work and overall) were observed within both groups, but only at 12 months.For secondary activity outcomes, Group ORG significantly improved in work prolonged sitting, time between sitting bouts and standing time, and overall prolonged sitting time (12 months), and in overall standing time (three- and 12 months); while Group ORG + Tracker, significantly improved in work prolonged sitting, standing, stepping and overall standing time (12 months). Adjusted for confounders, the only significant between-group differences were a greater stepping time and step count for Group