Breaking barriers: using the behavior change wheel to develop a tailored intervention to overcome workplace inhibitors to breaking up sitting time

Due to modernisation of society and technological advancements, there is now heavy reliance on computers in the workplace resulting in occupations being less physically demanding and more sedentary [1, 2]. Sedentary behaviour is any waking activity, such as sitting, reclining or lying which expends less than 1.5 metabolic equivalents [3]. From an operational standpoint, prolonged sitting at a desk is the type of sedentary behaviour typically observed in the office workplace. Seventy-three percent of the UK population aged 16–64 are currently in employment [4, 5] with a large number of these workers in office-related jobs [6]. Studies have identified that the workplace contributes to the majority of excessive daily sitting time in office-based employees [7, 8]. Self-reported occupational sitting time has been estimated at 6 h 30 min (IQR = 6 h 20–6 h 45 min) on a work day [9], which is in accordance with objective measurements of workplace sitting suggesting 71% [10] to 82% of the workday is spent seated [11]. Due to growing epidemiological evidence linking excessive sitting time to adverse cardiometabolic outcomes, such as cardiovascular disease, obesity, type 2 diabetes [6, 12–17] and poor mental wellbeing [13, 18–20], the workplace has become an important public health concern.

Two observational studies have shown that daily participation in moderate-to-vigorous physical activity (MVPA) for 60–75 min a day may eliminate the increased risk of premature mortality associated with high amounts of sitting [21, 22]. However, the majority of the population do not engage in such high levels of MVPA [23, 24]. For those who are unable to achieve these high levels of MVPA, and in order to mitigate the remaining cardiometabolic health risks, the workplace could be a potential intervention environment to break up and reduce excessive sitting [25–27]. To develop effective interventions to reduce and break up sitting, it is pertinent to understand what works and why [28].

Interventions targeted at changing behaviour need to be informed by theoretical, evidence-based frameworks. The Medical Research Council [29] has outlined recommendations that should be used when developing and evaluating complex interventions. These guidelines state that interventions should start with a theory phase before progressing to modelling and then an experimental phase [29–31]. Whilst this current work focuses on modelling, the theory phase involves the collection of evidence and analyses via theoretical frameworks through which an intervention can be developed and modelled. The modelling stage involves hypothesising what should be targeted (determinants of behaviour) and how this can be achieved (via behaviour change techniques) [32]. A wide range of theoretical models of behaviour have been developed including the Theory of Planned Behaviour [33] and the Health Belief Model [34]. One common limitation of these theories is that they only help to understand or predict behaviours [35] and do not help to understand behaviour change [36] or develop interventions.

While it is important to identify how a behaviour maps to COM-B, the intervention functions (that form the third layer of the BCW) selected as a result must also make practical considerations. One method that has been developed to assist researchers to narrow down feasible intervention functions is to consider Affordability, Practicability, Effectiveness and Cost-effectiveness, Acceptability, Side effects/safety and Equity through the APEASE criteria [37]. Use of these criteria allow researchers to look beyond the BCW and explore feasibility issues before trialling an intervention. Using the BCW to design interventions is becoming more common, and it has been successfully used to understand behaviour change in different contexts, such as sexual counselling [48], medication management [49, 50], auditory rehabilitation [51], and physical activity [52]. Development of interventions using the full BCW to reduce workplace sitting however, is limited, with only the Stand More AT Work (SMArT) study found to target hospital office workers [53]. Ten behaviour change techniques that could be used to target individual, environmental and organisation level barriers to reducing sitting were identified in the SMArT study [53]. This intervention decreased workplace sitting time by 50.6 min at 3 months and by 64.4 min at 6 months and with evidence of sustainable long-term positive effects on job performance [54]. Therefore, further studies using this BCW framework are needed. The aim of this work is to develop, through qualitative interviews, a tailored intervention package using the BCW that could be used in future interventions to reduce and break up sitting time in desk-based employees.

The first step involves defining the problem of interest that requires intervention in behavioural terms. This means identifying the problem, and specifying the behaviour and target population [37]. Previous evidence [10, 11, 55–58] suggests that increased sedentary time is a behavioural problem significantly associated with cardiometabolic risk and poor mental wellbeing [12, 16, 59]. With office workers engaging in sitting for approximately two-thirds of their total working time and their sitting bouts often lasting at least 30 min [6, 11, 60, 61], there are possibilities that the workplace may be a major contributor to increased cardiometabolic disease risk.

This step explains that long lists of all other behaviours that may influence the target behavioural problem need to be generated. This can then be systematically reduced by considering the possible impact of each of these behaviours. For this research, behaviours such as physical activity, sedentary behaviour and sitting time were considered.

Step three specifies the target behaviour by outlining the new behaviour in greater detail. Specifications should include: who needs to perform the behaviour, what do the persons need to do differently, when, where, how, and with whom will they do it [37]. In this research, the target behaviour is to break up and reduce sitting time at work which may follow guidelines in a recent expert statement [62], which states that office workers should initially reduce daily occupational sitting time by engaging in 2 h of standing or walking during working hours and gradually increasing this to 4 h per working day.

