Employer schemes to encourage walking to work: feasibility study incorporating an exploratory randomised controlled trial

Article history

The research reported in this issue of the journal was funded by the PHR programme as project number 10/3001/04. The contractual start date was in October 2011. The final report began editorial review in April 2014 and was accepted for publication in July 2014. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The PHR editors and production house have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the final report document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

Rona Campbell is a director of DECIPHer IMPACT Ltd, a university-owned not-for-profit company that licenses and supports the implementation of evidence-based health promotion interventions, for which she receives fees paid into a grant account held by the University of Bristol used to support further research activities. She is also a population and public health member of the Wellcome Trust’s Expert Research Group; a fee was paid for the time spent reviewing applications and attending board meetings in London. Adrian Davis is an independent public health consultant on transport planning and health, promoting evidence-based transport policy and practice, and has been paid by Bristol City Council in this capacity.

Copyright statement

© Queen’s Printer and Controller of HMSO 2015. This work was produced by Audrey et al. under the terms of a commissioning contract issued by the Secretary of State for Health. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

Background

Physical inactivity increases the risk of many chronic diseases, including coronary heart disease, type 2 diabetes, obesity and some cancers. 1 It is currently recommended that adults should aim to undertake at least 150 minutes of moderate intensity physical activity in bouts of 10 minutes or more throughout the week. 2,3 There are concerns that many adults in the UK and other high-income countries do not achieve this,1,3,4 although allowing for the accumulation of 150 minutes in bouts of 10 minutes has led to an estimated 61% of adults in England self-reporting that they do achieve the recommended levels. 5 Increasing physical activity levels, particularly among the most inactive, is an important aim of current public health policy in the UK. 1,6

In addition, there is increasing interest in the relationship between time spent sedentary (defined as any waking sitting or lying behaviour with low energy expenditure [≤ 1.5 metabolic equivalent of task (METs)] and health outcomes. 7 A large amount of time spent sitting has been associated with greater risk of cardiovascular disease, type 2 diabetes, and cardiovascular and all-cause mortality. In addition, a high volume of objectively measured sedentary time has been associated with a poorer metabolic profile in healthy adults and those at risk of and having developed type 2 diabetes. It is of note that these associations are independent of the volume of moderate to vigorous physical activity (MVPA), and consequently UK health guidelines recommend that adults should minimise the amount of time spent sedentary (sitting) in addition to increasing physical activity. 1

There is evidence of the link between adult obesity levels and travel behaviour, one indicator of which is that the countries with the highest levels of active travel generally have the lowest obesity rates. 8 Walking has been described as near-perfect exercise. 9 It is a popular, familiar, convenient and free form of exercise that can be incorporated into everyday life and sustained into older age. It is also a carbon-neutral mode of transport that has declined in recent decades in parallel with the growth in car use. 1 Even walking at a moderate pace of 5 km/hour (3 miles/hour) expends sufficient energy to meet the definition of moderate intensity physical activity. 10 Hence there are compelling reasons to encourage people to walk more, not only to improve their health but also to address the problems of climate change. 11–14

In the UK, there are substantial opportunities to increase walking by replacing short journeys undertaken by car. For example, the 2011 National Travel Survey showed that 22% of all car trips were shorter than 2 miles in length, while 18% of trips of less than 1 mile were made by car. 15 An opportunity for working adults to accumulate the recommended moderate activity levels is through the daily commute, and, in addition, replacing the car for short journeys is likely to reduce sedentary time. Experts in many World Health Organization (WHO) countries agree that significant public health benefits can be realised through greater use of active transport modes. 16 Furthermore, cost–benefit analysis for the UK Department for Transport suggests that the ratio of benefits to costs is high. 17

A systematic review comparing direct versus self-report measures for assessing physical activity in adults found that self-report measures were higher than objective measures in some cases and lower in others. 18 This calls into question the reliability of self-report measures, and indicates that there is no approach to correcting for self-report measures that will be valid in all cases. However, very few studies have objectively measured the contribution of walking, particularly walking to work, to adult physical activity levels; more evidence is needed. 19

In Sweden, two studies examined the association between neighbourhood walkability [measured using a geographic information system (GIS)] and objective physical activity (measured using accelerometers). 20,21 Both studies demonstrate how increased walking rates translate directly to increased MVPA. In the USA, a cross-sectional study included 2364 participants enrolled in the Coronary Artery Risk Development in Young Adults (CARDIA) study who worked outside the home during year 20 of the study (2005–6). 22 Associations were examined between walking or cycling to work and objective MVPA using accelerometers. Active commuting was found to be positively associated with fitness in men and women, and inversely associated with body mass index (BMI), obesity, triglyceride levels, blood pressure and insulin level in men. The authors concluded that active commuting should be investigated as a means of maintaining or improving health.

Systematic reviews have examined the effectiveness of interventions to promote physical activity in general,23–26 but there is less evidence about how best to promote walking to work. Available systematic review evidence has focused on interventions that promote walking; interventions that promote walking and cycling as an alternative to car use; and the effectiveness of workplace physical activity interventions. None focuses specifically on employer-led interventions that promote walking to work, although the studies that have been undertaken are included within the available systematic review evidence.

A systematic review of interventions to promote walking comprised 19 randomised controlled trials (RCTs) and 29 non-randomised controlled studies. 27 The review identified two general characteristics of interventions found to be effective: targeting and tailoring. Only six studies included even a rudimentary economic evaluation. A systematic review of promoting walking and cycling as an alternative to using cars26 identified 22 studies that met the inclusion criteria and found some evidence that targeted behaviour change programmes can change the behaviour of motivated subgroups. A systematic review of the literature regarding the effectiveness of workplace physical activity interventions, commissioned by the National Institute for Health and Care Excellence (NICE), graded 14 studies as high quality or good quality. 28 Three public sector studies provided evidence that workplace walking interventions using pedometers can increase daily step counts. One good-quality study reported a positive intervention effect on walking to work behaviour (active travel) in economically advantaged female employees. There was strong evidence that workplace counselling influenced physical activity behaviour, but the reviewers indicated that there was a dearth of evidence for small- and medium-sized enterprises.

National Institute for Health and Care Excellence public health guidance on workplace health promotion concluded that although a range of schemes exist to encourage employees to walk or cycle to work, little is known about their impact. 29 Few studies used robust data collection methods to measure the impact of workplace interventions on employees’ physical activity levels (most use self-report) and there is a lack of studies examining how workplace physical activity interventions are influenced by the size and type of workplace and the characteristics of employees.

