Appendix C The evidence

Introduction

This appendix lists the evidence statements and links them to the relevant recommendations. (See appendix B for the key to quality assessments.) Note: the evidence statements in review 1 (see appendix A) were amended by NICE and endorsed by the Programme Development Group (PDG). This appendix includes the amended evidence statements from review 1.

Appendix C also lists six expert reports and their links to the recommendations and sets out a brief summary of findings from the economic analysis.

The evidence statements are short summaries of evidence, in a review, report or paper (provided by an expert in the topic area). Each statement has a short code indicating which document the evidence has come from. The letter(s) in the code refer to the type of document the statement is from, and the numbers refer to the document number, and the number of the evidence statement in the document.

Evidence statement number R1.ES1 indicates that the linked statement is numbered 1 in the document 'Evidence statements on the effectiveness of local interventions to promote cycling and walking for recreational and travel purposes'. Evidence statement numbered R2.ES1 indicates that the linked statement is numbered 1 in the document 'Synthesis of evidence relating to barriers and facilitators to implementing interventions that promote cycling and walking, and to carrying out cycling and walking for recreational and travel purposes'. Evidence statement EM.ES1 indicates that the linked statement is numbered 1 in 'Interventions to promote cycling and walking for recreational and travel purposes: Health economic and modelling report'

The reviews, expert reports and economic analysis are available online. Where a recommendation is not directly taken from the evidence statements, but is inferred from the evidence, this is indicated by IDE (inference derived from the evidence).

Where the PDG has considered other evidence, it is linked to the appropriate recommendation below. It is also listed in the additional evidence section of this appendix.

Recommendation 1: IDE; Additional evidence expert papers 2, 4, 6

Recommendation 2: IDE; Additional evidence expert papers 2, 4, 6

Recommendation 3: Evidence statements R1.ES5, R1.ES6, R1.ES7; Additional evidence expert papers 2, 4, 6

Recommendation 4: Evidence statements R1.ES4, EM.ES4

Recommendation 5: Evidence statements R1.ES3, R1.ES5, R1.ES6, R1.ES7, R1.ES9, R1.ES12, R1.ES19, R2.ES9, R2.ES15, R2.ES18, EM.ES3, EM.ES5; Additional evidence expert papers 2, 3, 4, 5, 6

Recommendation 6: Evidence statements R1.ES1, R1.ES2, R1.ES7, R1.ES13, R1.ES18, R1.ES21, R1.ES22, R2.ES1, R2.ES2, R2.ES3, R2.ES5, R2.ES6, R2.ES10, R2.ES12, R2.ES13, EM.ES1, EM.ES3; Additional evidence expert papers 1, 5

Recommendation 7: Evidence statements R1.ES13, R1.ES14, R1.ES18, R1.ES21, R1.ES22, R2.ES3, R2.ES13, EM.ES2; Additional evidence expert paper 5

Recommendation 8: Evidence statements R1.ES8, R1.ES9, R1.ES10a, R1.ES10b, R1.ES10c, R2.ES15, R2.ES16, EM.ES1; Additional evidence expert paper 1

Recommendation 9: Evidence statements R1.ES11, R1.ES15, R1.ES16, R1.ES17, R1.ES23, R2.ES2, R2.ES4, R2.ES7, R2.ES9, R2.ES18; Additional evidence expert papers 1, 3

Recommendation 10: Evidence statements R1.ES20, R2.ES2, R2.ES4

Evidence statements

Please note that the wording of some evidence statements has been altered slightly from those in the evidence review(s) to make them more consistent with each other and NICE's standard house style.

Evidence statement R1.ES1: Population-level change in mass-media interventions to increase walking

There was inconsistent evidence from two studies1,2 (both [+]) on the effectiveness of mass-media interventions (which included paid advertisements [TV, radio, cable, newspapers], billboards/posters, public relations, educational activities and community participation), delivered in the community in increasing population levels of walking for leisure or travel in adults up to 1 year post intervention. One before-and-after (BA)1 study showed no effect on walking (the reporting of data in this study was poor) and one cross-sectional (CS)2 study showed a small, but positive effect on walking.

One (+) BA study1 (UK n=3476, 12 months) – 40-second TV advert supported by a telephone helpline – showed no change in the number of days spent walking for at least 30 minutes: mean of 4.26 days in 1995 and 4.13 days in 1996, no significance statistics given.

One (+) CS study2 (USA n=297, 5 months) – billboard, newspaper, radio, and poster advertisements – showed that those exposed to the campaign were more likely to walk for at least 10 minutes on more days of the week than the control group (5.2 days versus 4.52 days t[7]=2.34, p=0.02).

Population-level evidence on mass-media interventions to increase walking is partially applicable to the UK as one study was conducted in the UK. The differing environment in the USA must be considered in reference to the studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Wimbush (1998)

2 Wray (2005)

Evidence statement R1.ES2: Multi-component community-based interventions to promote walking

There was inconsistent evidence from six studies concerning the effectiveness of multi-component interventions on increasing population levels of walking for leisure or travel in the long term. Four non-randomised control studies (nRCT)1,2,3,4 papers (three [+] and one [++]) showed positive effects on walking and two nRCT5,6papers (one [++] and one [+]) indicated that the interventions were not effective in increasing walking.

One (+) nRCT1 (Australia n=two wards, 2 years) – park modifications, media campaign, walking maps – showed that those in the intervention ward were more likely than those in the control ward to have walked in the 2 weeks prior to follow up (89.3% versus 81.0% respectively; X2=11.51, p=0.001), and within-ward analysis indicated that walking increased from baseline in the intervention ward (X2[1]=5.85, p=0.016), but not in the control ward (X2[1]=0.07, p=0.794). There was no difference in the number reaching adequate levels of physical activity (health department recommendations).

One (++) nRCT5 (USA n=1233, 12 months) – individually tailored newsletters, interpersonal activities that stressed social support, community-wide events such as walk-a-thons – showed that rates of 7-day walking for any purpose or for exercise declined slightly in the intervention communities compared with the comparison sites (-1.4 min, p=0.91; and -5.6, p=0.37 respectively).

One (+) nRCT6 (USA n=1531, 12 months) as above found that the change in walking was higher in intervention (11.7 minutes) than comparison (6.5 minutes), although not statistically significant. Percentage of respondents who met the recommendation for walking was the same across the intervention and comparison areas: 22.2% and 21.6%, p=0.811.

One (+) nRCT2 (USA n=1472, 8 weeks) – paid advertising, public relations events to generate media coverage, public health educational activities at work sites, churches and local organisations – found a 23% increase in walking observations in the intervention community versus a 6% decrease in the comparison community (OR 1.31, 95% CI 1.14–1.50; p<0.0001).

One (++) nRCT3 (USA n=1472, 12 months) – paid advertisements (TV, radio, cable, newspapers), public relations and community participation – found that the least active group in the intervention population were more likely than control population to have increased daily walking (OR=1.72, 95%CI 1.01–2.95).

One (+) nRCT4 (USA n=4 communities, 8 weeks) – four interventions: Welch Walks (WW): paid media, media relations, community activities; Broome County (BC) walks (BC): WW components + website; Wheeling walks and West Virginia (WV) walks: BC components +12-week participatory planning, policy and environmental changes – found that 32% of insufficiently active persons in Wheeling Walks reported meeting the criteria for regular walking immediately post campaign compared to an 18% increase in the comparator community (OR=2.12, 95%CI 1.41–2.24). An increase in reaching regular walking was observed for the most sedentary group in WV walks (p<0.05). The intervention community in Welch walks demonstrated a twofold (OR=2.0 95%CI 1.01–3.97) gain in weekly walking by at least 30 minutes versus the comparison community. Forty one per cent of the BC walks intervention community increased walking by 30 minutes per week compared to 30% in the control (OR=1.56 95%CI 1.07–2.28).

The population-level evidence on multi-component interventions to increase walking is only partially applicable to the UK as studies were conducted in the US and Australia. The differing environment in the USA must be considered in reference to the studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 New South Wales Health Department (2002)

2 Reger (2002)

3 Reger-Nash (2005)

4 Reger-Nash (2006)

5 Brownson (2004)

6 Brownson (2005)

Evidence statement R1.ES3: Population-level change in mass-media interventions to increase walking and cycling – Australia 'Walk to work day'

Moderate evidence from one BA study reported in two papers1,2 (both [+]) suggests that the mass-media campaign 'Australia walk to work day' (a collaborative annual event in which members of the public are encouraged to walk [or cycle] to work) may be effective in increasing population levels of walking and cycling for travel in adults up to 1 year post intervention. This intervention resulted in positive effects on both walking and cycling.

One(+)1 study (n=1100, at least 1 year) found that overall, total weekly minutes of moderate physical activity increased by 20 minutes per week (t[1087]=4.76, p<0.005 with, an decrease in the proportion who were inactive -4.0% p<0.005). Significant population increase in total walk time (+16minutes per week t[780]=2.04, p<0.05) in participants who were employed, and in minutes spent walking increased by 21 minutes per week in 'passive commuters' (t[535] = 2.42, p< 0.05).

One (+)2 study (n=1100, 2 months) found a significant population-level increase in health enhancing active commuting (3.9%, p=0.01).

The evidence on mass-media interventions to increase walking and cycling is only partially applicable to the UK as studies were conducted in Australia. The differing environment in Australia must be considered in reference to these studies. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Merom (2005)

2 Merom (2008)

Evidence statement R1.ES4: Population-level change in TravelSmart as an intervention to increase walking and cycling

Weak evidence from a series of evaluation reports (ER)1,2( both [+]) suggests that TravelSmart is effective in increasing population levels of walking and cycling for travel in adults (who volunteered to participate) at least over 1 year. TravelSmart uses 'Individualised travel marketing' (ITM) which aims to highlight travel choices 'people may not know they have' by providing locally relevant information and support to households. The evidence is moderate as the reports only present percentage change data and limited methodologies. The intervention targets individuals, but data is reported at population level.

One (+) evaluation report1 (Australia n=5 regions, various) found household projects routinely showed decreases in car use of 4–15% and rise in use of walking, cycling and public transport.

One (+) evaluation report2 (UK n=19 regions, various) found cycling for travel increased by between 14% and 69%, travel by walking increased between 9% and 29%, travel by car decreased at each site by between 10 and 14%, overall sustainable travel trips increased at each site (between 9% and 29%).

The evidence on this intervention to increase walking and cycling is fully applicable to the UK as most of the data reported is from UK sites. However, the differing environment in Australia must be considered in reference to the data collected there. Individual local contexts as well as the setting will also impact on the applicability of data from individual sites.