The recommended method to understand what needs to change is interviews or focus group discussions [63], as this would ensure future interventions are participant-centred and co-created [64]. This research aims to inform Step 4 by using semi-structured interviews to explore sitting behaviour in office workers drawing from both the COM-B and TDF. To achieve this, 25 office-based workers (26–59 years) with self-reported daily occupational sitting time of at least 5.5 h were purposively recruited from local council offices and a university in the East of England region and interviewed by one researcher (SO) in their respective offices. Questions asked in the interview were developed using the TDF [39] with each lasting between 30 to 60 min. Amendments were deemed not necessary following pilot interviews with three of the participants. Two researchers (SO & AC) independently coded the transcripts and maintained anonymity throughout by using pseudonyms [65]. The COM-B model and TDF were employed as a combined deductive framework for the analysis covering all the relevant determinants of behaviour [66, 67]. Comparisons of codes were made, and discrepancies resolved by discussion to produce ‘behavioural diagnosis’ (a selection of barriers and facilitators) for breaking up and reducing prolonged sitting in the workplace. The interview data was managed using NVivo qualitative data analysis software (Version 10, QSR International, Melbourne, Australia) while SPSS (Version 23, IBM, Chicago, IL, USA) was used for descriptive data analysis of participant characteristics.

This study also aimed to identify relevant intervention functions and policy categories to be used following the COM-B and TDF analyses and how each of the intervention functions could be supported at an organisational level [37]. The BCW guide recommends that intervention functions and policy categories should be assessed through the use of the APEASE criteria [37]. However, as this screening process is largely contingent on resource availability, which might be different for intervention developers, the onus to use APEASE criteria would lie on individual intervention developers. In this present study, relevance of APEASE criteria is highlighted but not applied.

The research finally aimed to identify the most appropriate BCTs that could result in the desired breaking up and reduction of workplace sitting. BCTs mentioned within the qualitative interviews were individually identified and selected for the development of a future intervention by two members of the team (SO and MB). These were then discussed with the rest of the research team led by AC for consensus. Then, the most appropriate mode of delivery of each technique was deliberated upon and selected by the authors. Examples of modes of delivery include face-to-face or distance delivery at the individual or group level via phone (voice or text), print or digital media, broadcast media, outdoor media, or individually accessed computer programmes [37].

Steps 1–3 have been described in the methods above. This research has generated new qualitative data from Steps 4–8 as described below.

Responses from the interviews have been categorised into capability, opportunity and motivation in line with the COM-B model and includes links to the TDF domains within the text.

With respect to policy categories, only three out of the seven categories highlighted in the BCW guide [37] were identified. These included Communication/marketing (for instance, using verbal, electronic communication or flyers to create awareness of benefits of breaking up sitting and health consequences of prolonged sitting), Guidelines (examples of which include informing employees of sitting time guidelines), and Environmental/social planning (e.g, designing and controlling the logistics of height-adjustable desks within the office setting/office culture).

BCTs are considered as ‘active components’ when designing an intervention. In total, 39 out of the 93 BCTs in the BCT Taxonomy Version 1 [47] were identified from the interview data (Table 1). The list of BCTs identified include: ‘Instruction on how to perform the behaviour’, ‘Credible source’, ‘Information about health consequences’, ‘Information about social and environmental consequences’, ‘Feedback on behaviour’, ‘Behavioural practice/rehearsal’, ‘Behaviour substitution’, ‘Habit formation’, ‘Habit reversal’,’ Prompts/cues’, ‘Adding objects to the environment’, ‘Restructuring the physical environment’, ‘Self-monitoring of behaviour’, ‘Monitoring of behaviour by others without feedback’,’ Problem solving’, ‘Action planning’, ‘Body changes’, ‘Information about others’ approval’, ‘Social comparison’, ‘Restructuring the social environment’, ‘Social support (unspecified)’, ‘Social support (practical)’, ‘Demonstration of the behaviour’, ‘Goal setting’, ‘Verbal persuasion about capability’, ‘Mental rehearsal of successful performance’, ‘Material incentive (behaviour)’, ‘Material reward (behaviour)’, ‘Non-specific reward’, ‘Social reward’, Social incentive’, ‘Non-specific incentive’, ‘Self-reward’, ‘Information about emotional consequences’, ‘Reduce negative emotions’, ‘Self-monitoring of outcome(s) of behaviour’, ‘Behavioural experiments’, ‘Information about antecedents’ and ‘Incentive (outcome)’.

Intervention designers will need to select BCTs that are most appropriate for the population and location where the intervention will be conducted. This can be achieved by considering the APEASE criteria or by first choosing BCTs that were most frequently used within relevant intervention functions before those that were less frequently used as described in the BCW guide [37].

The appropriateness of mode of delivery depends on the target behaviour, target population and setting. Details on taxonomy of modes of delivery can be found in the BCW guide [37]. APEASE criteria should be used in selecting mode of delivery of choice. This could be either face-to-face or distance depending on setting. Where employees are spread over different offices and different locations, interventions could be delivered face-to-face, in clusters or individually. This can be achieved by giving out leaflets with detailed information about breaking up sitting, sitting guidelines for office workers and demonstrated using digital media.

The aim of this work was to use qualitative interviews with desk-based employees to highlight aspects of the BCW that can be used to develop a tailored intervention package that could be employed in breaking up and reducing workplace sitting. This research describes the systematic process used to model determinants of workplace sitting behaviour by qualitatively analysing sources of behaviour with the COM-B/TDF model, linking to subsequent intervention functions and policy within the BCW, and finally, identifying the appropriate behaviour change techniques to use when developing a tailored intervention to break up office workers’ sitting time. The majority of the participants in this study were not aware of any published recommendations for reducing sitting in the workplace [62]. However, participants expressed a keen interest in changing their sitting behaviour, suggesting that a workplace intervention targeted at sitting patterns would be acceptable.

The main reasons cited for prolonged sitting at work were the sedentary nature of the job, forgetfulness due to a heavy workload, an unsupportive physical workspace, and the organisational and social culture. These findings are consistent with previous studies that identified organisational cultural norms around “appropriate” workplace behaviour, environmental changes and workload pressures as barriers to breaking up workplace sedentary time [68–70]. The interview responses suggested that interventions should include education about sitting guidelines, health and emotional consequences of prolonged sitting and the benefits of reducing sitting time; prompts to serve as reminders to break up sitting; environmental modification, such as the provision of height-adjustable desks to alternate between sitting and standing without disrupting work; and changes to social and organisational support. Previous studies [71–73] have reported similar findings that breaks from prolonged sitting need to be seen as a “normal” activity in the workplace in order to prevent perceived criticism from colleagues. Organisational support would address this change. This could be an important strategy to prevent sedentary behaviour-induced diseases, due to a probable connection between social support, role-modelling, and social norms and the development of chronic diseases associated with prolonged sedentary behaviour [74].

In terms of the COM-B model, this study identified Psychological Capability, Social and Physical Opportunity as well as Reflective and Automatic Motivation as key targets for a behaviour change intervention for reducing and breaking up sitting time at work among office workers. In addition, the results from interviews with the participants suggested that Knowledge, Skills, Reinforcement, Goals, Intentions, Environmental context and resources, Social influences, Behavioural regulation, Emotion, and Memory, attention and decision processes were important TDF domains that need to be targeted in work-based sitting interventions. Consequently, seven intervention functions including Education, Training, Modelling, Persuasion, Enablement, Environmental restructuring and Incentivisation were identified as relevant for a sedentary workplace intervention. These results are in alignment with the SMArT study by Munir et al. [53] in which the BCW was also used to design a workplace sitting reduction intervention in hospital office workers. They identified the TDF domains of Knowledge, Social identity, Intentions, Beliefs about capabilities, and Self-regulation of behaviour, and consequently the key intervention functions of Education, Enablement, and Training. However, it should be noted that the present study identified a broader range of intervention functions due to the fact that the SMArT study applied the APEASE criteria to select the most relevant intervention function for the target population.

Consistent with the findings of this present study, Gardner, Smith [28] in their systematic review sub analysis of workplace interventions found 6 BCTs that frequently appeared in effective interventions to reduce sedentary behaviour: Review behavioural goals, Self-monitoring (behaviour), Instruction on how to perform behaviour, Information on health consequences, Behaviour substitution, and Adding objects to the environment. All but one of these BCTs, Review behavioural goals, was also found in the present study. The BCTs identified in this present study were identified from the qualitative data and it may be that Review behavioural goals was a BCT identified by interventionists from psychological theory sources. Therefore, when tailoring future interventions, researchers should consider including theoretically derived BCTs as well as those generated from the target population.

This study has identified possible components of a workplace intervention to break up and reduce sitting behaviour in the workplace based on the needs of office workers. This study emphasises the need for interventions to be targeted at multiple levels of influence on behaviour. Consequently, 39 BCTs have been identified and can be used as active ingredients in preparation for targeting the key determinants (Psychological Capability, Physical and Social Opportunity and Reflective and Automatic Motivation) of sitting behaviour in the workplace. Sedentary behaviour intervention designers should apply the APEASE criteria to determine the most appropriate intervention functions, policy categories and BCTs to use, drawing on the evidence presented here that identifies what needs to change. Future research can use the insight and modelling from this paper to test the effectiveness of an intervention based on the findings presented here, during an experimental phase as suggested by the Medical Research Council. This next phase could then provide an empirical basis for sitting behaviour policy implementation in the workplace.