Benefits and risks of walking as active travel

Physical activity is an important element of a healthy lifestyle. In England, the Chief Medical Officer has stated that the target of 30 minutes of moderate intensity activity for adults on at least 5 days per week, in order to promote health, will best be achieved by helping people to build activity into their daily lives. 1 Experts in many Organisation for Economic Co-operation and Development (OECD) countries agree that significant public health benefits can be realised through greater use of active transport modes. 16 Furthermore, the ratio of benefits to costs is high. 17 There are also potential benefits to walkers from reduced commuting costs and greater certainty about the timing of the journey to work. The potential benefits to employers who promote walking to work include a reduction in absenteeism; employees’ increased concentration and mental alertness, and better rapport with colleagues; a reduction in employees being late because of greater certainty about the timing of the journey; improved public image as a result of lowering the workplace’s carbon footprint; and savings in parking costs. 30

Despite the benefits, there are potential risks to pedestrians. For example, in Great Britain in 2008 there were 170,591 reported personal injury road accidents, of which 1 in 6 involved a pedestrian; 572 pedestrians were killed (23% of the total road accident fatalities); 6070 were seriously injured (23% of all seriously injured casualties); and 21,840 were slightly injured (11% of all slightly injured casualties). 31 In 54% of accidents, contributory factors were assigned to pedestrians only (‘pedestrian failed to look properly’ being the most common individual factor). In 25% of accidents, at least one factor was assigned to both a pedestrian casualty and a vehicle (the most common combination being both participants failing to look properly). In 21% of accidents, factors were associated only with vehicles involved (‘failed to look properly’ being the most common vehicle factor). There is also potential for harm in relation to personal safety of walkers where lighting is poor or there is potential for street crime.

Using behaviour change techniques to encourage active travel

Behaviour change techniques (BCTs) have been defined as the ‘active ingredients’ within an intervention designed to change behaviour that are observable, replicable and irreducible components which can be used alone or in combination. 32 A taxonomy of 26 BCTs was identified in 2008,33 with subsequent work undertaken to improve labels and definitions and to reach a wider consensus of agreed distinct BCTs. 32,34 The 2008 taxonomy has been successfully used to categorise the BCTs used in healthy eating and physical activity interventions, with ‘self-monitoring’ combined with at least one other technique identified as the most effective. 35 For walking and cycling interventions that have resulted in behaviour change, the most commonly used techniques are ‘self-monitoring’ and ‘intention formation’. 36 The UK’s NICE has recently issued recommendations advising that interventions should use BCTs based on goals and planning, feedback and monitoring, and social support. 37

The effectiveness of interventions to promote active travel tends to be measured using self-report surveys. Of the 46 walking and cycling controlled interventions coded for BCTs by Bird et al. ,36 21 reported a statistically significant effect using a mean number of BCTs of 6.43 [standard deviation (SD) = 3.92]. However, very few published studies have looked qualitatively at participants’ views of the BCTs used in the intervention.

The socioecological model

Individual behaviour does not occur in a vacuum, but is shaped by external factors. In acknowledging this, initiatives to encourage walking to work can be considered within a socioecological model,38 identifying and exploring intrapersonal, interpersonal, community and organisational factors within the overall policy context. Within this model, BCTs to encourage walking to work operate at the intrapersonal and interpersonal levels, but the ability to implement and act upon them is shaped by policies and practices in the workplace, the neighbourhood and wider society.

Rationale for a feasibility study focusing on walking to work

A number of high-profile active travel initiatives focus on cycling. 39,40 However, walking may be perceived by employees as a cheaper and safer option: it requires no special equipment and is less likely to involve direct competition with motorised traffic for road space. In addition, there is no requirement for employers to provide special parking facilities or changing and showering facilities. Compared with other forms of physical activity, walking is a popular, familiar, convenient, readily repeatable, self-reinforcing, habit-forming activity and the main option for increasing physical activity in sedentary populations. 9,26 There is a dose–response curve to physical activity, so that the greatest health benefits are achieved when the least active undertake some physical activity. In industrialised countries, higher levels of mortality and morbidity from obesity and physical inactivity-related diseases disproportionately affect those in poorer communities. Walking is the most obvious, low-cost, immediate and normative means by which to increase physical activity and may, therefore, help to address health inequalities as well as be beneficial to people across all socioeconomic groups.

At the time of writing the proposal, there was a paucity of information about the design and implementation of interventions specifically promoting walking to work (as opposed to walking for pleasure, walking as part of a rehabilitation programme for a specific health problem, or active transport initiatives which include cycling and walking but tend to promote cycling). This continues to be the case. There was also little, if any, information about differences between small and large workplaces in the delivery or receipt of such interventions. Under such circumstances, the Medical Research Council (MRC) framework for the evaluation of complex interventions emphasises the importance of undertaking a feasibility study before embarking on a phase 3 RCT. 41,42

Phase 1 focuses on existing evidence and theory to inform the development of the intervention and ways to evaluate it. Phase 2 involves testing the feasibility of a full-scale trial, including the ability to recruit and retain participants, the acceptability of the intervention and the methods used for its evaluation. An exploratory trial also provides a sound basis for calculating the sample size for a main trial. As workplace interventions are targeted at the cluster level, an estimate of the intracluster correlation coefficient (ICC) is needed. The Consolidated Standards of Reporting Trials (CONSORT) group has indicated that the conduct and reporting of cluster trials has been poor, including failure to account for clustering in the planning of trials or to account for clustering in the sample size calculations and analysis. 43 Without published data from a full-scale cluster RCT of a workplace intervention to promote walking to work, there was no estimate of the ICC. Estimates from other studies in different settings or using different outcomes may provide some information, but if the ICC is underestimated, then a full trial may be considerably underpowered. Consequently, there was a need to calculate an estimate of the ICC. Furthermore, in order to perform a formal sample size calculation for a definitive cluster trial, a good estimate of likely recruitment and adherence rates and feasible differences in outcome was required. Within a cluster trial, the average cluster size dramatically affects the inflation factor required for sample size.

As well as contributing to statistical analyses, an exploratory trial offers opportunities for process evaluation and estimations of costs. Conducting a process evaluation within a feasibility study enables the examination of the context and delivery of and response to an intervention. Consideration can also be given to the facilitators of and barriers to its successful implementation. In addition, an assessment of the costs and potential benefits of an intervention is required when considering if it is likely to be adopted into routine practice.

Study aim and objectives

The overall aim of the research was to build on existing knowledge and resources to develop an employer-led scheme to increase walking to work and to test the feasibility of implementing and evaluating it in a full-scale RCT.

The objectives were:

• to explore with employees the barriers to and facilitators of walking to work

• to explore with employers the barriers to and facilitators of employer-led schemes to promote walking to work

• to use existing resources and websites to develop a Walk to Work information pack to train work-based Walk to Work promoters

• to conduct an exploratory RCT of the intervention to:

  • pilot workplace and employee recruitment procedures

  • determine eligibility and consent rates

  • examine retention rates

  • pilot cost and outcome measures

  • inform a sample size calculation for a full RCT with estimation of potential feasible differences in outcome

  • gather information regarding variablility within and between workplaces

• to pilot the use of accelerometers and global positioning system (GPS) monitors to measure:

  • eligible employees’ levels of MVPA

  • daily steps taken in walking

  • temporal pattern of physical activity (to identify when activity has increased and whether or not there is a compensatory decrease in activity at other times)

  • route taken and physical activity associated with journey

• to explore any social patterning in uptake of walking to work, particularly in relation to socioeconomic status (SES), age and gender

• to examine whether or not the size or type of workplace influences uptake of walking to work

• to assess intervention costs to participating employers and employees

• to provide preliminary evidence on the cost and economic benefits of the intervention to employers, employees and society.

