Background
In recent years, a new research paradigm has emerged, highlighting the potential deleterious health impacts of sustained sedentary behaviour [
1], independent of the amount of physical activity undertaken [
2,
3]. Potential health consequences include an increased risk of: cardiovascular disease [
4‐
6]; metabolic syndrome/type 2 diabetes [
4,
6‐
9]; obesity [
10]; hypertension [
11]; some cancers [
12]; depression [
13]; and musculoskeletal problems [
14]. Furthermore, sustained sedentary behaviour has shown to be independently associated with an increased risk of premature mortality [
6,
9,
15]. As a result of these associated health impacts, reducing the amount of time adults spend being sedentary has been identified as an important public health initiative [
16].
Sedentary behaviour is defined as any waking behaviour characterised by an energy expenditure ≤1.5 Metabolic Equivalents (METs) while in a sitting or reclining posture [
17], and is distinct from light intensity activities such as standing and walking, which have been shown to confer some health benefits [
18,
19]. As it is impractical to measure energy expenditure in the majority of studies, sedentary behaviour can be most simply operationalised as sitting time [
20].
Prolonged workplace sitting is especially significant when considering interventions to reduce sitting time due to the increase in desk-based jobs over recent decades [
21] and the fact adults spend approximately 60 % of their waking hours in the workplace [
22,
23]. Furthermore, recent evidence has shown that office workers sit for an average of six hours in an eight-hour working day and sitting is most often accumulated through prolonged, unbroken bouts [
24,
25]. Observational studies looking at office work and sitting time have shown a positive association, which significantly contributes to overall daily sitting time [
24,
26]. It is this recognition of the importance of sitting in the workplace which has highlighted the need for the development of interventions to reduce workplace sitting [
27].
This paper describes the development and mixed methods evaluation of an intervention to reduce workplace sitting time, drawn from a four week uncontrolled study, which is intended to be a precursor to a larger controlled trial. The main aim of this study was to explore intervention acceptability and feasibility, which was addressed by documenting: pre- and post-intervention changes in daily sitting time; participants’ awareness of the various elements of the intervention; and participants’ views of the intervention. It also provides initial evidence on effectiveness of low-cost interventions, which could be used to inform future workplace studies.
Discussion
This uncontrolled trial explored the acceptability and feasibility of a low-cost, co-produced, multi-modal intervention in reducing daily workplace sitting time amongst university desk-based workers. Three sources of information were used: pre-post behavioural responses; awareness of the various elements of the intervention; and views on the intervention.
Baseline sitting time of 440 min/day amongst study participants was higher than sitting time data for professional/managerial staff obtained in a previous study by Miller et al. [
55] which looked at workplace sitting time by occupation. This could be explained by the differences in data collection: this present study used a sitting log, which was completed morning and afternoon by participants; whereas Miller et al. used the IPAQ. During this pilot, 82 % of “completers” reported a reduction in daily workplace sitting time, with a mean decrease of 26 min/workday (6 % reduction). Evaluating the effectiveness of this intervention was not an aim of this study, so it was not sufficiently powered to detect a statistically significant effect size (unless the effect size turned out to be very large); hence the quantitative findings need to be interpreted with caution. Nevertheless, the lack of statistical significance should not rule out potentially important benefits of the intervention. “Completers” were of comparable demographics to participants of similar studies [
41‐
43,
56,
57] i.e. mainly White British females aged 35–44 years with a similar educational attainment and baseline level of physical activity.
On the whole, studies that evaluated multi-modal interventions that demonstrated statistically significant changes, reported a 10-25 % reduction in workplace sitting [
41‐
43,
57]. The reasons for the reduction in sitting time in this study being lower than that of other multi-modal interventions may be explained by the lack of a true environmental element to this intervention such as sit-to-stand desks [
41,
43,
57] or portable pedal machines [
42]. Due to the need to develop low-cost interventions to support uptake amongst a variety of organisations, it was not considered appropriate to include such elements, which instead resulted in the development of a simple and pragmatic intervention.
The studies conducted by Neuhaus et al. [
41] and Carr et al. [
42] were based within a university setting similar to this pilot. A recent meta-analysis [
58] reported 76 % of studies targeting improvements in health behaviours, such as physical activity, conducted in tertiary educational settings demonstrated significant health improvements and suggested that these settings should serve as a platform to research such intervention strategies. Nevertheless, findings from studies carried out in these settings may not be generalisable, as participants are not representative of the wider working population due to key demographic differences such as educational attainment and socio-economic status. Despite this, a study by Matei et al. [
59], which looked at an intervention to reduce sitting time and increase physical activity amongst two different groups of older people, found different levels of uptake and impact amongst the different groups. Furthermore, a qualitative study by Bardus et al. [
60], which looked at the reasons for participating and not participating in an e-health workplace physical activity intervention, demonstrated the importance of focusing on employees’ needs and motivators to behaviour change. These two studies highlight the importance of developing interventions that are tailored to the specific needs of that particular population and setting, rather than using a generic “one-size-fits-all” approach. Hence, an intervention developed for staff in a university setting for example may be different to an intervention developed for staff in a private sector small-to-medium sized enterprise.
Participants (not involved in the intervention development) were all aware of the intervention, but with some variation in the awareness of the different elements. There was a lack of awareness of some elements of the intervention such as the Twitter™ updates and the presence of workplace champions, which therefore rendered them unhelpful to those participants. The intervention as a whole was generally well-received by participants and was felt to be acceptable and feasible due to the ability of the different elements to be easily incorporated into the working day. It was felt that the intervention had a positive impact on workplace sitting as a result of awareness raising, reminders and support from management. Management support in particular was valued by participants in previous studies that evaluated similar interventions as a means of validating changes to health behaviour [
41,
51]. There was a variety of feedback on the most helpful and least helpful elements, which seemed to be dependent on personal preference and awareness. Some participants felt that standing/walking meetings were helpful and easy to implement, whilst others felt that they were impractical and difficult to instigate. Barriers to reducing workplace sitting, which the intervention did not address, were highlighted and included: the nature of work; the workload/lack of time; attitudes of peers; and the workplace environment. The barriers highlighted by participants informed suggestions for improvements, which as far as possible need to be used to enhance further intervention development, highlighted by Neuhaus et al. [
28] as good practice.
Aside from the reduction in workplace sitting, further benefits that were also described by participants, included: increased productivity, improved workplace culture, decreased stress, increased awareness of the health benefits of sitting less, and improved physical health. There is currently uncertainty in the literature regarding what the clinically important difference in sitting time needs to be in order to positively affect health. In a large population study [
61], it was found that in the most sedentary, every hour/day increase in sitting time was associated with a 1.4 cm increase in waist circumference. Given that during this pilot daily, sitting time was reduced by an average of 26 min, it is possible that a change of this magnitude, if sustained, could result in positive health effects. However, determining health-related benefits was not an aim of this study and hence quantifiable data were not collected. Some studies have begun to evaluate whether multi-modal interventions to reduce workplace sitting correspond to improved health outcomes such as: anthropometric measures [
43]; cardiometabolic risk factors [
42,
43]; mood states [
57]; and musculoskeletal symptoms [
41,
57]. However, as a result of the short-term follow-up (4–12 weeks), mixed results have been reported. In addition, work-related outcomes, reported in this pilot as additional benefits, could be used to inform the argument for cost-effectiveness of such interventions. Nevertheless, determining work-related outcomes was not an aim of this study, so quantifiable data were not collected. Two studies evaluating multi-modal interventions [
41,
43], which did include an assessment of work-related outcomes (self-rated work performance, absenteeism, presenteeism), did not observe statistically significant improvements. At present there is a dearth of evidence relating to the impact of such interventions on health- and work-related outcomes. Longer-term evaluations of similar interventions should include an assessment of these outcomes.
Since the completion of this pilot, a review by Gardner at el. [
62] was published, which has highlighted the key elements of promising interventions to reduce workplace sitting. Firstly, the most promising interventions primarily aimed to change sedentary behaviour rather than physical activity. Secondly, this review found that interventions based on functions such as environmental restructuring and education, were most promising in reducing sedentary behaviour in the workplace. Finally, using behaviour change techniques such as self-monitoring of behaviour, adding objects to the environment, instruction on how to perform the behaviour, reviewing behavioural goals, providing information on health consequences, and behaviour substitution improved the promise of such interventions. Furthermore, the greater number of intervention functions and behaviour change techniques used, the more promising the workplace intervention. This pilot has many features of “promising interventions” as described by Gardner et al. including: the reduction in workplace sitting time as the primary aim; education being incorporated as a function of the intervention (e.g. awareness raising); utilising behaviour change techniques such as instruction on how to perform the behaviour (e.g. the use of prompts and reminder software to encourage regular breaks from sitting), providing information on health consequences (e.g. the educational You Tube™ video), and behaviour substitution (e.g. the use of standing/walking meetings). Despite this review by Gardner et al. being based on often low-quality evaluation methods, it has provided a good basis to improve future interventions aimed at reducing workplace sitting time. This pilot could usefully incorporate the findings of this review in any future work.
The main strength of this pilot was the adoption of a systematic and evidence-based intervention development process, incorporating conceptualisation and formative research. The fact that the intervention was grounded within a theoretical model (SEM) allowed the multiple levels of influence to be targeted, thereby ensuring that: individuals' autonomy and knowledge were increased; social networks were developed; and organisational support was obtained. The only element that this intervention unsatisfactorily addressed was the presence of true environmental changes. In addition, the use of a participatory approach has been demonstrated as an effective mechanism to reduce workplace sitting, which ensured a match between the needs of the staff and the suggested strategies [
56]. The participatory approach to intervention development, ensured that the intervention was more likely to be acceptable to (and feasible for) staff. This approach allowed the development of a pragmatic intervention for use in a “real-world” setting. Finally, a further strength was the mixed methods approach, which ensured that the main aim of the study, assessing the feasibility and acceptability of a low-cost intervention, could be demonstrated, and has provided a basis for future research.
The major limitations of this pilot were the subjective nature of the data collection tool and the small number who provided individual data, relative to the number of staff invited to participate in data collection. The measurements of sitting time were based on self-report only, which may have introduced reporting and recall bias, although the prospective nature of the completion of the log may have minimised some of the bias.
Further limitations included:
1.
The convenience sampling techniques used may have introduced selection bias. Participants may have been largely those already aware of the detrimental effects of prolonged workplace sitting or those who felt they might benefit the most from the intervention and hence were more committed to participating. Furthermore, the comparison between the “completers” and “non-completers” shows that, although there were no obvious differences between the two groups, it is unclear how representative the two groups were of the entire workforce. In addition, nothing was known about change in sitting time amongst the “non-completers”, so it is possible that “completers” were those who had a strong view about /were more aware of the intervention, which could have introduced further bias.
2.
The small sample size and the nature of the intervention (such as posters and changes to meetings) made individual randomisation impossible and even cluster randomisation very difficult with likely contamination of any control group, meaning it was not possible to control for confounders. The lack of a control group, small sample size and probable selection bias may mean that the findings of this study are not widely representative of all ScHARR staff and hence are limited to the population studied and setting in which the study was conducted.
3.
The lack of long-term monitoring of the effects of this pilot intervention along with the short intervention duration means that the potential for sustained reductions in workplace sitting have not been determined.
Implications for policy and practice
This study has demonstrated the acceptability of low-cost interventions to reduce workplace sitting. The implications of such interventions on productivity at work requires further research, but the present evidence suggests that there is at least no decline in productivity or other work-related outcomes and it is possible that a reduction in sitting may improve productivity according to the findings from this study. Therefore, UK workplaces could consider pragmatic methods for reducing workplace sitting, which address multiple levels of influence and includes a participatory approach to ensure that the strategy is tailored to the needs of the workplace.
Implications for future research
The findings of this pilot contribute to this emerging research paradigm and could be used to inform the development of a larger, cluster-randomised control trial. This more robust study design would allow for control of confounding and selection bias and has the ability to explore effect modification, whilst also producing a more precise estimate of effects. Furthermore, interventions tested over a greater duration including longer-term follow-up and objective monitoring techniques, to determine accurate and sustainable effects, are required. Any negative implications of such intervention and suggestions for intervention improvements should be incorporated into any further intervention development and evaluation ensuring the iterative nature of such a process. Including an environmental/ergonomic element within such multi-modal interventions is likely to yield more successful results, although will be more costly and hence may be a significant barrier to uptake amongst some employers.
Establishing whether there were differences between “completers”, “non-completers” and those who did not participate at all, may have provided further insight into the findings of this pilot and maximise the generalisability. Future research should ensure that everyone who is to be exposed to an intervention is recruited to participate in the study, not simply a self-selected sample as in this pilot. To support this, providing an incentive in return for participants’ time spent during the data collection process, in the form of either a financial incentive or simply feedback on how their sitting time compares with their peers, could be beneficial. In addition, conducting similar pilots in a variety of different sedentary workplaces (e.g. public sector, private sector, small-to-medium sized enterprises, and larger corporations) may provide a greater understanding of the steps that need to be taken to develop and evaluate successful interventions aiming to reduce workplace sitting, and hence also increase the generalisability of the findings.
This pilot has highlighted the importance of a systematic, evidence-based intervention development process, which should include a pilot study, so further research into interventions to reduce workplace sitting need to ensure a similar process is adopted. Furthermore, there is a requirement for future research to focus not only on whether an intervention successfully reduced workplace sitting, but on the impact such interventions could have on both health- and work-related outcomes. It is these findings which will allow a full assessment of the cost-effectiveness and hence sustainability of such interventions in UK workplaces.
Competing interests
All the authors declare that they have no competing interest.
Authors’ contributions
KM conceived the study idea, designed the intervention trial and procedures, collected and analysed data, and drafted the manuscript. All of this was in fulfilment of the requirements for KM’s Masters in Public Health dissertation project. EG was the main supervisor of this dissertation project providing essential support particularly relating to the developing the aim of the study, the methods used, the type of analysis conducted and the writing-up of the original report. FE supplied some of the evidence-based prompts, that he had been involved in the design and trialling of, which were used as part of the final intervention. FE also provided intellectual knowledge relating to the use of prompts and contributed particularly to the critique of the evidence-base relating to point-of-decision prompts. Both EG and FE contributed to the design of the study, providing important advice and support to KM throughout the study period. EG and FE helped to revise the drafted manuscript by providing important critiques on both the content and the style of the final manuscript. All authors edited the manuscript, and read and approved the final submission. All authors agree to be accountable for all aspects of the work.
KM carried out this trial to fulfil the requirements for a Masters in Public Health dissertation project (completed at ScHARR, University of Sheffield); as a result there was no specific funding for this trial. KM is a Specialty Registrar in Public Health currently planning to apply for a Doctoral Research Fellowship to build on the work presented here.