Intervening to reduce workplace sitting: mediating role of social-cognitive constructs during a cluster randomised controlled trial

Nyssa T. Hadgraft, Elisabeth A. H. Winkler, Genevieve N. Healy, Brigid M. Lynch, Maike Neuhaus, Elizabeth G. Eakin, David W. Dunstan, Neville Owen, Brianna S. Fjeldsoe

The Stand Up Victoria multi-component intervention successfully reduced workplace sitting time in both the short (three months) and long (12 months) term. To further understand how this intervention worked, we aimed to assess the impact of the intervention on four social-cognitive constructs, and examined whether these constructs mediated intervention effects on workplace sitting time at 3 and 12 months post-baseline. Two hundred and thirty one office-based workers (14 worksites, single government employer) were randomised to intervention or control conditions by worksite. The intervention comprised organisational, environmental, and individual level elements. Participant characteristics and social-cognitive constructs (perceived behavioural control, barrier self-efficacy, perceived organisational norms and knowledge) were measured through a self-administered online survey at baseline, 3 months and 12 months. Workplace sitting time (min/8 h day) was measured with the activPAL3 device. Single multi-level mediation models were performed for each construct at both time points. There were significant intervention effects at 3 months on perceived behavioural control, barrier self-efficacy and perceived organisational norms. Effects on perceived organisational norms were not significant at 12 months. Perceived behavioural control significantly mediated intervention effects at 3 months, accounting for a small portion of the total effect (indirect effect: -8.6 min/8 h day, 95% CI: -18.5, -3.6 min; 7.5% of total effect). At 12 months, barrier self-efficacy significantly mediated the intervention effects on workplace sitting time (indirect effect: -10.3 min/8 h day, 95% CI: -27.3, -2.2; 13.9% of total effect). No significant effects were observed for knowledge at either time point. Strategies that aim to increase workers' perceived control and self-efficacy over their sitting time may be helpful components of sedentary behaviour interventions in the workplace. However, social-cognitive factors only partially explain variation in workplace sitting reduction. Understanding the importance of other levels of influence (particularly interpersonal and environmental) for initiating and maintaining workplace sedentary behaviour change will be informative for intervention development and refinement. This study was prospectively registered with the Australian New Zealand Clinical Trials register ( ACTRN12611000742976 ) on 15 July 2011.