Research design

During phase 1, a review of current resources that promoted walking (and in particular the benefits of walking to work) was undertaken, and three focus groups with employees plus three interviews with employers were conducted in workplaces (small, medium and large) outside of Bristol to finalise the intervention design. Phase 2 comprised an exploratory randomised trial in Bristol. The trial included an integral process evaluation and an assessment of intervention costs to participating employers and employees.

Setting

The intervention took place in Bristol. With a population of 433,100 people, Bristol is the largest city in the south-west and one of the eight core cities in England, excluding London. 44 The areas of highest population growth are all concentrated around the city centre and nearly half (46%) of all jobs are located in the city centre. 45 A substantial number of people in Bristol make short journeys to work by car. The 2011 census indicated that there are 44,000 people who travel < 5 km to work yet still go by car, of whom 13,000 drive < 2 km. 46 The intervention was implemented in 17 workplaces (eight small, five medium and four large). Seven workplaces received the intervention, and 10 workplaces constituted the control arm.

Workplace recruitment and randomisation

Workplaces were approached through BusinessWest for initial expressions of interest, including willingness to allocate employee time for study activities. A publicly available list of major employers was also used to identify large employers who were not on the BusinessWest mailing list. The e-mail contact included an information leaflet giving details of the study and the expectations of participating employers and employees, with the contact details of the research team. In addition, paper copies were sent out, as it was felt that, for some employers who receive large numbers of e-mails, the e-mail invitation might be deleted as low priority. The paper copy included a pre-paid reply envelope. Workplaces expressing an interest were sent a short questionnaire to enable purposive sampling of pairs (one intervention, one control), with each pair containing workplaces as similar as possible with respect to total number of employees (up to 50, 51–250, 250+),47 location characteristics and type of business. Assignment of workplaces to the intervention group employed computer-generated allocation.

Study participants

Employees living within 2 miles of the workplace (‘walking distance’) were given information about the study, by e-mail or letter as appropriate, to be distributed through the workplace with an invitation to participate. As the study progressed, it was felt that this was too restrictive, and a second round of recuitment was undertaken to include people who lived further away and were willing to incorporate some walking into their daily commute.

The Walk to Work intervention

The socioecological model informed the Walk to Work intervention, which was designed to encourage employers and employees to consider, and where possible address, the facilitators and barriers at each level of the socioecological model (Table 1).

The intervention was based on the idea of recruiting and training, with the permission and support of the employer, a ‘champion’ (Walk to Work promoter) within each participating workplace to promote walking to work. The role of these Walk to Work promoters was to promote walking to work with their employer and among eligible employees in their workplace, to be ‘role models’ for the Walk to Work intervention and to be the recognised ‘point of contact’ for the Walk to Work intervention in their place of work.

The intervention focused on nine BCTs which included three categories (goals and planning, feedback and monitoring, and social support) as recommended by NICE. 37 The nine BCTs and the contact points at which they were used are listed in Table 2.

Following the resource review and focus groups during phase 1, members of the research team developed booklets for Walk to Work promoters and participating employees. The packs included information about the health, environmental, economic and social benefits of walking to work derived from, for example, Walk4Life (www.walk4life.info), Walkit.com urban route planner and Living Streets (www.livingstreets.org.uk). Specific BCTs used were providing information on the link between walking and health; identifying barriers and ways to overcome them; goal setting and review of goals; prompting self-monitoring by use of a travel diary and pedometer; providing social support and encouragement; and relapse prevention. There is evidence that these techniques can effect behaviour change. 33,36

There were four main stages of the intervention:

• Eligible employees were identified through workplace records (matching postcodes of workplace and home address, and checking for disability status or other disqualifying factors, e.g. delivery drivers).

• Walk to Work promoters, either volunteers or those nominated by participating employers, were trained by expert members of the research team about the health, social, economic and environmental benefits of walking to work and how to identify and promote safe walking routes for employees. This half-day training took place at the University of Bristol for those who were able to attend. Those who were not were given an individual training session at their place of work. All Walk to Work promoters were given a booklet to guide them in their role, as well as Walk to Work booklets and pedometers for participating employees. They were also trained to access relevant websites and toolkits (e.g. Walkit.com, Living Streets). The aim was a maximum of 25 participants to each Walk to Work promoter.

• The Walk to Work promoters were asked to discuss the Walk to Work intervention with study participants in their workplace and to ask those who were interested in walking to work to ‘sign up’ at this stage. The promoters were also encouraged to identify safe, feasible walking routes for their colleagues and to help them to set goals for walking to work.

• Further encouragement was to be provided through four contacts from the Walk to Work promoter over the following 10 weeks (face to face, e-mail or telephone, as appropriate for the workplace). The Walk to Work promoters were encouraged to focus on specific BCTs as outlined in Table 2.

During the first contact with the Walk to Work promoters, those who agreed to try changing their travel behaviour were asked to identify barriers, propose solutions and develop a plan of how they might increase walking. This involved setting short-, intermediate- and long-term goals, for example walking 1 day per week in the first week and then increasing it during the course of the intervention. Participants were issued with a Walk to Work booklet and pedometer. They were also encouraged to complete diary sheets and record whether or not they had walked and, for those using pedometers, how many steps had been registered.

The promoters were asked to make three further contacts with participants, in person, by e-mail or by telephone according to the needs of the workplace and employees. Week 3 (contact 2) focused on social support from other people, such as colleagues, family or friends; week 5 (contact 3) stressed a review of goals to see whether they had been achieved or needed to be adjusted; and from week 7 (contact 4) until the end of the intervention, the aim was to prevent relapse by supporting and encouraging participants to continue working towards their goals.

Assessment of harms

Although this was a low-risk intervention, the team were mindful of the potential for harm in terms of road traffic accidents; personal safety of walkers where lighting is poor or there is potential for street crime; difficulties experienced by Walk to Work promoters, including disrupting usual working relationships and employers’ attitudes towards time taken out of usual work activities; and costs to employers, including disruption to work routines, of permitting the intervention during working hours. It was also possible that people with low activity and no history of walking would suffer initial muscle stiffness. To address these issues, the Walk to Work booklets included advice about setting appropriate goals and personal safety. Participating employers were given information in advance about the level of time commitments and the potential benefits of the scheme. Employers and employees were given the contact details of the principal investigator to report any adverse incidents, which would then be recorded and kept on file, with any relevant participants informed immediately (e.g. other employees taking a similar route across a dangerous road or through a dimly lit area with a high rate of street crime).

Outcome measures

Because this was a feasibility study, the outcomes did not relate to the effectiveness of the intervention but to the feasibility of, and requirements for, a full-scale trial. Several items were identified relating to the trial design, physical activity measurement, the context, delivery and receipt of the intervention, and the assessment of costs. The main study outcomes are listed in Table 3.