1 TravelSmart (2006)

2 TravelSmart (2011)

Evidence statement R1.ES5: Population-level change in cycle demonstration towns as interventions to increase cycling

There is moderate evidence indicating that cycling demonstration towns (CDT) (multi-component interventions to increase cycling in six towns) are effective in increasing population levels of cycling for active travel in the general population up to 10 years post intervention. One (-) ER1, one (+) BA2 and one (+) interrupted time series (ITS) study3 showed positive effects on cycling in cycle demonstration towns, although the significance of the effects is not reported. See also R1.ES7 and R1.ES6.

One (+) ITS3 (UK n= six towns, 4 years) found automatic counter data indicated an average increase in cycles counted of 27%. Proportion of pupils cycling to school at least once a week increased from 12% pre-survey to 26% post-survey.

One (-) ER1 (UK n=6 towns, 10 years) found data from automatic cycle counts indicated a 12% increase overall in usage of cycle routes and up to 60% at specific sites (this report also uses data from other interventions).

One (+) BA2 (UK n=1500, 4 years) found the proportion of adult cycling for at least 30 minutes once or more per month increased from 11.8% in 2006 to 15.1% in 2008, an increase of 3.3%-points or 28%.

The evidence on cycle demonstration town is directly applicable as it was conducted in the UK.

1 Cope (2011)

2 Sloman (2009)

3 Cope (2009)

Evidence statement R1.ES6: Population-level change in multi-component interventions to increase cycling

Weak evidence from one (+) nRCT1 study suggests that multi-component interventions are not effective in increasing population levels of cycling in the general population up to 2 years post intervention, but may result in increased use of bicycle paths and increase in cycling among new/beginner cyclists. See also R1.ES5.

One (+) nRCT1 (n=909, 2 years) – multi-component community-based intervention including: organised bike rides and events, cycling skills courses, distribution of cycling maps of the area, local press coverage – found significantly greater use of the bicycle paths in the intervention area (28.3%) at follow-up compared with the comparison area (16.2%) p<0.001. No self-reported increase in residents who said they cycled in the last year, however, significantly more 'novice'/beginner riders had cycled in the last year in the intervention area (11.5% versus1.4% in the comparison area; p=0.013).

The population-level evidence on multi-component interventions to increase cycling is only partially applicable to the UK as the study was conducted in Australia. The differing environment in Australia must be considered in all studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Rissel (2010)

Evidence statement R1.ES7: Population-level change in multi-component interventions to increase walking and cycling in adults

Weak evidence from four (two [+], two [-]) of five1,2,3,4,5 studies indicates that multi-component interventions delivered in the community are effective in increasing population levels of walking and cycling for travel and/or leisure up to 9 years post-intervention. Evidence from the three BA1,2,3, and one ITS4, showed mostly positive effects of community interventions to encourage cycling and walking for travel and/or leisure. One (+) nRCT5 indicated that multi-component interventions may reduce a natural decline in walking in women and that among those with a low educational level, cycling may show a small increase. See also R1.ES5 and R1.ES6.

One (+) BA1 (Belgium n=438, 1 year) – physical activity promoted in the entire city of Ghent. Central theme of '10,000 steps/day', with secondary taglines of 'every step counts') and 'every revolution (of bicycle pedals) counts', pedometers given – found that 47.5% increased average step counts by 896 steps/day or more at 1-year follow-up (no statistical analysis; cycling was 'converted' to step counts).

One (-) BA2 (USA n=not reported, 12 months) – multi-component intervention to increase safe physical activity opportunities and encourage walking and biking for short trips – found the number of people seen using active transportation increased from 1028 in 2005 to 1853 in 2006 (63% increase). Walking to school more than doubled at three of four schools engaged for at least 2 years (no other analysis).

One (+) BA3 (UK n=at least 12,000, 4 years) – three 'Sustainable travel towns' which implemented intensive town-wide Smarter Choice Programmes to encourage use of non-car options; bus use, cycling and walking, and less single occupancy cars – found that cycle trips per head grew substantially in all three towns by 26–30%. Comparison towns cycle trips decreased. Walking trips per head grew substantially by 10–13% compared to a national decline in similar towns.

One (-) ITS4 (USA n=not reported, 1 year) – Project U-Turn, active transportation (biking, walking, and transit use) through an integrated approach to active living, ran for 5 years, targeting 36,000. City-wide count of people using active transport, showed an increase of 63% over 1 year, limited study details provided. Also had a major schools component and reported an increase in walking over time, no statistics given.

One (+) nRCT5 (Netherlands n=3114, 5 years) – community-based project with 790 lifestyle interventions, 361 were physical activity focused, example: printed guides of walking and cycling routes – found that there was a smaller decline in walking in women in the intervention compared to control region (-0.3 hours/week versus -2.3 hours/week; p≤0.05); and among those with a low education level there was a significant difference in change in cycling and walking in the intervention versus control region (0.2 hours/week versus -0.3 hours week respectively for cycling and 0.0 hours/week versus -2.2 hours week for walking; both p≤0.05).

The population-level evidence on multi-component interventions to increase walking and cycling in adults is only partially applicable to the UK as one studies was conducted in the UK. The differing environment in the USA and Europe must be considered in reference to the studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 De Cocker (2009)

2 Hendricks (2009)

3 Sloman (2010)

4 TenBrink (2009)

5 Wendel-Vos (2009)

Evidence statement R1.ES8: Population-level change in multi-component interventions to increase walking and cycling in children

Inconsistent evidence from three studies1,2,3 on the effectiveness of school-based multi-component interventions to increase levels of walking and cycling for children. Evidence from two (+) BA studies1,2 showed positive effects on school population-level walking in children however evidence from one (++) cluster randomised control study (RCT)3 showed no effect on cycling and walking for school travel.

One (+) BA1 (UK n=179, 41 months) – school travel plan group developed a walking bus scheme, incentive scheme 'going for gold' included children cycling or scooting to school, also cycle training, pedestrian training, park and walk scheme, curriculum work, school assemblies and newsletters – found walking to school increased from 30% to 58.8%, cycling to school increased from 0– 4%.

One (++) cluster RCT3 (UK n=21 schools, 12 months) – multi-component school travel plans were developed by a school travel coordinator – found the proportion of children walking or cycling to school was not affected by the intervention.

One (+) BA2 (UK n=11 schools, up to approximately 18 months) – 'Safe routes to school'– identified and created safe routes to school, invites community-wide involvement, full-time educator employed to develop curriculum and volunteer team leader in each school – found an increase in number of school trips made by walking (64%) and biking (114%).

The population-level evidence on multi-component interventions to increase walking and cycling in children is applicable to the UK as all studies were conducted in the UK. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Cairns (2006a)

2 Staunton (2003)

3 Rowland (2003)

Evidence statement R1.ES9: School-based change in interventions to increase cycling in children

Weak evidence from one (+) BA study1 suggests that school-based multi-component interventions may be effective in increasing school population levels of cycling in children. Evidence showed positive effects on walking at the school population level.

The study1 (UK n=52 schools, 1 year) – 'Bike it': school travel plans, cycling champions in schools to demonstrate to parents and pupils that cycling is a popular choice. Percentage of school pupils cycling to school every day increased from 3% to 10%. Number of pupils cycling at least once a week increased from 10% to 27%. Number of pupils who never cycled decreased from 80% to 55%.

The evidence on multi-component interventions to increase cycling in children is applicable in the UK as the study was carried out in the UK.

1 Sustrans (2008)

Evidence statement R1.ES10a: Walking school bus interventions to increase walking

Moderate evidence from three (+) BA studies1,2,3and one (+) nRCT4 suggests that walking school bus interventions may be effective in increasing levels of walking at the school population level for children up to 30 months post-intervention.

One (+) BA1 (UK n=309, 14 months) – walking school buses supported by environmental interventions such as street lighting on walking routes – found that participants walking increased from 60% to 68.3%, 25% of that was due to walking buses.

One (+) nRCT2 (USA n=3 primary schools, follow up 6 months after baseline) – Walking School Bus (WSB). The school implemented three routes staffed by parent volunteers, and were compared to two nearby schools without a WSB – found that the number of children who walked to school increased from baseline to follow up by 25% (from 19–26%). Comparison schools showed a decrease in the proportion of children walking to school over the same period (no data given).

One (+) BA3 (UK n=64, 18–30 months) – walking buses at five schools. Information sent home to parents to encourage participation – found that there was an overall average increase of 513 metres walked per day. For children that had previously walked to school the WSB resulted in an average increase of only 19 metres/day, for those that previously travelled to school by a mixture of car and walking: average increase of 309 metres/day and for those that previously regularly travelled by car to get to school: average increase of 1549 metres/day (no statistical analyses reported). Participation in the walking buses declined over time.

One (+) nRCT4 (USA n=643, 12 months) – WSB run by a part-time coordinator and parent volunteers. The intervention included three routes which ranged from 0.3 to 1.5 miles and took 15–40 minutes. The WSB operated once or twice a week – found that higher proportions of students walked to the intervention (25% +/- 2%) versus the control schools (7% +/-1%: p<0.001). Significant increase in walking to school in intervention school from 20% (+/-2%) at baseline.

The evidence on school-based walking sessions to increase walking is partially applicable to the UK as two studies were conducted in the UK. The differing environment in the USA must be considered in reference to the studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Bickerstaff (2000)

2 Johnston (2006)

3 Mackett (2005)

4 Mendoza (2009)

Evidence statement R1.ES10b: School-based interventions using pedometers to increase walking

Moderate evidence from one (+) cluster RCT1and one (+) ITS2 suggests that school-based walking interventions which incorporate pedometers may be effective in increasing levels of walking at the school population level for children up to 12 weeks post intervention.

One (+) ITS2 (USA n=169, 6 weeks) – pedometers and a 'Fit bits' programme to implement physical activity breaks in the classroom – mean steps increased from 19,149 (95%CI 18,224–20,073) week 1 to 21,248 (95%CI 19,730-22,765) week 6 (p<0.001) found that overall, walking peaked at week 3; and younger students had a stronger response to the intervention.

One (+) cluster RCT1 (New Zealand n=85, 12 weeks) – physical activity self-monitoring and educative programme – the pedometer (PED) group set daily step targets, and the minutes (MIN) group set daily time based activity goals – found that both intervention groups had significant increase in steps between baseline and week 12 (p<0.001), no significant differences between time points for the control group (p=0.23).

The evidence on school-based walking sessions to increase walking is only partially applicable to the UK as studies were undertaken in the USA and New Zealand. The differing environments in these countries must be considered in reference to the studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Schofield (2005)

2 Cirignano (2010)

Evidence statement R1.ES10c: School-based walking session interventions to increase walking

Inconsistent evidence from five studies (reported in six papers1,2,3,4,5,6) on the effectiveness of school-based walking session interventions in increasing levels of walking at the school population level for children up to 48 months post intervention. Evidence from one (+) nRCT1 and two (+) BA studies (reported in three papers2,3,4) showed positive effects on school population walking. However one (+) nRCT5 showed no effect on walking and one (+) cluster RCT6 had conflicting evidence concerning the intervention effect on walking for school travel.

One (+) BA2 (UK n=585, 48 months) –'Walk on Tuesday and Thursday' (WOTT) encouraged walking to school, included incentives – found that walking to school increased from 53.3% to 58.7% (percentages only reported). Also reported in a second (+) BA3.

One (+) nRCT1 (UK n=60, 10 weeks) – school-based active travel project. Active travel was integrated into the curriculum, and participants used interactive travel planning resources at home – found that mean distance travelled to school by walking increased significantly more in the intervention (389%) than the control (17%: t[38]=-4.679, p<0.001, 95% CI -315 to -795 m).

One (+) nRCT5 (UK n=13 schools, 4 weeks) – interventions linked to national walk to school week – found no difference between intervention and control schools in walking before or after the intervention.

One (+) cluster RCT6 (Australia n=24 schools, 2 months) – health promoting schools policy: classroom activities, pedometer-based walking activities (some schools) development of school travel access guides, parent newsletters, and improving environments with local councils – found that, based on student survey data while both intervention and control groups increased walking by about 4% from baseline, there was no statistically significant difference in mean percentages of change in mode of transport to or from school from baseline to follow-up between the intervention and control groups (no data given).But parent survey data (n=807) indicated a significant increase in walking trips by students in the intervention compared to control schools (28.8% versus 19%, p=0.05).

One (+) BA4 (Australia n=234, 4 weeks) – classroom activities supported by a weekly newsletter to encourage walking to school – found the percentage of walking trips increased by 3.4% and car trips decreased by 3.4%.

The evidence on school-based walking sessions to increase walking is partially applicable to the UK as three studies were conducted in the UK. The differing environments in Australia must be considered in reference to the studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 McKee (2007)

2 Cairns (2006b)

3 Cairns (2006c)

4 Zaccari (2003)

5 Tapestry (2003)

6 Wen (2008)

Evidence statement R1.ES11: Population-level change in workplace-based interventions to increase independent walking and cycling

Weak evidence from one (+) BA study1 and one (+) ITS2 indicates that multi-component interventions delivered in the workplace are effective in increasing population levels of walking and cycling

One (+) ITS1 (UK n=1850 to 2829 in each of four staff surveys, 9 years) – university transport plan: limiting the number of available parking spaces and permits, improving, installing secure cycle storage, subsidised cycle purchase scheme, car share scheme, free bus travel, and discounted season tickets – found that respondents who usually walked to work increased from 19 to 30% (Z=4.24, p<0.001) and regular cyclists increased from 7.0% to 11.8% (not significant).

One (+) BA2 (UK n=2240, 3 years) – Well@Work programmes which consisted of a diverse set of initiatives and actions aimed at promoting and supporting healthy lifestyles – found an increase of 9% in the proportion of employees participating in active travel (walking or cycling), significant increase in employees cycling (4%) or walking (8%) to work.

The population-level evidence on multi-component interventions to increase walking and cycling in adults is applicable to the UK as both studies were conducted in the UK. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Bull (2008)

2 Brockman (2011)

Evidence statement R1.ES12: Individual-level change from participation event to increase cycling

Weak evidence from one study suggests that a mass participation intervention may be effective in increasing individual-level cycling for leisure in adults. Evidence from one (+) BA study1 showed a positive effect on cycling 1 month after the intervention.

One (+) BA1 (Australia n=918, 2 months) – mass cycling event – found that participants with low pre-event self-reported cycling ability reported an average of four sessions of cycling in the month before the event and an average of 6.8 sessions in the month after the event (t=5.25, p<0.001).

The evidence on mass participation event intervention to increase cycling is only partially applicable to the UK as the study was conducted in Australia. The differing environment in Australia must be considered in reference to the studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Bowles (2006)

Evidence statement R1.ES13: Individual-level change in community delivered targeted health information interventions to increase walking

Moderate evidence from six studies1,2,3,4,5,6suggests that individual, targeted provision of health information (including printed media, telephone support and text messages) delivered in the community are effective in increasing individual levels of walking for leisure or travel in adults up to 1 year post intervention. Five (++) RCTs1,2,3,4,5 showed positive effects on walking. One further (++) RCT6 also showed positive effects on walking, but was designed to test intervention fidelity.

One (++) RCT1 (USA n=117, 3 months) – ten weekly emails containing links to a webpage with an interactive information tailoring tool to promote physical activity – found that walking increased at a faster rate in the intervention group than the control group (β=15.04 [SE=8.38], p=.035 [one-tailed]). Intervention group increased walking by 69 minutes/week versus 32 minutes/week in control.

One (++) RCT2 (Australia n=399, 10 weeks) – print only (participants were mailed self-help brochures weekly for 3 weeks) or print plus telephone (participants received the same print programme plus three weekly telephone support calls – found that both intervention groups significantly increased time reported walking for exercise per week (from 130 to 147 minutes: t[1,277] =-3.50, p<0.001; and from 132 to 150 minutes, t[1,106]= -2.44, p=0.016).

One (++) RCT3 (USA n=197, 6 months) – counselling weekly telephone calls to assess physical activity levels and problem solve how to fit adequate walking activity into their week – found that women in the intervention group reported more time walked each day than the control women (F [1,191]=4.10, p<0.05).

One (++) RCT4 (USA n=253, 12 months) – telephone calls with or without counselling, or a control video – found that women in the intervention group showed a linear increase in walking from baseline to 6 months (latent growth analysis to assess the relationship between time and intervention group membership).

One (++) RCT5 (UK n=149, 4 weeks) – two theory-based interventions consisting of forming 'implementation intentions' along with text message reminders to achieve walking-related plans or goals – found a differential change across groups in brisk walking or fast walking (F [2,130]= 3.12, p=0.048). Two intervention groups which differed in having a plan reminder or goal reminder had a 45% and 42% increase of at least 2 days a week meeting physical activity daily guidelines respectively, with a 22% increase in the control group.

One (++) RCT6 (USA n=50, 12 months) – two interventions consisting of forming 'implementation intentions' along with text message reminders to achieve walking-related plans or goals using social cognitive theory (SCT) – found the greatest increase in walking in interventions that adhered more closely to SCT. High fidelity intervention increased walking by 34.23 minutes a week (+/-81.91) compared to a low fidelity increase of 7.91 minutes a week (+/-47.93, F=3.207 p=0.08).

The evidence on community delivered health information interventions is only partially applicable to the UK as most studies were conducted in Australia or the USA with only one UK study included. The differing environment in Australia and the USA must be considered in reference to the studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

Note: pedometers are a technology which offers an opportunity to present individualised information about walking and so are closely linked to the studies above. Use of pedometers is related to goal setting and monitoring rather than to delivery of information about health benefits or methods to overcome barriers. Studies may use pedometers as one of a number of factors to support increases in walking, in common with other approaches, or may use pedometers solely as a means of measuring change. Pedometer studies are considered below.

1 Dunton (2008)

2 Humpel (2004)

3 Nies (2003)

4 Nies (2006)

5 Prestwich (2010)

6 Rovniak (2005)

Evidence statement R1.ES14: Individual-level change in community-based pedometer interventions to increase walking

Moderate evidence from 12 studies suggests that pedometer-based interventions delivered in the community are effective in adults (or women only) to increase individual levels of walking for leisure or travel, up to 6 months post intervention. Evidence from five (++) RCTs1,2,3,4,12 and 2 BA studies5,6, (one [+] and one[-]) showed positive effects on walking for leisure and/or travel in adults. This is supported by data from a (-) CS study7. However, one (++) RCT8 found that short-term improvements in walking 4 weeks post intervention had decreased by 12 months follow-up. Evidence from one (++) RCT9 and one (+) BA study10 showed substantial positive effects on walking for leisure and/or travel in women. An additional (++) RCT11 found that a pedometer-based intervention increased walking in environments with low aesthetics, but not in those with aesthetically pleasing environments.

One (++) RCT8 (UK n=61, 52 weeks) – walking programme with goals set in minutes, or steps or using a pedometer – found that the pedometer group increased walking at 4 weeks (p<0.001), but decreased between 4 weeks and 12 months. No change in minutes or control groups.

One (++) RCT1 (UK n=130, 4 weeks) – motivational component had three stages: participants were shown 10 statements about what would make it easier for them to walk more, asked to complete a scale to show how confident they would be about walking in each situation, and discussed with facilitator and walking plan developed; pedometers were worn – found a significant difference in number of minutes spent walking to week 2 between the control group (M=138.7, SD=93.3) and the intervention group (M=22.5 SD=100.3), from a mean of 19.8 minutes to 32.2 minutes per day (increase of over 60%). Also a significant increase in the number of minutes spent walking per week for intervention group from week 1 to week 4 (mean 287.3, SD=129.4 t[46]=8.12, p<0.001).

One (+) BA10 (USA n=36, 6 weeks) – women who were designated as insufficiently active were given brochures and pedometers and were sent emails. Participants received a pedometer, 6 weeks of step log sheets, self-addressed envelopes, and three commercial brochures describing strategies based on transtheoretical model (TTM) for increasing physical activity and the risks and benefits of physical activity – found that participants significantly increased their total walking minutes from baseline (median 55) to post intervention (median 245: Z=4.03, p=0.001) including walking while at work (Z=2.79, p=0.005, d=0.63), for transport (Z=2.86, p=0.004, d=0.60) and during leisure time (Z=3.54, p=0.001, d=0.81).

One (++) RCT2 (Japan n=68,12 weeks) – feedback based on accelerometer daily physical activity, number of daily steps and time spent performing daily moderate physical activity (MPA) which was provided to each participant every 2 weeks. Participants were recommended to accumulate 9000 steps and 30 minutes of MPA per day – found a significant group interaction for steps (f=10.53, p<0.01). The intervention group increased their steps by 16% (7811 +/-3268 to 9046 +/-2620 steps). There was no significant change in the control group.

One (++) RCT3 (Australia n=314, 3 months) – self-help booklet based on social cognitive theory constructs, plus six weekly diaries printed on reply-paid postcards (WP group), plus a pedometer (WPP group). Three incremental stages, starting with short walks (<15 minutes) 3 days a week, typically by incidental walking, gradually increasing the duration of walks to 3 to 4 days, then (continuously) walking briskly for 30 minutes – found that the mean change in total sessions of all-purpose walking/week increased within all groups from baseline, but increased the most within WPP. The control group had a mean increase of 1.2 sessions/week (95% CI: 0.6-1.8, t=3.97, p<0.001); WP: 1.3 sessions/week (0.5–2.0, t=3.32, p<0.001); WPP: 2.3 sessions/week (1.6–3.1, t=6.30, p<0.001). Leisure time walking sessions/week for the previous 3 months also increased within all groups, with both WP (2.0 sessions/week 1.6–2.4, t=9.49, p<0.001) and WPP (2.1 sessions/week 1.7–2.6, t=9.63, p<0.001) showing a significantly larger increase than the control group (0.9 sessions/week 0.6–1.2, t=5.82, p<0.001). There was a similar pattern for leisure time walking minutes/week for the previous 3 months, but only the WPP group (66 minutes/week 50–82, t=8.05, p<0.001) showed a significant increase compared to the control group (34 minutes/week 21–48, t=5.03, p<0.001). The WPP group was also more likely than controls to meet physical activity recommendations. Unclear if the provision of pedometers provides benefit over and above standardised structure walking programme.

One (++) RCT11 (Australia n=369, 3 months) – participants received a single mail-out of a self-help walking programme (WP) or the same programme plus a pedometer (WPP) – found that only the WPP group were significantly more likely than controls to increase total walking time (Exp [b] = 2.53, p<0.01) and to undertake regular walking (OR=5.85, 95% CI 2.60–12.2) where environment aesthetics (level of greenery and interesting scenery) were perceived to be low; while in aesthetically pleasing environments, the differences in walking measures between intervention and control groups were non-significant.

One (+) BA5 (Japan n=56, 4 months) – subjects were given a pedometer and instructed to walk at least 7,500 steps each day. They were also given additional monthly advice on healthy diet and lifestyle provided in a newsletter – found the mean steps per day increased significantly from 9389 to 11846 (p<0.01).

One (++) RCT9 (USA n=24, 24 weeks) – given pedometer and initially, all (post-menopausal) women were prescribed a distance of 1.4 km/day above their baseline. Distance was then increased by 0.5 km/day until the desired walking distance was met – found that the intervention group increased their daily walking by 4300 steps (2.9 ± 0.2 km/day); significantly different from baseline and from the control group (both p<0.05).

One (++) RCT4 (Australia n=26, 12 weeks) – participants (overweight middle-aged women) in the pedometer group were told to record their pedometer steps on a daily basis for 12 weeks; those in the control group were asked to wear a sealed pedometer for 12 weeks with weekly recording. The pedometer group was also encouraged to reach a daily step goal of 10,000 steps/day – found that the pedometer group daily average number of steps at weeks 6 (8321 ± 884 steps/day) and 12 (9703 ± 921 steps/day) were significantly higher than the baseline daily average (of 6242 ± 541 steps/day: p=0.046 and p=0.035) respectively.

One (-) BA6 (USA n=12, 2 weeks) – participants over 65 years of age; site-specific walking route maps, health counselling session with individualised goal-setting and pedometers – average daily pedometer steps increased between baseline (M=3020, SD=1858) and week 1 (M=4314, SD=2627; t[11]= -2.99, p=0.012) and week 2 (M=4246, SD=2331; t[11]=3.42, p=0.006) found that all participants met their daily step goals in week 1 while 50% met their step goals in week 2.

One (-) CS7 (Canada n=41, 6 months) – lending pedometers to patrons of five public libraries. The pedometers were loaned for maximum of 9 weeks. Education packages were handed out with the pedometer including: information on pedometer use, physical activity/walking recommendations, maps of local trails, and a walking challenge questionnaire – found that 39.5% indicated they walked more since borrowing the pedometer and 60.5% reported walking about the same.

One (++) RCT12 (n=79 12 weeks) – the sessions were based on the Transtheoretical Model of exercise behaviour change. Strategies used included enhancing motivation, overcoming barriers and developing appropriate walking plans. Followed a 12-week pedometer-based walking program – found a significant increase in steps/day for the intervention group between baseline (M=6802, SD=3212) and week 12 (M=9977, SD=4669, t(38)=-6.06, p<0.001, d=0.79, CI 2,115–4236). No significant difference was observed in the control group (t(39)= -0.50, p=0.618, CI -463–770).

The evidence on community pedometer interventions to increase walking is only partially applicable to the UK. Three studies were conducted in the UK, with the majority in the USA, Australian, Canada, and Japan. The differing environments must be considered in reference to the studies, particularly for those conducted in Japan. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Darker (2010)

2 Koizumi (2009)

3 Merom (2007)

4 Pal (2009)

5 Miyazaki (2011)

6 Rosenberg (2009)

7 Ryder (2009)

8 Baker (2011)

9 Moreau (2001)

10 Dinger (2005)

11 Merom (2009)

12 Baker (2008b)

Evidence statement R1.ES15: Individual-level change in workplace pedometer interventions to increase walking

Moderate evidence from 11 studies suggests that pedometer-based interventions delivered in the workplace may be effective in increasing individual levels of walking for leisure or travel, up to 12 months post intervention. Evidence from three (++) RCTs2,3,4, one (+) nRCT5, two (+) BA6,7 and two (+) ITS studies8,9 showed positive effects on walking for leisure and/or travel in the short term (up to 12 weeks). However, one (+) ITS study10 which used a competition format, saw the initial increase in walking decline over 12 weeks. One (+) nRCT11 found significant increases in walking 12 months after the intervention, while another (++) RCT1 found that initial increase in walking declined by 52 weeks follow-up.

A (++) RCT1 (UK n=50, 52 weeks) – walking programme with goals set in steps using an open pedometer for feedback – found that both groups significantly increased step counts from baseline to week 4. Significantly greater number of participants in the intervention (77%) compared with the control (54%) achieved their week 4 goals (X2= 4.752, p=0.03). There was no significant change in step counts from week 4–16 and a significant decrease from week 16–52.

A (+) ITS10 (USA n=640 (in 64 teams of 10), 12 weeks) – competition-based employer sponsored physical activity programme using pedometers. Employees formed groups of 10 to undertake the challenge of attaining 10,000 steps per participant per day – found that total weekly steps for all teams combined increased between weeks 1 and 8 (p<0.0001), but declined from weeks 9–12. Increase in total weekly step count between week 1 and 12 not significant. Significant difference in team steps, with post-hoc comparisons indicating significant differences from baseline step counts during weeks 6–8 (F=71.15, p<0.001) but not at the end of the programme.

A (+) nRCT11 (Australia n=205, 12 months) – staff defined as inactive received a 3-month self-help walking programme and pedometer plus four maintenance newsletters over 9 months to assist them to maintain their new activity levels. Control received pedometer and programme but no maintenance – found that both intervention groups significantly increased minutes walking (p=0.01). Change in moderate or vigorous physical activity (MVPA) minutes was significantly higher in the standard plus maintenance group compared with the standard group (118 minutes versus 69 minutes, P=0.029). No significant differences between groups were observed for total physical activity (161 minutes versus 117 minutes, P=0.187).

A (+) ITS9 (Canada n=106, 12 weeks) – adoption phase: participants met in workplace-based groups with a facilitator for 30–60 minutes each week during a lunch break. Set individual steps per day goals and self-monitored their progress using a pedometer to record daily accumulated steps taken. Then adherence measured for 8 weeks – found that steps per day increased (from 7,029 +/- 3,100 [SD] at baseline to a plateau of 10,480 +/- 3,224 steps/day by 3.96 +/- 3.28 weeks of the intervention). Some decreases in activity relative to baseline steps per day, (ranging from -2.4% to -20.6% [12.0% ± 7.6%]).

A (+) nRCT5 (Australia n=56, 6 weeks) – the intervention group received a pedometer and step logs. Set a daily step goal based on the previous week's step counts. They received weekly email reminders to wear the pedometer and return that week's log. They also received three commercial brochures. The control group received the intervention but without commercial brochures, intervention emails contained transtheoretical model (TTM)-based strategies – found that daily steps increased significantly (from 6419 ± 2386 during week 1 to 7984 ± 2742 during week 6: p<0.001) for both groups combined. Increases did not differ between groups.

A (+) ITS8 (USA n=206, 10 weeks) – each day participants put on pedometers upon arriving at work, prior to getting out of their cars. To increase motivation, participants were encouraged to develop teams, and each team chose a team leader. Weekly motivational emails were sent to participants – found a significant increase in the number of steps per week for weeks 2, 3, 4, 6 and 8 compared to baseline (p=0.001).

A (++) RCT2 (UK n=64, 10 weeks) – walking routes which employed prescribed walks around campus with participants asked to complete at least 15 minutes continuous brisk walking every day and 'Walking in task' which encouraged the accumulation of step counts through the working day – found a decrease in steps for the control group (-767 steps/day) and increases in intervention groups for walking routes (+926 steps/day) and walking in tasks (+997 steps/day). (Control versus walking routes p<0.008, control versus walking in tasks p<0.005).

A (++) RCT3 (UK, Australia and Spain n=64, 70 and 80 respectively, 10 weeks) – participants in the first intervention group were directed to increase their step count through brisk, sustained, route-based walking during work breaks. The second intervention group was asked to engage in incidental walking and accumulate step counts during working tasks, both groups were instructed to use pedometers to motivate and regulate walking – found that average step count data decreased in the control group (-391 steps/day t=1.76; p <0.08) and significant increases in both the routes (968 steps/day; t=3.9; p<0.001) and the incidental group (699 steps/day; t=2.5; p<0.014).

A (+) BA6 (USA n=290, 12 weeks) – participants wore a pedometer at least 5 days per week for 12 weeks and completed questionnaires assessing demographic information. After baseline (week 1) they were given suggested number of steps to meet recommendations, instructions for goal-setting and other behaviour-change strategies to gradually increase number of daily steps – found that the average number of steps increased from week 1 to week 6 (p<0.001) and week 12 (p=0.002).

A (++) RCT4 (Canada n=63, 1 week) – intervention group pedometer was worn for 1 week for all waking hours to encourage walking. Control (non-pedometer) participants were informed they could wear a pedometer the following week – found that, compared to the no pedometer group, the pedometer group reported more walking (F=5.22, p=0.03).

A (+) BA7 (USA n=188, 10 weeks) – participants were provided with pedometers and given personalised daily and weekly step goals over the 10 week intervention. Local strategies available to the participants included walking groups, marked walking circuits and posted walking maps – found a mean increase of 1503 steps (38% increase over baseline). Mean weekly step counts values for all intervention weeks were significantly higher than baseline (p<0.01).

The evidence on workplace pedometer interventions to increase walking is partially applicable to the UK. Three studies were conducted in the UK but most studies were conducted abroad: in USA, Australia, Canada or Spain which may limit the applicability in some cases. The differing environments must be considered in reference to the studies. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Baker (2008a)

2 Gilson (2007)

3 Gilson (2009)

4 Spence (2009)

5 Dinger (2007)

6 Jackson (2008)

7 Warren (2010)

8 Faghri (2008)

9 Chan (2004)

10 Behrens (2007)

11 Borg (2010)

Evidence statement R1.ES16: Individual-level change in workplace delivered targeted health information interventions to increase walking

Weak evidence from two studies suggests that individual, targeted provision of health information delivered in the workplace (including flyers, email, telephone calls, website postings, and information booths) may be effective in increasing individual levels of walking for leisure or travel in adults up to 24 weeks post intervention. One (+) RCT study1 showed a positive effect on walking and one (+) BA study2 showed a small (borderline significance) positive effect on walking.

A (+) RCT1 (USA n=135, 24 weeks) – phone calls once a week versus every 3 weeks, and structured versus non-structured feedback – survival curves indicated that there was a significant effect on walking for treated (the combined four treatment conditions) versus the control condition (LD=17.661 p<0.001) and for frequency of prompting (those prompted once a week against every 3 weeks) (LD=17.719, p<0.0001).

A (+) BA2 (USA, n=not reported, 2 weeks) – promotional material distributed via flyers, email, website postings, and at bi-weekly information booths – borderline statistically significant increases in walking counts on a route ('Path to health') from baseline midway through the campaign (p=0.069) and following the campaign: (p=0.075 – p values only reported).

The evidence on workplace health information interventions is only partially applicable to the UK as the studies were conducted in the USA. The differing environment in the USA must be considered in reference to the studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Lombard (1995)

2 Napolitano (2006)

Evidence statement R1.ES17: Individual-level change in workplace delivered targeted health information interventions to increase walking and cycling

Moderate evidence from one (++) RCT study1 suggests that individual, targeted provision of health information (including a booklet of interactive materials, social marketing and individualised marketing strategies) in the workplace may be effective in increasing individual levels of walking, but not cycling, for travel in adults for up to 6 months post intervention. See also R1.ES4.

A (++) RCT1 (UK n=295, 6 months) – interactive materials based on the transtheoretical model of behaviour change: choosing routes, maintaining personal safety, shower and safe cycle storage information, and useful contacts – found a significant increase in time per week spent walking to work (mean 125 minutes/week intervention versus 61 minutes/week control), but no difference in average weekly minutes of cycling between cyclists in the intervention group (n=9) and control group (n=9).

The evidence on health information intervention to increase walking and cycling is applicable to the UK as the study was conducted in the UK. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Mutrie (2002)

Evidence statement R1.ES18: Individual-level change in multi-component interventions to increase walking

Weak evidence from two (++) BA studies1,2 suggests that multi-component interventions have a positive effective on increasing individual levels of walking for leisure or travel up to three months post intervention.

A (+) BA1 (USA n=124, 8 weeks) – multi physical activity and dietary program, pedometers – found post intervention that 46.2% (n=43) met the 10,000 steps/day criteria for high activity (no further statistics). This increased from 11.8% at baseline. Average steps increased from 5969stpes/day to 9757 steps/day

A (+) BA2 (USA n=53, 3 months) – sponsored walking groups, improving walking routes, providing information about walking options, and advocating for pedestrian safety – found self reported walking activity increased from 65 to 109 minutes per day: 44.1% increase (95%CI= 28.0-60.2, p=0.001). The proportion that reported being at least moderately active for at least 150 minutes per week increased from 62% to 81% (19.2 % increase 95% CI= 2.2, 36.3 P=018);

The individual level evidence on multi-component interventions to increase walking is only partially applicable to the UK as studies were conducted in the USA. The differing environment in the USA must be considered in reference to the studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Clarke (2007)

2 Krieger (2009)

Evidence statement R1.ES19: Individual-level change from cycle training interventions to increase cycling

Weak evidence from one BA (+) study1 suggests that cycle training interventions may be effective in increasing individual levels of cycling for active travel amongst those not cycling at baseline, up to 2 months post intervention.

A (+) BA1 (Aus n=81, 2 months) – practical skills development and supervised on road or cycle path training. Free courses for beginner and intermediate level cyclists were conducted. Promoted through flyers, posters, media releases, articles and TV and newspaper adverts – found non cyclists at baseline reported significant increase (p<0.001) in minutes cycling.

The individual level evidence on multi-component interventions to increase cycling is only partially applicable to the UK as the study was conducted in Australia. The differing environment in Australia must be considered in all studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Telfer (2006)

Evidence statement R1.ES20: Individual-level change in healthcare delivered multi-component interventions to increase both walking and cycling in adults

Moderate evidence from 1 (++) RCT study1 concerning the effect of multi-component interventions on increasing individual levels of both walking and cycling for travel and/or leisure up to 18 months post intervention indicated a positive effect on cycling but no effect on walking.

A (++) RCT1 (Sweden n=120, 18 months) – physician meetings, physical activity prescriptions, group counselling, and bicycle provision; control and intervention groups received pedometers – found the intervention group was more likely to achieve recommended level of cycling than controls (38.7% versus. 8.9%, OR=7.8, 95%CI 4.0-15.0, p<0.001) but there was no difference in compliance with the walking recommendation (45.7 versus. 39.3%, OR 1.2, 95%CI 0.7-2.0, p=0.5). Commuting by car and public transport were reduced by 34% (P<0.01) and 37% (P<0.001), respectively in the whole sample, with no differences between groups.

The individual level evidence on multi-component interventions to increase walking and cycling in adults is only partially applicable to the UK as the study was conducted in the Sweden. The differing environment in Sweden must be considered in reference to this study. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Hemmingsson

Evidence statement R1ES21: Individual-level change in community-based led walking group interventions to increase walking

Inconsistent evidence from 5 studies1,2,3,4,5 on the effects of a community-based led walking group interventions on walking. 1 (++) RCT5, 1 (++) clustered RCT2, 1 (++) nRCT1 and 1 BA (+) study3 showed positive effects on walking from community-based walking group interventions; but evidence from a further (++) RCT4 showed no difference between groups at 12 months.

A (++) nRCT1 (UK n=7883, 12 weeks) – 'Get walking, keep walking': Bespoke, led walks and sessions and walking packs aimed at encouraging (predominantly) inactive people, those from deprived communities, black and minority backgrounds, women and younger adults to walk – found 67% of participants increased the amount of exercise they did each week. Walking from 'place to place' increased by 1.1 day/week and walking for leisure by 1 day/week.

A (++) cluster RCT2 (USA n=501, 6 months) – leader-led walking group activity or an information-only control group – found significant increase observed in walking activity: p <0.05.

A (+) BA3 (Australia n=169, 6 months) – walk leaders received a prescriptive progressive weekly exercise program guided by social cognitive theory, that contained written information on the appropriate length for the walking program; stretching exercises; and ball skills, such as side twist leader ball, participants aged 65-74 – found baseline mean walking time for recreation was one hour (SD =1.65), increasing to 2.69 hours (SD =2.02) per week by the end of the program.

A (++) RCT4 (UK n=260, 12 months) – accompanied walks were provided at several different times in the day and evening, during the week and at weekends, and were led by lay volunteers – found at 12 months, although both walking and control groups increased activity (by 35.7% and 22.6% respectively; 95% CI 0.003% to 25.9%) p=0.05), there was no significant difference between them.

A (++) RCT5 (USA n=114, 20 weeks) – efficacy based Exercise classes were conducted by trained exercise specialists and employed brisk walking as the aerobic component – found at the end of the 20 week program, subjects in the intervention group walked more miles per week than the control group: p<0.05.Intervention group subjects also walked more often (p<0.01) and accumulated more minutes (p<0.01) than control

The evidence on community based walking group sessions to increase walking is only partially applicable to the UK as only two studies were conducted in the UK. The differing environment in the USA and Australia must be considered in reference to the studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 CLES (2011)

2 Fisher (2004)

3 Jancey (2008)

4 Lamb (2002)

5 McAuley (1994)

Evidence statement R1.ES22: Individual-level change in interventions to increase independent community-based walking

Weak evidence from 2 (+) BA studies1,2 and 1 (++) RCT3 suggests that interventions to increase independent community based walking may be effective in increasing individual walking for leisure, exercise or travel up to 13 weeks post intervention in adults or the whole community.

A (+) BA1 (Canada n=39, 8 weeks) – 'mall walking programme', participants provided with pedometers. Participants self-selected the pace, time, and frequency of walking. Encouraged to attend as often as possible between 8am and 10am Monday to Friday – found average daily mall walk steps increased from 5055 (SD 1374) to 5969 (SD 1543): p=0.002, and average daily mall walk time increased from 42.9 (SD 10.6) min to 50.4 (SD 13.5) min: p=0.002.

A (+) RCT3 (Aus n=88, 13 weeks) – participants: postnatal women; information, goal setting consultations, activity and self-monitoring daily planner, tailored SMS, nominated social support person – found frequency of walking for exercise (days/week) increased over time in the intervention compared to control group (time×group interaction effect F(2,85)=5.38, p=0.023, medium effect size partial η2=0.06); while change in duration of walking did not show a significant time×group interaction effect (p=0.081; effect size partial η2=0.05), there was a significant group effect with increases in walking duration in the intervention compared to control (p=0.005; medium to large effect size partial η2=0.09).

A (+) BA2 (USA n=16, 12 weeks) – walking intervention facilitated by community health workers. Weekly sessions encouraged participants to accumulate at least 30 min of moderate intensity walking on most/all days of the week – found that exposure to the programme resulted in significant increase in walking: 915.8 metabolic equivalent min/week, p=0.002.

The evidence on interventions to increase independent community based walking may not be applicable to the UK as studies were conducted in the USA and Canada. The differing environment in the USA must be considered in reference to the studies conducted there. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Culos-Reed (2008)

2 Mier (2011)

3 Fjeldsoe (2010)

Evidence statement R1.ES23: Individual-level change in workplace-based interventions to increase independent walking

Inconsistent evidence from 2 (++) RCT studies1,2 concerning workplace walking session interventions (conducted in universities) effectiveness in increasing individual levels of walking for staff and/or student participants up to 12 months post intervention. Evidence from 1 RCT1 showed positive effects on walking while one RCT2 showed no effect on walking.

A (++) RCT1 (USA n=32, 32 weeks) – sedentary adults; walking prescription: 3 brisk walking conditions: 30 continuous minutes, 3 10-minute bouts, or 30 minutes made up of any combination of bouts each at least 5 minutes long; 1 hour information and modelling session followed by weekly meetings with an activity counsellor for 15 weeks. Behavioural methods used to promote adherence: goal setting and mastery, self-management techniques, weekly personal feedback, problem solving, behavioural contracting participants paid $50, refunded on successful completion – found self-reported walking for all intervention groups significantly increased throughout the program: F(6, 186)= 26.16; p<0.001.

A (++) RCT2 (USA n=26, 6 weeks) – two 8-week walking for fitness classes – found that neither group increased walking time or number of steps significantly over time.

The evidence on workplace (university) based walking sessions to increase walking is only partially applicable to the UK as the studies were conducted in the USA. The differing environments must be considered in reference to the studies conducted in the US. Individual local contexts as well as the setting will also impact on the applicability of individual studies.

1 Coleman (1999)

2 Eastep (2004)

Evidence statement R2.ES1: Providers' and researchers' views of barriers and facilitators to planning and delivering interventions to increase walking

Moderate evidence from four studies suggests that facilitators to planning and delivering interventions included organisational support and sufficient planning time. It may be beneficial to include volunteer leaders at the planning stage.

Having previous experience in marketing and a conceptual framework facilitated recruitment efforts. Personal satisfaction, social interaction and a positive rapport with group members were motivational effects of leading walking groups.

Barriers to planning and delivery included lack of inter-organisational collaboration. This was facilitated by introducing staff in different organisations to each other and being clear about shared goals. Employing an individual to co-ordinate between organisations was a facilitator to implementation.

De-motivators to being involved in organising and monitoring groups included researchers' perceived workload, efforts required for effective recruitment, lack of support from and feelings of responsibility for group members.

A (+) UK evaluation1 suggested that sufficient planning time is required for successful implementation of a family-based intervention. Involvement of proposed walking leads at the planning stage was suggested as a way of increasing their engagement with the programme.

A US (+) pilot evaluation2 reported that walking group policy makers supported the walking group by promoting the intervention and assisting with recruitment. Administrative support was also supplied, and events were organised.

A (+) study3 that included interviews from the UK reported that the process of recruiting members to a walking group was draining on time and resources for the organisers, and some volunteers lacked skills in recruitment. Having experience in marketing and a conceptual framework around recruitment was a facilitator to recruiting new members. However, word of mouth was regarded as the most effective recruitment strategy.

A pilot evaluation2 reported that running the walking group provided a sense of personal satisfaction for organisers as well as an opportunity for personal development and health promotion. Interaction with club members was a motivator for organisers.

Collaboration with other organisations was an issue in two studies1,2, due to a focus on their own organisation and lack of communication. In one study2 this meant that walking routes were not shared and events were less well attended. Club directors could also feel isolated. In the other study1, collaboration between a walking association and a family support group was improved through members getting to know each other and being clear that goals were to be shared, and that interventions would run alongside each other rather than new initiatives replacing existing ones. Coordination by one designated officer also facilitated implementation.

One study2 reported that group organisers expressed views about their burden of responsibility for the wellbeing and safety of members, especially if leadership was not shared. Recruitment and maintenance of membership numbers were regarded as a burden, and strategies were developed by the club to limit drop-out. Having to walk at a slow pace with other members was a de-motivator.

One UK (+) interview study4 found that carrying out routine physiological measurements in a pedometer study was regarded as a burden for researchers.

Findings from these studies have partial applicability to other walking groups. The organisation of walking interventions will differ across countries, regions and groups. Groups may have different goals, and recruit specific populations. There is no reason to believe that the barriers and facilitators described are not applicable to other similar interventions.

1 Milton et al. (2011)

2 Nguyen et al. (2005)

3 Matthews et al. (undated)

4 Shaw et al. (2011)

Evidence statement R2.ES2 Participants' views about motivators and barriers to participating in interventions to increase walking

Moderate evidence from five studies suggests that participating in a walking intervention motivated people to walk through the presence of role models, organised routes, and the support of being part of a group.

Families were motivated by the opportunity for children to participate in an activity that was free of charge. For others, the opportunity to improve health and enjoy fresh air and nature were motivational.

Barriers to motivation include conflicts between walking activities and work / school schedules, and cultural lack of acceptance in regard to work-based activity.

One US (+) pilot evaluation1 reported that having access to a role model and to organised walk routes were motivators to attendance. For women, having the support and security of a group was a motivator (one [++] focus groups, US) 2. For families, the opportunity for children to participate in activities with the family, free of charge, and outside of nursery hours were incentives (one [+] evaluation, UK)3. For adults, a sense of routine and structure was valued for those who were not in employment (one [+] focus groups, UK) 4.

Participants in one (+) UK focus groups study4 were motivated by the opportunity to improve their health and be out in the fresh air and natural environment.

However, barriers to participation included conflicting schedules with school attendance3 or workplace responsibilities (one [+] interviews, UK) 5. In a workplace setting, they also reported that increasing walking time required acceptance from colleagues, and this varied depending on the status of the employee within the organisation.

Applicability: The findings from these studies are applicable to other walking groups. The acceptability of walking interventions will depend upon specific walking group characteristics, settings and aims. There is no reason to believe that the barriers and facilitators reported are not applicable to interventions implemented in the UK.

1 Nguyen et al. (2005)

2 Burroughs et al. (2006)

3 Milton et al. (2011)

4 Hynds and Allibone (2009)

5 Gilson et al. (2008)

Evidence statement R2.ES3 Participants' views about maintaining participation in interventions to increase walking

Moderate evidence from ten studies provided evidence regarding factors associated with maintenance of participation.

Social interaction and social support were major factors in maintaining participation. Maintenance was also related to the extent to which activities could be integrated into daily life.

Monitoring activity, providing people remembered to self-monitor, could increase motivation, though it could also introduce unwanted competition between members.

Other motivators included variation in walking routes, and incentives such as gifts.

Barriers to maintenance included the difficulty of integrating walking and attendance at clubs into daily routines. Boredom, dissatisfaction with elements of the club, and incongruent aims were reported factors associated with discontinued membership.

The social factor associated with walking in groups was supported by six studies1,2,3,4,5,6. The social factor was particularly strong for women and older adults. One (+) UK focus group study5 reported a strong bond and sense of loyalty to the group that facilitated attendance. For men, the social factor was not so important with males tending to prefer walking alone (one [++] focus groups, US)7.

Support was also important; in one intervention7, feedback from providers was welcome, though e-mail was the preferred mode.

A US (+) RCT2 highlighted the importance of family and friends in supporting the maintenance of walking behaviours. Walking also had a positive effect on interactions with family members.

A UK (+) interview study8 reported that walking to deliver messages at work instead of e-mailing created a greater sense of community.

An important aspect of walking was the ability to integrate interventions into daily life. The ability to turn up without booking was a positive factor for some, and a sense of routine and structure was valued for those who were not in employment (one [+] focus groups, UK) 5.

However, one study1 reported that women in particular found difficulty integrating extra walking into daily routines. Life changes, coinciding schedules and other commitments were also a barrier2,4,5,9. Wearing female-oriented clothing such as high heels was a barrier to walking while at work2,8,9. For African–American women, it was difficult to focus on self-based activities4.

Monitoring activities was reported as a motivator. Two studies1,10 reported that pedometer use and the process of self-monitoring increased walking behaviours. One5 reported that step counting gave a sense of achievement.

However, a US observation and interviews (+) study6 found that in older adults (mainly female), pedometer use and fitness objectives conflicted with the moral economy (shared values regarding social interaction) of the walking group, which was based on sociability rather than competition. In addition, people often forget to complete logs, or to use their pedometer10.

Other incentives included rewards and gifts7.

One study9 reported that the atmosphere of the club, mismatch between aims of the club and aims of the participant, as well as the pace required to walk could be barriers to participation in walking interventions. Another1 also added that boredom could dissuade attendance, and for African–American women, one study reported lack of objectives as potential barriers4.

Applicability: The findings from these studies are applicable to other walking groups. The motivation to maintain walking behaviour within an intervention will depend upon individual circumstances and requirements as well as the characteristics and aims of the club. There is no reason to believe that the barriers and facilitators reported are not applicable in the UK.

1 Shaw et al. (2011)

2 Nies and Motyka (2006)

3 Milton et al. (2011)

4 Dunn (2008)

5 Hynds and Allibone (2009)

6 Copleton (2009)

7 Burroughs et al. (2006)

8 Gilson et al. (2008)

9 Nguyen et al. (2005)

10 Zoellner et al. (2009)

Evidence statement R2.ES4: Participants' views of the benefits of participating in a walking intervention

Moderate evidence from eight studies highlighted the reported benefits of walking as part of a walking intervention.

Perceived benefits to walking were reported to facilitate motivation and hence walking behaviour (one [+] focus groups, US) 1. Such benefits could be emphasised when encouraging participation in interventions.

Reported benefits included physical and psychological benefits, adding variety to the day and getting out of the house or office. Walking could provide a sense of peace and solitude, and was also fun, providing an opportunity to be out in fresh air and see the sights.

Reported physical benefits were feeling healthy1, (one [++] focus groups, US)2, and fit (one [+] pilot evaluation, US)3, (one [+] RCT, US)4, increased energy4, (one [+] interview studies, UK)5, lower blood pressure4, weight loss1,4 and improved body shape1.

Psychological benefits included enhanced mood4,5, stress reduction1,2,4, mental and emotional satisfaction4, feeling rejuvenated4, and having meditative or spiritual feelings1. Feeling tired at the end of a walk was associated with a sense of achievement (one [+] focus groups, UK)6.

In a workplace intervention, walking was reported to add variety to the day and improved output at work5. For a group of previously sedentary adults, walking became fun, and was a chance to get out of the house3. Walking for one group of mid-age women allowed them time to think, time out of the office, time with the family and fresh air4.

Benefits reported from two (+) pedometer-based interventions included seeing the sights (interviews, UK)7, and socialising with members of the group (observation and interviews, US)8.

Applicability: The findings from these studies are applicable to other walking groups. Benefits of walking may differ by setting, though there is no reason to believe that the benefits reported are not applicable in those settings within the UK.

1 Dunn (2008)

2 Burroughs et al. (2006)

3 Nguyen (2005)

4 Nies and Motyka (2006)

5 Gilson et al. (2008)

6 Hynds and Allibone (2009)

7 Shaw (2011)

8 Copleton (2009)

Evidence statement R2.ES5: Walking intervention participant's views of perceived barriers to walking

Moderate evidence from seven studies highlighted perceived barriers to walking for participants of walking interventions. These included physical and psychological limitations, environmental barriers, and poor weather conditions.

Physical barriers to continuing with the walking programme included health problems such as arthritis (one [+] focus groups, US) 1, and physical limitations such as illness and injuries (one [+] RCT,US)2. Tiredness and depression also prevented some women from continuing attendance1.

Poor weather conditions or hot weather were reported disincentives to walking1,2 (one [+] interviews,UK3; one [+] pilot evaluation, US4;one [++] focus groups, US5;one [+] focus groups, UK6). One study reported costs of participation as a barrier4.

Lack of access to the walking route, and obstacles such as poorly maintained stiles along the walking route were also reported barriers6.

Applicability: The findings from these studies are applicable to other walking groups. The barriers to participation in walking interventions might depend upon individual circumstances, such as age and physical fitness as well as seasonal weather conditions. Weather conditions may be better, or more extreme, in the US, Canada and Australia than in the UK, though there is no reason to believe that the barriers reported are not applicable in the UK.

1 Dunn (2008)

2 Nies and Motyka (2006)

3 Shaw (2011)

4 Nguyen et al. (2005)

5 Burroughs et al. (2006)

6 Hynds and Allibone (2009)

Evidence statement R2.ES6: Suggested strategies to overcoming barriers to maintaining walking in a walking intervention

Moderate evidence from two studies highlighted reported strategies to overcome perceived barriers to participating in walking interventions. These included making time, and integrating walking into daily life as well as thinking positively.

A (+) US RCT1 reported strategies including scheduling time to walk, problem-solving and using motivators such as positive thinking and focusing on the long-term benefits. Goals were more achievable if walking was made a priority and was fitted into daily life as much as possible. Similarly, a (+) US study2 (focus groups) reported that for African–American women, weaving walking into family life was a strategy that allowed themselves and the family to participate.

Applicability: The findings from these studies are applicable to other walking groups. The ability to implement strategies to overcome barriers to participation in walking interventions will depend upon individual circumstances.

1 Nies and Motyka (2006)

2 Dunn (2008)

Evidence statement R2. ES7: Providers' views about effective intervention components that motivate walking and cycling

Moderate evidence from one study suggests that workplace efforts to encourage walking and cycling are most successful where they attend to cultural attitude, access, security and available facilities. Incentives and provision of equipment are also motivating.

The (+) study (survey and interviews, UK)1 provides evidence that, across 20 workplace initiatives, walking and cycling are increased where good onsite and offsite access is available, along with provision of showers, drying and changing facilities. Organised walks at lunchtime and cycling groups were an incentive.

Organisational attitude was important, with some workplaces marketing the benefits of walking to staff. Motivators such as complementary products or financial incentives were used.

For cycling, the ability to borrow equipment or receive discounts on cycling equipment was important, as was having secure parking for cycles.

Applicability: Findings from this study were taken from a range of workplace initiatives within the UK and so are applicable in UK workplace settings.

1 Cairns et al. (2010)

Evidence statement R2.ES9: Participants' views about taking part in interventions to increase cycling

Moderate evidence from one exploratory study and one evaluation showed that facilitators to a led cycling intervention were a feeling of safety and acceptance that was obtained from cycling in a group.

Provision of acceptable equipment and the need not to wear a helmet was a facilitator for boys.

In a workplace-based cycling intervention, facilitators included the provision of storage and changing facilities and raised awareness about benefits.

The (++) exploratory study (focus groups, UK)1 elicited community members' views about use of a cycle trail and a proposed intervention that included led cycling groups.

The main facilitator to using the trail for led cycle groups was the protection of riding together in a group. For young women, the image of cycling as 'uncool' was an issue, but this barrier would be lessened if they were cycling with friends.

Image was also an issue for boys, whose participation would be facilitated by the provision of the 'right' bike, and not having to wear a cycling helmet.

The (+) UK evaluation study2 (found that the main influences on increase in cycling following an intervention were the provision of workplace cycling facilities, a house or job move that made cycling more attractive, and heightened awareness of the importance of physical activity for health. Welcomed and best used measures were secure cycle parking, showering and changing facilities, and cycle purchase loans.

Applicability: The findings from these UK-based studies are applicable to other potential cycling interventions. The motivation to participate in cycling interventions might depend upon individual circumstances, as well as local geography and usage of the proposed site. Some areas of the UK may be more or less attractive as cycling venues than the one described here. Workplaces will also differ in provision of facilities, and interventions may be affected by factors outside the control of organisers, such as weather conditions.

1 Cavill and Watkins (2007)

2 Cleary et al. (2000)

Evidence statement R2.ES10: Young people's views about walking for travel or leisure (not related to an intervention)

Moderate evidence from one interview study and one survey study suggests that walking for leisure was facilitated by walking as a social event or as part of a challenge.

Barriers to walking for travel or leisure for young people are mainly related to lack of time. In addition, having a lot to carry and wearing shoes that were not comfortable were disincentives. Young people report busy lives as a barrier to walking for transport. For men, walking was not sufficiently vigorous to be considered 'exercise'.

The (++) UK interview study1 reported that young people, and especially young men, did not regard walking as vigorous enough to provide exercise. Walking for transport required too much time out of a busy day. Walking for leisure was only acceptable if it included some form of teamwork or challenge. For those that did walk for transport, listening to music was a facilitator as it drowned out noise from traffic and construction sites.

The (+) US survey study2 reported that undergraduates found that lack of time, having a lot to carry, and wearing shoes that were uncomfortable were the most highly rated barriers.

Applicability: The findings from these studies are applicable to young people in the UK and US. Evidence reflects aspects of daily life that alter with changes through the life course. Participants in this study are constricted by timescales associated with the working day that might not apply to some other populations. There are also specific gender differences in perceptions of walking for fitness.

1 Darker et al. (2007)

2 Dunton et al. (2006)

Evidence statement R2.ES12: Older people's views about walking for travel or leisure (not related to an intervention)

Moderate evidence from six studies suggests that the main facilitator to walking for travel or leisure in older adults was social interaction.

Barriers to walking for travel or leisure for older adults are related to limited mobility and fears for safety. These factors were mediated by the external environment, with fears of falling or of swift traffic being commonly voiced.

Walking indoors was a relatively safe and comfortable alternative if designed appropriately. Walking indoors also incorporated a social aspect to walking.

Older adults reported factors that impacted on safety as the main barriers. When walking outside, narrow pavements and obstacles such as parked cars on pavements, and construction sites were barriers to access (one [-] interviews, UK)1. Traffic was also an issue, with cycle tracks and bus lanes creating hazards. Suggested improvements were wider pavements and better provision for cyclists.

In addition, two focus group studies from Canada (one [++]2 and one [+]3) reported that fear of falling was a barrier to older adults, particularly in icy weather. Uneven pavements and car parks that are not designed for pedestrians were hazards. Older adults often require more time to cross roads, and it was reported that fast roads and poor visibility at crossroads were barriers to outdoor walking.

Suggestions for improving the walking experience for this group were access to toilets and seating, as well as adequate access to local amenities and pedestrianised shopping areas. Making sure that pavements were smooth and clear of snow and ice was also a factor2.

A (+) UK survey 4 reported that obstructions to mobility included crossings without dropped kerbs, narrow footpaths, and a dropped curb with a steep angle. The authors report that 19% of people aged >80 years could not reach key places if they need to pass through a gap of 1000 mm.

Two studies assessed indoor walking for older adults. A (++) observations and interviews study from the UK5 reported on mall walking that not only contributed to improved physical activity, but also provided a social network and a meaningful work replacement following retirement. Routines were adapted and events were organised in a relatively safe environment compared to outdoors.

For older adults in assisted living facilities, a (++) focus group study from the US6 reported similar facilitators in corridor walking, such as relative safety of being indoors, and the social incentive of meeting people in the corridors. Handrails were valued, as well as appropriate flooring, seating in corridors and adequate toilet arrangements. Public rooms needed to be thoughtfully placed to allow residents optimum access.

Reported barriers to this activity6 were the lack of varied things to see compared with outside. Facilities with outdoor walking areas provided an opportunity to overcome this barrier providing the walking surfaces were adequate.

Applicability: The findings from these studies are applicable to older adults in the UK and North America. The evidence reflected safety concerns that alter with changes through the life course such as ageing. Participants in this study were constricted by limited mobility that might not apply to some other populations. Social interaction is important for this population to prevent social exclusion.

1 Newton et al. (undated)

2 Lockett (2005)

3 Ripat et al. (2010)

4 Mackett et al. (2001)

5 Duncan et al. (1995)

6 Lu et al. (2011)

Evidence statement R2.ES13: Views of people from deprived areas about walking for travel or leisure (not related to an intervention)

Moderate evidence from two studies suggests that the main barriers to walking for travel or leisure in people from deprived areas were safety, lack of time and lack of motivation.

Women were constricted by perceived dangers from the external environment, family commitments, lack of motivation and lack of walking companions.

There was evidence that participants were either out of the habit of walking, or that walking was enforced due to a lack of options.

For men, walking was not sufficiently vigorous to be considered 'exercise'.

Two studies assessed the views of populations from deprived groups. One (+) UK interview study1 reported that males did not associate walking with exercise as it is not strenuous enough. Women more often preferred to walk with someone else rather than alone, so walking with a friend, or children was an incentive. Walking with a dog was a motivator for men or women.

Though health benefits such as weight management and reducing aggression or boredom were recognised by those that did maintain walking activities, there was a habit of not walking that needed to be broken. Lack of motivation, other commitments, lack of time and bad weather were all barriers to continuing walking1.

A (+) UK interview study2 examined the experiences of women without access to a car and reported feelings of social exclusion due to having to walk in neglected areas and often with very young children, who were tired. Women often had to walk long distances to shops, and feared for their children's safety at busy roads.

Applicability: The findings from these studies are applicable to people living in deprived areas in the UK. The evidence reflected safety concerns associated with perceived environmental dangers. Participants in this study were constricted by reduced options that might not apply to some other populations. Social interaction is important for this population to increase the feeling of safety, particularly for women. There were also specific gender differences in perceptions of walking for fitness.

1 Ipsos/MORI (unpublished report)

2 Bostock (2001)

Evidence statement R2.ES15: Views about barriers and facilitators to active travel to school (walking and/or cycling for transport)

Moderate evidence from nine studies suggested that the main facilitators to active travel included the social aspect of walking and spending time with friends, or having quality time with parents.

Barriers for schoolchildren contemplating active travel to and from school were parental and children's lack of time and dangers from traffic and from intimidation or attack by other people. The missed opportunity by schools to develop children's existing awareness, and displaying conflicting messages was also a barrier. Peer pressure was an important factor for this age group in terms of choices.

Other reported barriers included distance, carrying heavy bags, and poor weather conditions. Parental habits and commitments as well as fears for their children's safety were also influential on decisions about walking.

Barriers to cycling for children included a lack of cycle lanes and a lack of facilities to store bicycles.

The perceived image of cycling, and a dislike of wearing cycling helmets was also reported to be a barrier.

Walking or cycling

Three studies (one [++] focus groups, UK1; one [++] focus groups, US2; and one [+] survey and interviews, UK3) identified recognition in parents and children that walking or cycling would be beneficial to health and could increase a child's confidence and sense of independence around roads. In addition, two studies1, (one [+] focus groups, UK)4 reported that walking with a parent provided valuable time together. Spending time with friends was an important social aspect for older children1.

However, barriers to walking or cycling included lack of time1,2,3,4; parents often needed to accompany children to different schools and arrive at their place of work in time. Children and parents would need to get out of bed much earlier in the morning in order to fit in walking. Laziness was reported as a reason for not using active travel1.

Peer pressure and the trend toward car ownership was a factor, particularly for cycling, which for some groups was socially unacceptable. Schools may also miss opportunities to develop children's knowledge about sustainable transport choices3.

A US (+) survey5and an Australian (+) survey6 found that among children that did not walk to school, distance was the most commonly reported barrier, followed by traffic danger. Parents restricted their children to playing close to home on their bicycles (one [+] focus groups, UK)7.

Children having to carry heavy bags of books and equipment was a barrier to both walking and cycling1,3,4, as were bad weather, dark mornings1,2,4 and hilly terrain4.

For older children who travel without an adult, there were fears for personal safety1,2, of accidents and abductions2, of strangers and bullies4,7 and of busy traffic1,2,4,7. Environmental factors such as poor lighting, secluded areas or woodland on the journey exacerbated these fears1,2,4,7.

A (+) survey from Australia8 showed that parental perceptions were a factor in decisions to walk. These included parents own physical activity habits, parental working schedules, and parental concerns about safety. Having to attend out-of-school activities was also a factor.

Cycling

Cycling was associated with particular barriers, such as lack of cycle lanes, and general support for cycling at school such as provision to store bicycles and helmets1,4. Fear of having a bicycle stolen was a disincentive1,7.

The image that cycling conveyed was an issue for some. For teenage girls, cycling was perceived as childish4. For children that did cycle, the 'coolest' bike was required4, and cycling helmets were regarded as 'uncool'1, (one [+] action research, Australia9), lacking in style and fit, with consequences such as negative comments from others9. In addition, cycling impacted on personal appearance; for example, cycling helmets dishevelled one's hair1.

Applicability: The findings from these studies are partially applicable as the findings are specific to schoolchildren. While some barriers and facilitators to active travel are applicable to any population, schoolchildren and their parents face particular issues pertaining to safety and practicalities for children. Some barriers differ by age group and gender.

1 Kirby (2008)

2 Ahlport et al. (2008)

3 Halden Consultancy (2003)

4 Granville et al. (2002)

5 Beck et al. (2008)

6 Yeung et al. (2008)

7 Davis and Jones (1996) and Davis (2001)

8 Ziviani et al. (2004)

9 Stevenson and Lennie (1992)

Evidence statement R2.ES16: Suggestions for strategies to encourage active travel to school (walking and/or cycling for transport)

Moderate evidence from five studies provided suggestions for strategies that might encourage safe active travel in schoolchildren.

Suggested strategies included environmental improvements to increase safety, changing attitudes to car use, school-based campaigns to assist in cycling skills and awareness, and personal-level encouragement by provision of storage facilities and better design of cycling helmets.

Suggested strategies that may overcome some of the reported barriers included employing crossing patrols near to schools (one [++] focus groups, US1), escort schemes, traffic calming schemes, and pedestrian training (one [+] focus groups, UK2).

A (+) survey from the UK3 reported that modifying attitudes to car-centredness would be a useful policy; more so than promoting general environmental awareness.

To reduce cycling accidents, improved cycle paths and compulsory helmet wearing was suggested in one (+) study (action research, Australia)4.

Other suggestions included schools organising walking and cycling groups, providing training in cycling proficiency, and support such as storage for wet clothes and bicycles (one [++] focus groups, UK5),2 4.

Improved design of cycling helmets might impact on their use and on cycling behaviour by children4.

Applicability: The findings from these studies are partially applicable as the findings are specific to schoolchildren. While some suggestions to encourage active travel are applicable to any population, schoolchildren and their parents face particular issues pertaining to safety and practicalities for this age group.

1 Ahlport et al. (2008)

2 Granville et al. (2002)

3 Black et al. (2001)

4 Stevenson and Lennie (1992)

5 Kirby (2008)

Evidence statement R2.ES18: Adult views about cycling for transport

Moderate evidence from five studies was available regarding barriers and facilitators to adult cycling for transport.

Benefits of cycling for transport were reported motivators, such as the ability to travel relatively quickly through traffic, the feeling of autonomy and freedom, and benefits for health and the environment. Cycling rather than driving could be encouraged by workplace initiatives.

Barriers to cycling were reported such as obstacles in the road, pollution and poor weather. Carrying bags and changes of clothing required after getting wet were also reported disincentives.

Cycling for transport requires negotiating space on the road; major barriers were traffic volume, inconsiderate driving and lack of adequate cycling tracks.

Some cycling behaviours were perceived as inappropriate by some other road users, giving cyclists a poor image and limited relationship with drivers.

Cycling was perceived as male, white and middle class. There was evidence that resistance to this image from female cyclists includes adopting and disseminating ideas for a feminine cycling image.

Reported benefits from commuting by bicycle included swiftness of travel through busy traffic, not having to rely on public transport, and improved fitness (for men) or body shape (for women). An additional factor was reassurance that the environment is being protected (one [+] interviews, UK1).

Parents were reported to drive less to work when cycling was encouraged by their workplace (one [+] survey, Australia2).

However, cyclists in the city report a number of obstacles that can interrupt the journey, such as poor road surfaces, manhole covers, glass, rough gutters, hilly terrain, parked cars and buses. In addition, pollution and bad weather can be a disincentive (one [++] interviews, UK3; one [+] survey and interviews, UK4). A (+) survey from Australia5 reported that women cyclists preferred off-road paths compared to roads with no facilities, and off-road paths compared to on-road lanes.

Commuting by cycle often involved carrying extra clothes to work and extra time at work to get changed from cycling outfits to work attire, including re-styling hair after wearing a helmet1. Lack of available facilities was a barrier to cycling, as were saddle-soreness and tiredness4.

Cycling on the road also requires negotiation with other road users. Cyclists reported fears of traffic and of accidents1 which meant having to be constantly alert for other traffic in order not to collide, and feeling vulnerable when crossing traffic to turn right3.

Cyclists reported feeling segregated and invisible on the road3. In areas where cycling is traditionally less prominent, there was a 'strangeness' about cycling, which was internalised by cyclists. There was also a perception that cycling is a male (predominantly white) activity, and some women felt the need to construct their own cycling identity, which could mean resisting the 'blokey' image and embracing femininity (for example, wearing heels while cycling; using blogs to reinforce identity)1.

Applicability: The findings from these studies are applicable to cyclists who commute in the UK and Australia. Differences in experiences between cycling populations (gender, ethnicity e.t.c.) and between settings in their promotion and support of cycling need to be taken into account.

1 Steinbach et al. (2011)

2 Wen et al. (2010)

3 McKenna and Whatling (2007)

4 Gaterslaben et al. (2007)

5 Garrard et al. (2008)

Evidence statement R2.ES19: Views about cycling identities

Moderate evidence from one (+) UK focus groups and interviews study1 that obtained car driver views of adult cycling identities.

Cycling for transport requires negotiating space on the road. Some cycling behaviours were perceived as inappropriate by some other road users, giving cyclists a poor image and limited relationship with drivers.

Car drivers reported being fearful of collisions, since cars and cycles travel at different speeds, and gave cyclists a wide berth. Some cyclists were reported as behaving poorly on the roads, for example passing through red lights, and this contributed for some, to cyclists having a negative image. Drivers that cycled were more likely to have empathy with cyclists on the road. Cycling proficiency testing, road taxes and compulsory helmet wearing were suggestions for improving the status of cyclists on the road.

Applicability: Findings from this study are applicable to car drivers in the UK. How cyclists are perceived by other road users and the impact that this may have for cyclists needs to be taken into account.

1 Granville et al. (2001)

Evidence statement EM.ES1: Led walking including 'walking school bus'

Moderate evidence from four studies suggests that led walking interventions (seven different interventions analysed in four studies) could be cost effective.

A Spanish study1: 6-month programme to promote walking-based exercise via a supervised exercise programme with three 50-minute sessions per week. Incremental cost per QALY range of €94– 871 per QALY.

A US study2: community-based social support strategies, including organised walking groups, home visits and phone calls, and newsletters, maps and handouts. Incremental cost per QALY of $27,373 and $39,690 for the two different led walking interventions versus do nothing.

A UK study3: organised community walking groups. The two organised walking group interventions showed a cost per QALY of £301 and £475.

Another UK study4:walking bus intervention designed to encourage schoolchildren to walk to school. Incremental cost per QALY estimated to be approx. £4007 per QALY gained.

The evidence is partially applicable to the UK, with two of the studies UK-based, and the other international studies concerning interventions that could be of UK relevance.

1 Gusi (2008)

2 Roux (2008)

3 Pringle (2010)

4 Fordham (2008)

Evidence statement EM.ES2: Pedometers

Moderate evidence from one Australian study1 suggests pedometer interventions could be cost effective: pedometer interventions, which used a meta-analysis of eight randomised control trials. Pedometer interventions maintained a net saving even when the intervention effect was modelled to decay completely by the end of the first year. That is, the modelled lifetime cost savings to the health service outweighed the pedometer costs as well as providing health benefits.

The evidence is partially applicable to the UK as similar pedometer interventions are of relevance.

1 Cobiac (2009)

Evidence statement EM.ES3: Media campaigns

Moderate evidence from one UK study1suggests media campaigns could be cost-effective: media campaigns circulating maps of walking and cycling routes. The cost-per-QALY of £86 for provision of a healthy living map with walking and cycling routes, and £288 for the promotion of walking and cycling through printed media.

The evidence is applicable to the UK.

1 Pringle (2010)

Evidence statement EM.ES4: Community health information (TravelSmart)

Moderate evidence from one Australian study1 suggests TravelSmart interventions could be cost-effective: TravelSmart intervention with individualised information to households on travel choices measuring change in the number of walking and cycling trips made per week. The TravelSmart programme resulted in a cost of $18,000 per disability-adjusted life year (DALY) assuming 50% decay per annum. The TravelSmart programme had net savings with annual decay rates of 0% and 25%, but costs rose to $41,000 per DALY at 75%, and $63,000 per DALY at 100% decay.

The evidence is partially applicable to the UK as the TravelSmart style intervention is relevant in the UK.

1 Cobiac (2009)

Evidence statement EM.ES5: Multi-component (Cycling Demonstration Towns)

Moderate evidence from one UK study1 suggests that the Cycling Demonstration Towns projects have a good benefit/cost rate.

The study: infrastructure measures such as the building of cycle paths, combined with a programme of education and marketing aimed at the general population. Benefits converted to monetary values and compared with the initial investment and running costs to produce a benefit–cost ratio. A range of 2.6–3.5 was given, reflecting the different approaches available for estimating accident and absenteeism benefits. Under all but the most pessimistic of scenarios considered, the benefit–cost ratio remained above one.

The evidence on cycle demonstration town is directly applicable as it was conducted in the UK.

1 Cope (2010)

Additional evidence

  • Expert paper 1 'Paving the way for everyday walking: 'Living Streets' interventions and public health'

  • Expert paper 2 'Making walking and cycling normal: key findings from the understanding walking and cycling research project':

  • Expert paper 3 'Programmes to promote cycling – evidence for NICE from CTC':

  • