Perceived Barriers and Facilitators to Breaking Up Sitting Time among Desk-Based Office Workers: A Qualitative Investigation Using the TDF and COM-B

Sedentary behaviour is defined as a waking energy expenditure ≤1.5 metabolic equivalents (METs) while sitting, reclining or lying down [1]. High levels of sedentary behaviour has been identified as a risk factor for a number of cardiometabolic health issues, including cardiovascular disease, type 2 diabetes, and obesity [2,3,4]. The workplace has been identified as a high-risk site for excessive sitting time in many western countries [5,6], with evidence that office workers spend the majority of their working day sitting [7]. Desk-based workers may sit for up to 82% of their working hours [5].

Studies have shown that breaking up long periods of sitting with short bouts of standing, light and moderate-intensity activity can reduce cardiometabolic risk markers including postprandial glucose, insulin, triglycerides and blood pressure [8,9,10,11,12]. Numerous interventions have been designed to break up sitting, including environmental changes to the workplace, such as provision of height-adjustable desks to switch between siting and standing, treadmill desks, and cycling desks [13,14,15], computer prompts [16,17,18,19], policy changes, counselling and provision of information [20]. However, while some of these interventions appear to be effective in reducing workplace sitting time [17,21,22], results from studies have been inconsistent. In addition, those using active workstations provide low quality evidence and there is a dearth of literature on long-term use according to a Cochrane review [23]. For instance, a study investigating the effectiveness of standing ‘hot desks’ on sedentary work time has reported no change in overall sedentary behaviour [24]. Similarly, a Swedish cross-sectional observational study reported that height-adjustable desks only resulted in a 5% decrease in call-centre workers’ total sitting time [15]. Moreover, a study carried out in four different companies reported that 60% of male and female employees who recently received height-adjustable desks stated using the desks only once a month or less [25]. This suggests that access to a height-adjustable desk does not necessarily translate to using it to alternate between sitting and standing [26]. Similarly, the effectiveness of using treadmill desks to increase light physical activity among desk-based office workers was hampered by poor adherence due to drop-out, injury, work conflict and being out of office [27].

A possible reason for the lack of adherence to interventions could be that they were not tailored to meet the needs of the participants. Few studies have evaluated the perception of employees to reducing workplace sitting before developing interventions [28,29]. Cole et al. [30] explored the perceptions of professionals with respect to factors that might hinder or promote sedentary behaviour at work. This was necessary as it allowed the participants to be part of the processes involved in making decisions towards changing behaviour [31]. Future research should use appropriate behaviour change theory to develop tailored interventions that consider specific preferences, work practices and settings [30].

The use of behaviour-change theory and models have been recommended for many contexts including addiction, prevention of disease, and professional practice [32,33]. It is helpful to assess barriers to change prior to the tailoring of interventions [34]. The Behaviour Change Wheel (BCW) is a layered framework designed to enable developers of interventions to move from the analysis of a behavioural problem to intervention design informed by evidence [35]. This framework provides a systematic and transparent way of identifying intervention functions and policy categories that provoke change [35]. The Medical Research Council guidance [36] for developing complex interventions outlines that researchers need to have a theoretical understanding of probable process of change. This can be achieved by drawing on existing evidence and theory, and if necessary, supplemented with qualitative interviews with stakeholders and/or the target population [36]. The COM-B (Capability, Opportunity, Motivation—Behaviour) model is at the hub of the BCW and is used to determine what needs to change. The COM-B model establishes that human behaviour is a result of the interaction that exists between Capability, Opportunity and Motivation [35]. Capability refers to individual’s physical and psychological ability to enact the target behaviour, which usually involves having the required cognitive processing ability, such as attention, memory, knowledge, and skills. Motivation on the other hand refers to the belief systems that strengthen and guide behaviour such as beliefs about capability, intention, and outcome expectancies. It further encompasses habitual processes, and emotional responses. Opportunity points to other factors beyond the individual, both social and physical, which can influence the enablement of the behaviour [35]. Each of the three components can be further subdivided to capture important distinctions within the constructs. For instance, (1) ‘psychological capability’ has to do with the capacity to engage in essential thought processes, while (2) ‘physical capability’ is concerned with skills to execute the behaviour. In terms of opportunity, (3) ‘physical opportunity’ is what is afforded by the environment to enable the behaviour, while (4) ‘social opportunity’ centres around cultural and social influences. With respect to motivation, (5) ‘reflective motivation’ involves thinking with the head, which includes evaluations and intentions while (6) ‘automatic motivation’ involves emotions and habits [35]. This system does not place priority on any of the individual components, however, it provides a way of identifying to what extent changing one or more components could lead to the desired change in behaviour [35]. COM-B has been found to be an effective model in explaining physical activity behaviours [37], however, less is known of its ability to predict sedentary behaviour.

The Theoretical Domains Framework (TDF) [38] was designed to help with the understanding of the concept of behaviour theoretically to effectively target processes for change [32]. The TDF has been used as the basis for identifying factors that predict guideline adherence; for structuring an interview schedule to capture drivers of behaviour using framework analysis [39,40] and for structuring the presentation of data analysis [41,42]. However, the best approach to identifying the most appropriate domains to be targeted for intervention is still unclear [43]. The TDF has 14 domains which are; ‘Knowledge’, ‘Skills’, ‘Social/Professional Role and Identity’, ‘Beliefs about Capabilities’, ‘Optimism’, ‘Beliefs about Consequences’, ‘Reinforcement’, ‘Intentions’, ‘Goals’, ‘Memory, Attention and Decision Processes’, ‘Environmental Context and Resources’, ‘Social Influences’, ‘Emotions’, and ‘Behavioural Regulation’ [38].

More recently, research has begun to use the TDF alongside COM-B. Qualitative studies investigating smoking cessation, stroke rehabilitation, diabetes prevention and antibiotic prescribing behaviour [44,45,46,47,48] have adopted this approach. This enables the expansion of each of the components of COM-B, allowing for more detail by using the TDF alongside [44,45,46,47,48] creating a parsimonious arrangement of probable influencing determinants of behaviour [49].

There is only one known study relating to sedentary behaviour, that has used the combination of the TDF and COM-B to develop an intervention, entitled Stand More AT Work (SMArT). This intervention aimed to reduce workplace sitting time in desk-based National Health Service (NHS) hospital staff [49]. Within three months of intervention, the study found that workplace sitting time had reduced by 50.6 min and by 64.4 min at six months in addition to sustainable long-term improved job performance [50]. With sedentary behaviour research applying both the COM-B model and TDF still at infancy stage, it is difficult to ascertain the generalisability of the current evidence. Therefore, further studies that target employees in different settings with purely administrative roles and high levels of sitting time are warranted [51]. This current study will draw on both the COM-B model and TDF combined to investigate the perceived barriers and facilitators to breaking up and reducing workplace sitting time among desk-based office workers to inform the development of future interventions.

Barriers to breaking up and reducing sitting time at work were attributed to the core themes of: ‘Knowledge-deficit sitting behaviour’, ‘Willingness to change’, ‘Tied to the desk’, ‘Organisational support and interpersonal influences’, ‘Competing motivations’, ‘Emotional influences’, and ‘Inadequate cognitive resources for action’. These themes were linked to five of the six COM-B elements, without strong evidence for Physical Capability barriers. This is likely due to the exclusion of people with musculoskeletal problems in the study design.

The use of the COM-B model [35] was instrumental in identifying an overview of factors contributing to prolonged sitting time in the workplace. Mapping the data further to the TDF [38] ensured an exhaustive behavioural diagnosis of the determinants of the behaviour. Eleven of the TDF domains (Knowledge, Beliefs about capabilities, Intentions, Beliefs about consequences, Environmental context and resources, Social influences, Reinforcement, Social/Professional role and identity, Emotion, Memory, attention and decision processes, Behavioural regulation) were highlighted deductively following the initial inductive thematic analysis, without strong evidence for Skills, Optimism and Goals. This may again be linked to the inclusion criteria (free from musculoskeletal ill health) and the nature of the interviews investigating barriers as opposed to issues that may highlight optimistic views or goal setting.

Participants often did not engage in breaking up sitting behaviour as they lacked the Psychological Capability in the form of knowledge or guidance for sitting at work and stated that it would be important for them to understand why it would be necessary. However, despite not knowing recommendations for sitting time, the participants still believed their sitting pattern was bad and showed positive intentions towards participating in a behaviour change intervention to break up sitting in the workplace. This is consistent with findings from the SMArT study that used focus groups to explore what needed to change to reduce NHS office-based employees’ sitting time based on questions developed from the COM-B model and TDF [49]. In this present study, participants’ knowledge was found to be an important TDF domain.

Both Physical and Social Opportunities were found to be crucial to breaking up and reducing sitting behaviour in the present study. This is consistent with work by Mackenzie et al. [60], in which most of the participants blamed the sedentary nature of their job, tight deadlines, peers’ influence, an unsupportive organisational culture, fear of being judged due to the open plan structure of their office that does not allow for any privacy, and an excessive workload, for their inability to break up sitting time. The participants in the current study felt it would be helpful to be provided with opportunity, which would require their work environment to be restructured, including the use of a height-adjustable desk, on-screen computer prompts, scheduled breaks, and adequate support from the organisational management. These findings are consistent with those identified in focus group discussions among employees and executives to identify strategies that could be used to influence workplace sedentary time in a study by De Cocker and colleagues [28]. However, concerns were raised about the cost implication of a height-adjustable desk and whether organisations would be willing to invest in such a scheme to enhance the Physical Opportunity to sit less. There is evidence that suggests the long-term indirect costs of employees’ presenteeism and absenteeism stemming from sedentary behaviour-associated illness might be higher than the cost of height-adjustable desks. For instance, height-adjustable desks cost ≥ £279 (US $375) per unit for a single display [61] and ≥ £305 (US $400) per unit for a dual display [62], which compares favourably with the indirect cost of presenteeism for businesses (which is far more than the cost of absenteeism) that was approximately £194 (US $255) per employee per annum [63]. Previous work has shown that organisational managers and employers are perceived as gatekeepers that determine the implementation of the recommendations to break up prolonged sitting in the workplace [64], which was also identified in the current study. Similarly, the social opportunity provided by being within a team or culture in which everybody (employees and managers) implemented the plan, was desirable, highlighting the importance of Social Opportunity to facilitate breaking up sitting behaviour.

Both Automatic and Reflective Motivation were also found to be key determinants of participants’ sitting behaviour and their intention to be involved in an intervention that aimed to break up sitting time. In terms of Automatic Motivation, regardless of opportunity, laziness, lack of drive and habitual behaviour such as eating at the desk could hinder employees from breaking up their sitting time. However, this could be mediated by their Reflective Motivation regarding beliefs about the consequences of breaking up sitting time, such as benefits to health. Automatic Motivation had a varied effect; mood, for instance, was found to either increase or decrease participants’ sitting behaviour, and sometimes, appeared to have no effect. However, motivation to increase engagement with breaking up sitting could arise from the introduction of a reward system. This suggests that reinforcement could be an important TDF domain to influence emotions and therefore should be taken into consideration during intervention development.

The findings from the analysis of the interviews in this study suggest that the drivers of behaviour could be either intrinsic or extrinsic. Intrinsic drivers are those that are personal to employees, while extrinsic drivers are determined by the environment and employers. However, this implies that breaking up sitting time will require a more complex intervention using concerted behavioural planning that involves both office workers and their employers, as found in a previous study [49]. It was evident that there is currently no such collaborative scheme, where employees are knowledgeable, ready, willing and able, and feel supported to break up their sitting in the workplace. To trigger a change, there is an urgent need for intervention developers to focus on educating office workers about risks attached to prolonged sitting and what is currently considered as acceptable sitting time durations during a working day. To ensure such a change occurs would require a culture of collaborative behavioural planning and the development of a complex intervention that would need to take Psychological Capability, Social and Physical Opportunity, and Reflective and Automatic Motivation into consideration. However, going by the ‘less is more’ principle of intervention design, only a fraction of target drivers of behaviour should be selected for rigorous intervention testing, instead of targeting several with no focus [65]. Subsequent studies on intervention development involving identification of intervention functions and active ingredients should be mindful of these considerations.

One of the limitations of qualitative studies is concern over generalisability due to a small sample size in comparison with quantitative studies and their subjectivity with regards to researchers’ interpretation. However, this study has provided an in-depth analysis of the barriers to breaking up sitting time at work and potential facilitators that can be targeted in intervention design. A further limitation is that physical activity of the participant was not considered during recruitment. However, as the focus of this study was on workplace sitting time, the time spent sitting during working hours was one of the criteria for inclusion rather than level of physical activity, which can be mutually exclusive. Future consideration should be given to occupational health surveillance and the identification of those at risk of excess sitting.

The present study took a novel approach adopting both the COM-B model and TDF to identify factors that could influence desk-based employee sitting behaviour, which ensured that the study was theoretically-underpinned and office worker-led, rather than researcher-led [66,67]. It also ensured that any future implementation would be person-centred [68]. Future research should take this behavioural diagnosis work and consider following the systematic approach provided by the developers of the COM-B model and Behaviour Change Wheel [35] to further identify intervention functions, policy categories and behaviour change techniques to support the development and examination of an intervention to change sitting behaviour of office workers.