Assessment and follow-up

At the baseline data collection (DC1), participants in the intervention and control arms were asked to complete a questionnaire giving basic personal data, including postcode (to assess distance from home to work), job title, mode of transport to work, before- and after-work ‘routines’ affecting travel mode (e.g. school run), typical commuting costs, household car ownership, commute-related adverse events, health service use and views about walking. Eligible employees were also asked to wear an accelerometer for 7 days from waking in the morning until going to bed at night to provide an objective measurement of physical activity (including intensity and step counts), and a personal GPS receiver during the journey to and from work to confirm the duration of the journey and quantify its contribution to overall physical activity. The GPS recorded location and speed at 10-second intervals while outdoors. Participants were required to turn it on at the start of the journey to work and off when the journey home ended, as the battery life was only 24 hours. Participants were also given small chargers for the monitors, with written instructions. A £10 gift voucher was given to participants who returned accelerometers and GPS receivers.

Immediately post intervention (DC2), questionnaires were administered in the intervention and control arms to explore attitudes towards and experiences of walking to work, including perceived barriers and facilitators and emotional and physical well-being. Additional questions about the acceptability of the intervention were included for the intervention arm only. At 1-year follow-up (DC3), questionnaires, accelerometers and GPS receivers were administered again in the intervention and control arms (as per baseline protocol).

Measuring physical activity

To address the potential for the time of year to confound differences in the trial arms, the outcome data were collected simultaneously in each ‘pair’ of workplaces.

Physical activity was measured objectively using accelerometers (ActiGraph GT3X+; ActiGraph LLC, FL, USA) worn on a belt around the waist during waking hours for 7 days and removed for swimming and bathing. Accelerometers were set to record data at 30 Hz. The protocol, decisions and outcomes for analysis of accelerometer data are listed in Table 4.

Participants also wore a personal GPS receiver (QStarz BT-Q1000XT, Qstarz International Co. Ltd, Taiwan) on the same belt during their commuting journeys for 7 days to allow the journey to and from work to be spatially described. GPS data were recorded at 10-second intervals and, where possible, the ‘assisted GPS’ mode was used to enhance the precision of the GPS location data.

Raw accelerometer data were downloaded using Actilife 6 software (ActiGraph LLC, FL, USA) and reintegrated to 10-second epochs for analysis and matching with GPS data. Reintegrated accelerometer data were processed using Kinesoft (v3.3.62; KineSoft, Saskatchewan, Canada) data reduction software to generate outcome variables. Continuous periods of 20 minutes of zero values were considered to be ‘non-wear’ time and removed. Outcome variables were physical activity volume [mean daily accelerometer counts per minute (cpm)] and intensity (MVPA and sedentary time) defined using validated thresholds (sedentary < 100 cpm; MVPA > 952 cpm). 48

Combining accelerometer and global positioning system data

Accelerometer and GPS data were combined (accGPS) based on the timestamp of the ActiGraph data. For measurement of the journeys to and from work, the participant’s workplace and home were geocoded using the full postcode, and imported into GIS software (ArcMap v10, Esri, Redlands, CA, USA). The merged accGPS files were then imported into ArcMap and journeys to and from work were visually identified and segmented from other accGPS data using the ‘identify’ tool. Journeys were identified as a continuous (or near-continuous) sequence of GPS locations between the participant’s home and workplace, and thus included trips to other destinations (e.g. supermarkets) if taken as part of the journey to or from work.

Baseline data analysis: the contribution of walking to work to physical activity

A cross-sectional study was undertaken using the baseline data to assess the contribution of walking to work to physical activity levels. 49 Analyses were confined to data recorded between 6.00 a.m. and midnight. Participants recorded travel mode to and from work for each day, and only days where participants used the same mode of transport both to and from work were included in the initial analyses, to allow classification as either an active commuter (walk or cycle) or a car user on each day. For participants who recorded using a mixed mode (e.g. walk and bus) for a journey, the mode of transport of the longest duration was considered to be their mode for that journey. If the modes were of equal duration, that day of data was excluded.

Participants who cycled to work (n = 19) were excluded from analyses because of the inability of waist-worn accelerometers to accurately record physical activity during cycling, as were the small number (n = 7) using ‘other’ modes of travel. Where a travel diary was not completed, usual travel mode was determined from the baseline behaviour questionnaire where possible (n = 10). Differences in physical activity between travel modes (walking/car) were analysed by one-way analysis of variance (ANOVA). Paired samples t-tests were used to compare weekday and weekend values for total physical activity (cpm), MVPA and time spent sedentary, and to investigate differences in the volume of MVPA accumulated between overall accelerometer data and spatially segmented trips. Linear regression was used to explore the association between travel mode (walking/car) and total weekday physical activity (cpm) and MVPA (minutes per day). Models were adjusted for possible confounders [age, sex, education (educated to degree level or not)], and accelerometer wear time.

Statistical methods

Statistical analyses were largely descriptive. Reasons for non-eligibility, refusing consent and withdrawing from the study were fully documented. Within the intervention group, differences between employees who do and do not walk to work were explored, particularly in relation to SES, age and gender. Such exploration is hypothesis generating but may be investigated further in a full trial. In addition, such differences were examined to inform changes to the implementation or design of the ‘Walk to Work’ intervention and its evaluation prior to a full RCT.

Analyses of outcomes to be used as primary outcomes for a full trial were carried out at the employee level using multilevel regression models to account for the effects of clustering within workplaces. Models were adjusted for baseline and covariates including age, gender and SES. As an exploratory trial, the proposed study was not powered to provide a definitive comparison between the intervention and control groups. However, estimates of effect and 95% confidence intervals (CIs) were generated. CIs informed feasible effect sizes to be incorporated into sample size calculations for a full trial. The intracluster correlation was estimated to provide some information about variability within and between workplaces to inform the sample size calculation for a full trial.

Process evaluation

The process evaluation examined the context, delivery and receipt of the intervention from the perspectives of employers, Walk to Work promoters and employees. Researcher notes were taken at each stage of the intervention, i.e. recruitment of workplaces, recruitment of study participants, and the recruitment and training of Walk to Work promoters. In addition, post-intervention interviews were conducted with a purposive sample of employees who increased walking to work and employees who did not. A senior manager and the Walk to Work promoter(s) in each workplace were also interviewed. Process data were also gathered through the behaviour questionnaires administered at baseline, immediately post intervention and at 1-year follow-up. An outline of the process evaluation is provided in Table 5.

Descriptive statistics were compiled, for example of recruitment and retention. Qualitative analyses employed the framework method of data management, with textual data placed in charts in relation to specific research questions and then scrutinised for differences and similarities within emerging themes, keeping in mind the context in which these arose.

Qualitative data collection and analysis

Interviews with employers/managers from five workplaces, and focus groups with employees at three workplaces, were conducted during the phase 1 developmental work; these workplaces were not then recruited to the phase 2 exploratory trial. The aim was to explore the participants’ views of walking to work and, in particular, the proposed intervention and evaluation methods. During phase 2, baseline interviews were conducted with employers/managers (n = 12) of participating workplaces to understand the context of the workplaces, their views about travel to work, reasons for participating in the study and their opinions about the proposed intervention and its evaluation.

Immediately post intervention, 36 interviews were conducted with a purposive sample of participating employers and managers, Walk to Work promoters, and employees who had attempted to change their travel behaviour and those who had not. In addition, to further investigate issues relating to the recruitment of workplaces, semistructured interviews were conducted at six workplaces with employers/managers who had initially expressed an interest in the study but did not participate.

The focus groups and semistructured interviews were conducted in a private room at the workplaces of the participants. A topic guide was developed as part of the process evaluation to examine the context, delivery and receipt of the intervention. The topic guide allowed flexibility for participants to raise issues and follow their own train of thought, as well as including specific prompts. All interviews were audio recorded, fully transcribed and anonymised, and electronically stored in a secure folder.

The framework approach for data management was used to aid qualitative analysis. 50,51 The interview transcripts were read and reread, and textual data were placed in charts focusing on key research questions. The charts were scrutinised and each unit of data was subsequently coded. For example, the charts relating to the use of BCTs focused on barriers and enablers to walking to work, and the extent to which the BCTs facilitated change in travel behaviour. The charts relating to employers’ views of the intervention focused on employers’ views of the advantages and disadvantages of employer-led schemes to promote walking to work, and whether or not they felt that it was possible for employers to promote active travel.

Assessment of costs

Trainer time was costed using basic salary, national insurance and superannuation, and, where external consultants were employed, their day rate was divided by eight in order to identify a unit cost per hour. The number of promoters and number of individuals participating in the intervention at each workplace was recorded, which allowed us to estimate the number of promoter and employee booklets required. The booklets were printed by the University of Bristol print services and paid for by a small grant from Bristol City Council. The unit cost of pedometers, which were provided free of charge to the trial, was estimated from an online supplier (Be-Active Ltd, www.be-activeltd.co.uk/step-counter.htm).

The delivery of the promoter training day was planned to take place outside the workplace. However, it was not feasible for all of the promoters to attend these training days and therefore some promoter training was provided in the workplace. For training days outside the promoter workplace, costs included trainer and consultant time, lunch and drinks, and room hire. The unit cost of promoter time was calculated by dividing the upper quartile weekly earnings by the median number of hours worked per week. 52 The upper quartile was applied as it was believed that promoters were likely to be on a higher wage than the average worker. Promoter expenditure was also included with respect to travel to the promoter training day location. The cost per mile was calculated using the AA schedule of motoring costs. 53 Promoter training at the workplace included the cost of the trainer, travel costs to the workplace for the trainer, and promoter time.

Health service use in the past 4 weeks was self-reported by study participants at baseline and 1-year follow-up. Table 6 provides the unit costs that were used to value health service use. Primary care, including general practitioner (GP) visits, nurse visits, GP calls and nurse calls, was costed using national unit cost estimates. 54 All hospital-based care was costed using Department of Health reference costs. 55 Acupuncture and chiropractic care were costed using the NHS choices website. 56,57 Physiotherapy, chiropractic care, acupuncture, midwife visits, counselling and walk-in services were grouped together under ‘other care’ for the analysis. Routine dental and optician appointments were excluded from the analysis and a few health-care items were excluded because of lack of information on what service the participant had received. Medications were costed using the Prescription Cost Analysis. 58 Where medications could not be costed because of missing information on the number of days taken, it was assumed for chronic conditions and contraception that the medications were taken daily. For medications with missing information on the number of doses per day, an assumption of one dose was made. Any medications that could not be costed due to lack of sufficient information were excluded.

Self-assessed productivity based on the extent to which health problems affected it in the past 7 days was measured on a 10-point scale, where 1 indicated that health problems had no effect on an individual’s work and 10 indicated that health problems completely prevented work. Absence from work was also self-reported, whereby participants were asked to report the number of hours of work they had missed in the past 7 days because of health problems. 59

Information regarding the commute was recorded for 1 week at baseline and at 12-month follow-up in a travel diary. Time spent commuting by mode of transport, daily expenses (e.g. bus fare) associated with the commute and occasional expenses of commute (including bus and train passes and parking permits) were all recorded. A cost per mile of 0.665 pence53 and an average speed of 15.7 mph60 were used to calculate a cost per minute of driving, which was multiplied by the number of minutes driving in order to calculate the cost of commuting by car. Cost and duration of parking permits, bus and train passes were used to calculate weekly parking permit, bus and train costs. Car sharing was incorporated into the costs of car travel by dividing the cost of driving and the cost of a parking permit by the number of people in the car.

Study timetable and milestones

The study timetable is summarised in Table 7. To allow for seasonality, data collections took place concurrently in each ‘matched’ pair of workplaces (one in the intervention, the other in the control arm). In addition, baseline (DC1) and 1-year follow-up (DC3) data collections were undertaken in the same season (April to June).

Public involvement

The public were involved in shaping the intervention through the phase 1 focus groups with employees in three workplaces and the individual interviews with employers in five workplaces. Their views were sought on the design of the Walk to Work intervention and its evaluation, including the use of accelerometers and GPS monitors. Data from these focus groups and interviews shaped the intervention and its evaluation. Additional interviews were conducted with employers (n = 15) and employees (n = 33) following implementation of the intervention during the phase 2 exploratory trial. Participants identified a number of improvements that could be made, should an application for a full-scale trial be made: a simplified recruitment process, greater flexibility for attending the Walk to Work promoter training, and provision of additional information to employers in the intervention arm about changes they can make in the workplace to support employees who are attempting to increase walking to work. Representatives of non-academic organisations also influenced the study: a director of Sustrans advised on promoting active travel; a transport consultant with Bristol City Council helped to design and implement the training programme; and representatives of BusinessWest advised on, and supported, the recruitment of workplaces.

A feedback event was held in February 2014. Employees, employers and Walk to Work promoters attended a half-day event at which the research team presented findings, and participants were invited to give feedback on the intervention and its evaluation. Expressions of interest were collected for a public advisory group should a full-scale trial be funded for wider implementation.

Changes to protocol

There were no major changes to protocol. The number of workplaces recruited was increased because of the smaller-than-anticipated cluster size. Recruitment of employees was extended to include employees who lived further than 2 miles from the workplace. A request for a 3-month non-financial extension was granted towards the end of the study.

Wider context

Two important factors are likely to have influenced recruitment to the study and the actions and attitudes of participants. The study took place in the aftermath of a global banking crisis. 61 This resulted in economic insecurity, with businesses restructuring and an increase in unemployment. Under these circumstances, health promotion interventions may be seen as less important to employers and employees than the survival of businesses and the retention of jobs. Secondly, the intervention was implemented during the wettest summer in the UK for 100 years. 62 Weather conditions have been identified as an important barrier to walking and this is likely to have been exacerbated by the weather during the summer of 2012.

Workplace recruitment and retention

A summary of workplace recruitment is included in Figure 1. Following an e-mail sent through the Chambers of Commerce contact list, and a paper mailing to workplaces included in a publicly available list of large employers, 55 workplaces expressed an interest in the study. Small and medium-sized workplaces appeared more interested than large workplaces. A short questionnaire was sent to those expressing an interest, asking about the type of business and size of workplace, and with a request to assess the number of employees living within 2 miles of the workplace by inserting relevant postcodes into a calculator on the Walkit.com website. A list of the first four digits of postcodes likely to contain relevant employees was also supplied, to reduce the number of postcodes to be checked. Nevertheless, this task appears to have been too onerous for some employers, and 19 workplaces returned the completed questionnaire, of which 17 continued through to DC1 (Table 8). These workplaces were willing to continue throughout the study, although in one large factory in the control arm (workplace 28) there was only one participant, who retired before the 1-year follow-up data collection.

FIGURE 1.

Consolidated Standards of Reporting Trials: Walk to Work flow of workplaces and participants. a, Target participants were those living < 2 miles from the workplace who did not walk or cycle to work.

Participant recruitment and retention

Characteristics of the 187 participants recruited are summarised in Table 9 and the flow of workplaces and participants through the study is described in Figure 1. There was a balance of males and females and an overall mean age of 37.8 years, with a range from 17.3 to 67 years. The participants were predominantly white (77%), well educated (with 60% having a degree or higher qualification) and employed in sedentary (desk-based) occupations.

Walk to Work promoters’ recruitment and training

Walk to Work promoters were recruited in each of the workplaces randomised to the intervention arm (Table 10). Of the 10 Walk to Work promoters trained, six were female and four were male.

A single half-day group training event for all Walk to Work promoters had been planned. However, it did not prove possible to timetable a training event to suit all workplaces. Consequently, two smaller training events were organised at the University of Bristol, and three visits were made to individual workplaces for on-site training.

At the training, the Walk to Work promoters received general information about the benefits of walking to work and were taken through the stages of a Walk to Work promoters’ booklet outlining their role and the BCTs to be used with participants during each of the four contacts over the 10-week intervention period. The session ended with further information about publicly available resources and websites that could be also be accessed and used to promote walking.

Qualitative data from the post-intervention (DC2) interviews suggest that the promoters found the training day useful: ‘I thought it was useful, I mean a lot of it was sort of stuff I knew already but it all sort of brought it together really’ (male, aged 30, workplace 25); ‘It was good, it was really informative that day’ (male, aged 29, workplace 20). One promoter (female, aged 45, workplace 27) commented that ‘It was good to work with other people from other companies’. Overall, the promoters found that their booklets were ‘well set out’ (male, aged 38, workplace 18) and gave clear guidance on the role: ‘It’s not too intimidating, it’s quite accessible, it’s got a, sort of a structure you can work through, when the Walk to Work promoter meets with the employees to discuss and set goals and review them’ (male, aged 30, workplace 25). However, one promoter said that he was ‘disappointed with myself getting out of sync and not following it properly because it guides you well but I lost sight of which week I was on and what I was supposed to be achieving’ (male, aged 38, workplace 18). Another suggested that more support and encouragement for the promoters would have been helpful, for example ‘each week, like, e-mail a newsletter just to remind you about the entire thing and kind of give you more information to convince you to keep on doing it’ (male, aged 29, workplace 20).

Assessment of costs

The contribution of walking to work to physical activity levels

Analyses of the baseline data highlight the considerable contribution that walking to work can make to adult physical activity levels. 49 The initial sample comprised 147 adults who lived within 2 miles of their workplace. Seventy-four participants (50.3% of the original sample) were categorised as walkers and 33 (22.4%) as car users. Five of these participants did not provide any valid accelerometer data, and were excluded from analyses. The final sample comprised 102 adults (mean age 36.3 ± 11.6 years; 56.9% female). Participants recorded 236 return journeys by foot and 86 by car. Mean self-reported journey time to work (each way) was 19.7 ± 8.3 minutes by foot and 10.7 ± 7.6 minutes by car.

Participants wore the accelerometer for a mean of 723.1 ± 155.2 minutes each day, with no difference between travel modes or sexes. Physical activity did not differ between males and females (497.9 ± 206.0 vs. 460.5 ± 146.1 cpm respectively; p = 0.285) and therefore the sexes were analysed together. When analysed by main travel mode (Table 20), participants who walked to work had higher levels of overall weekday physical activity than those who travelled by car, and also recorded more minutes of MVPA, but there was little difference in sedentary time.

TABLE 20 - Weekday and weekend physical activity by usual travel mode to work on weekdays (mean ± SD)

	All (n = 102)	Walk (n = 69)	Car (n = 33)	p-value
Weekday
Overall daily physical activity (accelerometer cpm)	476.6 ± 174.6	530.2 ± 165.1	364.6 ± 138.4	<0.001
MVPA (minutes/day)	69.5 ± 27.9	78.9 ± 24.0	49.8 ± 25.2	<0.001
Sedentary time (minutes/day)	586.2 ± 71.9	579.9 ± 76.4	599.2 ± 60.5	0.206
Weekend
Overall daily physical activity (accelerometer cpm)	418.2 ± 192.9	434.5 ± 207.3	385.1 ± 158.6	0.329
MVPA (minutes/day)	53.7 ± 30.0	55.5 ± 31.2	50.2 ± 27.7	0.505
Sedentary time (minutes/day)	510.4 ± 90.1	511.1 ± 91.1	509.0 ± 90.0	0.928

For those who walked to work, physical activity and MVPA values were substantially reduced at the weekend. The physical activity of car users was essentially unchanged so that there was no difference in weekend physical activity between the two travel groups. In linear regression models, walking to work was associated with higher overall weekday physical activity (B = 164.6 cpm, 95% CI 98.8 to 230.5; p < 0.001) and MVPA (B = 28.5 minutes, 95% CI 18.3 to 38.8; p < 0.001), but not with a difference in sedentary time.

Age and sex were not predictors of daily physical activity, MVPA or sedentary time in these models. Education was a predictor of total physical activity and MVPA, with those educated to degree level or above more active than those less well qualified (overall physical activity: B = 81.6 cpm, 95% CI 14.0 to 149.3; p = 0.019; MVPA: B = 12.3 minutes, 95% CI 1.8 to 22.8; p = 0.022).

Mean hourly physical activity was plotted to identify when differences in physical activity occurred during the day. Figure 2 shows that the main differences between car users and walkers occurred in the morning and late afternoon, potentially when commuting to or from work. There was no difference in physical activity between the travel groups during the main working hours (9 a.m. to 4 p.m.) (walk 347.2 ± 187.4 cpm vs. car 318.7 ± 194.4 cpm; p = 0.480).

FIGURE 2.

Mean hourly physical activity by mode of travel to/from work on weekdays.

To explore the contribution of walking to work to total physical activity, accGPS traces recorded between 6.00 a.m. and 10.00 a.m., and between 4.00 p.m. and 8.00 p.m., were examined. Of the 74 participants who walked to work, 58 recorded GPS data for at least one journey. Overall, 321 journeys (182 to work and 139 home from work, from 54 participants) were recorded. Participants generally took broadly the same route to and from work each day. A high level of physical activity was recorded on the journeys to (4260.7 ± 943.5 cpm) and from (3806.3 ± 915.8 cpm) work. Time spent on the journey to work was less than on the journey home (21.9 ± 7.8 vs. 28.6 ± 18.5 minutes), as in some cases participants interrupted their journey to visit shops on the way home (Figure 3). These visits were considered to be part of the journey. However, the minutes of MVPA on the journeys to and from work were highly similar (19.8 ± 7.1 vs. 21.0 ± 8.9 minutes of MVPA), as time spent in shops was not MVPA. A comparison of total MVPA (6.00 a.m. to midnight) with MVPA recorded during the journey for the 58 participants providing any GPS data showed that the walk to and from work contributed 47.3% of participants’ total daily MVPA (38.0 of 80.3 minutes).

FIGURE 3.

Global positioning system trace of participant walking directly to work and returning by an alternative route (precise start and end are cropped to protect confidentiality).

Participants’ views of using behaviour change techniques to promote walking to work

To examine participants’ views of the BCTs used, 22 interview transcripts were analysed. 64 The characteristics of the interviewees, who included Walk to Work promoters and intervention participants, are detailed in Table 21. Fourteen interviewees were employees who attempted to change their travel behaviour, and eight were Walk to Work promoters (of whom four already walked to work and four were attempting to change their behaviour). The age range was 22 to 65 years, with a range of household incomes, and 55% were male. Most were city-centre based and all were in sedentary occupations. Participants’ IDs reflect the data collection (DC2), workplace ID and participant number.

Employers’ views about encouraging employees to walk to work

Interviews were conducted with employers/managers at various stages of the study (Table 22). 65 Participant ID numbers reflect the stage at which the interview took place: during the phase 1 development work (P1); during the exploratory trial at baseline (DC1) or post intervention (DC2); and at the end of the intervention period with employers who had expressed an interest originally but did not continue with the study (prefixed with the digit 9). These are followed by the numerical workplace ID and participant ID.

Five employers/managers were interviewed during the development phase of the study. During the phase 2 exploratory trial, 12 interviews were conducted with employers/managers at baseline and a further seven (with eight participants, as one was a paired interview) immediately post intervention. An additional six interviews (with eight participants) were conducted with employers/managers who expressed an interest at the workplace recruitment stage but did not go on to participate in the study.

The interviews were conducted in a range of small, medium and large workplaces in both city centre and suburban environments. Two-thirds of the interviewees were female. In this section of the report, we concentrate on employers’/managers’ perceptions of the benefits and disadvantages of walking to work, and whether or not employers can or should encourage their employees to walk to work.

Lessons learned from the feasibility study

The feasibility study was important for developing the intervention and testing the methods for its implementation and evaluation. In particular, several modifications were suggested for a full-scale trial. The workplace recruitment process should be simplified so that employers are not required to calculate the number of employees who live within 2 miles of the workplace before being recruited to the study. This proved too onerous for some workplaces, which did not sign up for the study after an initial expression of interest. The calculation of distances is valuable to assess how many participants are within ‘walking distance’ and might be encouraged to walk the full distance, and how many others might be encouraged to use mixed mode. However, the study researchers or the Walk to Work promoters could calculate these distances after obtaining consent during DC1 and encourage the Walk to Work promoters to tailor their advice to suit the circumstances.

All employees who do not already always walk or cycle to work, and are not disabled in relation to walking, should be eligible to participate, rather than the initial focus being solely on those living within 2 miles of the workplace. The Walk to Work promoters’ booklets, and the participants’ booklets, will give information and advice in relation to walking the full distance and about combining walking with other modes of transport.

Workplaces with staff on predominantly part-time, zero-hour or short-term contracts are not suitable because of the need for a 1-year follow-up data collection. The attrition level (29% at 1-year follow-up) compared well with other workplace-based interventions. No attempt was made to contact those who had left the workplace. For any future trial, asking for the contact details of those who leave the workplace between baseline and follow-up would enable data collection packs to be forwarded to them, and returned to the researchers, by post.

Flexibility with regard to the training and support of Walk to Work promoters is required. During the feasibility study, it was proposed that the Walk to Work promoters would be invited to one external training event. However, it became clear that not all of the workplaces would be able to release an employee in this way on the same date. Two smaller external training sessions were organised and some individual workplace-based tuition was provided. A choice should, therefore, be offered. Some prompts and encouragement for the Walk to Work promoters themselves during the 10-week intervention period were also suggested during the post-intervention interviews.

Assessment of costs

The intervention was developed and delivered at an affordable average cost per workplace (£441.05). However, the actual cost per workplace varied substantially depending on the number of promoters trained and the location of the training. There were no significant differences in health service or commuting costs at 1-year follow-up. Self-reported measures of productivity were also similar between the intervention and control groups at follow-up.

Despite studies demonstrating the health benefits of active commuting, assessments of the cost-effectiveness of these interventions are relatively scarce. Where economic evaluations have been performed, cost-effectiveness analyses have been conducted66 and benefit-to-cost ratios have been calculated. 67 Our study provides preliminary information on the costs and benefits borne by employers, employees and others, which will be relevant to government bodies and employers when deciding whether or not to adopt such an intervention. The study has demonstrated that it is feasible to conduct an economic evaluation alongside a RCT of an employer-led scheme to increase walking to work.

Future research should incorporate a measure of the time spent by promoters and employees discussing the intervention, in order to calculate a more accurate cost of delivery. Although the 12-month follow-up is long enough to provide information on the medium-term effectiveness of the intervention with respect to commuter behaviour and health service use, it is also important to consider the implication of increased walking on longer-term health-related benefits. Future research should, therefore, include other measures of health in order to capture any long-term benefits. The Health Economic Assessment Tool (HEAT) developed in 2011 could be included, as this assesses resource savings and reduced mortality or increased life-years associated with an increased in time spent walking. 67

Promoting active travel through the workplace

The interviews with employers suggest that there is some uncertainty about whether encouraging change in commuting behaviour should be the responsibility of the employer, other policy-makers, or both. Furthermore, employers need to justify the costs to the organisation and balance the scheme’s activities with other business priorities. In some cases, employers were concerned that attempts to influence travel mode might be interpreted as a cost-cutting exercise or inappropriate interference with the lives of employees outside of the working environment, rather than as a sign of a caring employer.

Some employers were unclear about how to give more general support in the workplace for employees who walk to work. In addition, there was evidence that those who had been charged with promoting ‘green’ travel initiatives in the past could be jaded by the experience. This suggests that additional booklets for employers would be valuable. This booklet could include ideas for promoting walking, such as strategically placed posters and leaflets; providing information about walking distances to train and bus stops; providing lockers or improved cloakroom facilities; financial assistance for public transport season tickets or walking clothes/shoes; free incentive items for those who switch to walking, for example umbrellas, rucksacks or breakfast vouchers; and ideas for ‘competitions’ and challenges for those who enjoy taking part in such activities.

Strengths and limitations

The role of Walk to Work promoters

The Walk to Work promoters either were volunteers or were asked by their employer to take on the role, and showed varying degrees of interest in active travel. Although they all received training to understand and use the nine BCTs, this was also approached with varying levels of enthusiasm: most appearing keen to help and generally encouraging, with one perhaps being overenthusiastic and another struggling to fulfil the role because of other work commitments. Where one promoter was of a ‘lower grade’ than the colleague they were supporting to walk to work, there was evidence of uncertainty about how to deal with this ‘role reversal’.

Similarly, participants also varied in their appreciation of the support provided by the promoters, from a feeling that it was too ‘light touch’ to concern that they were being pressurised into meeting unrealistic expectations. This would suggest that training should include awareness of how to assess and respond to the varying needs of participants. The availability of a range of approaches, including face-to-face contact with a promoter, guidance booklets, self-monitoring equipment and interactive websites, may help to ensure that the different preferences of participants can be catered for.

An evidence review undertaken on behalf of NICE identified a number of characteristics and competencies for those delivering BCT interventions, which included being supportive, motivating and empathetic, as well as having knowledge and communication skills associated with assessing individuals, signposting support, developing motivation and action, providing feedback, action planning, goal setting and problem solving, encouraging self-management, group counselling and maintaining change and relapse prevention. 38 It is unrealistic to expect volunteers in the workplace to have all of these characteristics and competencies, together with the time and space to implement them. For the Walk to Work intervention, the promoters were provided with a half-day training session supported by booklets and signposts to websites and other useful resources. The appropriate use of these resources, combined with a supportive and empathetic approach, may be an appropriate compromise for a workplace public health intervention of this kind. Nevertheless, workplace volunteers who undertake this role may require some additional training to develop skills and competencies, or external support from those with greater knowledge and experience of health promotion interventions. 37,81

The socioecological model

Advocates of the socioecological model argue for the importance of examining influences at the policy, community, organisational, interpersonal and intrapersonal levels. 38 BCTs are primarily focused on the intrapersonal and interpersonal levels, but participants in the Walk to Work study also highlighted the importance of support at the other levels of the model, particularly the workplace (organisational level), through, for example, employers providing free breakfast for walkers or improved facilities for washing and changing. The availability of car parking was identified as an important barrier, a finding echoed by others who suggest that limiting workplace parking may encourage walking to work. 82–84 For example, a workplace travel plan at the University of Bristol which included controlling the supply and cost of car parking spaces for staff resulted in a self-reported increase in walking to work from 19% to 30% and a reduction in commuting by car from 50% to 33%. 85 In the current study, environmental barriers such as unpleasant or potentially unsafe routes were mentioned but were not as prominent as in other studies. 82–84 This may support the view of Guell et al. ,82 who point out that some participants will walk despite adverse environmental conditions, having overcome the issue through experience or weighing up the perceived benefits and costs.

It is also important to acknowledge the contextual issues which may undermine behaviour change interventions. In the case of the Walk to Work study, it appeared that the legacy of the global financial crisis meant that some employers and employees were unable to prioritise the intervention over other work commitments. Furthermore, the extraordinarily wet weather was identified as an important barrier to walking by participants. However, even within this context, there was evidence of willingness to consider the BCTs and change travel behaviour.

The benefits of walking to work

There is compelling evidence that regular physical activity is effective in the prevention of chronic diseases (including cardiovascular disease, type 2 diabetes, some cancers, hypertension, obesity, depression and osteoporosis) and premature death, with the greatest improvements in health status seen when people who are least active become physically active. 86 Active commuting (walking and cycling) has specifically been associated with an overall reduction in cardiovascular disease. 87 Consequently, those who shift from car use to regular walking during the commute to work are likely to benefit from improvements to health. Other benefits include reduced commuting costs and greater certainty about the timing of the journey. The suggested benefits to employers of promoting walking schemes include increased productivity, a reduction in sick leave, improved public image as a result of lowering the workplace’s carbon footprint, and savings in providing car parking facilities. 27,83,87–89

Public Health England recently announced several ‘high-level enduring priorities’ guiding their work, two of which are relevant to this study: helping people to live longer and more healthy lives by reducing preventable deaths, and improving health in the workplace by encouraging employers to support their staff, and those moving into and out of the workforce, to lead healthier lives. 90 Similar aims are shared by governments and health practitioners throughout the world. Encouraging employees to walk to work has the potential to make a considerable contribution to addressing these priorities.

Conclusion and recommendations

The aim of the Walk to Work study was to test the feasibility of developing and evaluating an employer-led intervention to increase walking to work. Consequently, the implications of this study do not, at this stage, relate to practice but to the feasibility and design of a follow-on full-scale RCT.

Contributions of authors

Dr Suzanne Audrey (Research Fellow) was the chief investigator for the study. She conceived and designed the study, supervised its conduct, led on the analysis of employers’ views and led the writing of the report.

Dr Sunita Procter (Research Associate) was the study manager, contributed to the design of the intervention, led the analysis of the data relating to the use of behaviour change techniques and contributed to the writing of the report.

Professor Ashley Cooper (Professor of Physical Activity and Public Health) was a co-applicant, led on the measurement of physical activity and contributed to the sections of the report relating to physical activity outcomes.

Professor Nanette Mutrie (Chair in Physical Activity for Health) was a co-applicant and contributed to the design of the intervention and the section of the report relating to the use of behaviour change techniques.

Professor William Hollingworth (Professor of Health Economics) was a co-applicant, led on the assessment of costs and contributed to the section of the report relating to the assessment of costs.

Dr Adrian Davis (Public Health and Transport Consultant) was a co-applicant and contributed to the design of the intervention and the section of the report relating to the use of behaviour change techniques.

Dr Ruth Kipping (Research Fellow and NHS Consultant in Public Health) was a co-applicant and contributed to the study design.

Mr Philip Insall (Health Director at Sustrans) was a co-applicant and contributed a lay perspective throughout the study.

Ms Kirsty Garfield (Research Assistant in Health Economics) undertook analyses and contributed to the section of the report relating to the assessment of costs.

Professor Rona Campbell (Professor of Health Services Research) advised on the study design and conduct.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the PHR programme or the Department of Health. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the PHR programme or the Department of Health.

Publications

Audrey S, Procter S, Cooper AR. The contribution of walking to work to adult physical activity levels: a cross sectional study. Int J Behav Nutr Phys Act 2014;11:37.

Procter S, Mutrie N, Davis A, Audrey S. Participants’ views of using behavior change techniques to encourage walking to work: a qualitative study. BMC Public Health 2014;14:868.

Audrey S, Procter S. Employers’ views of promoting walking to work: a qualitative study. Int J Behav Nutr Phys Act 2015;12:12.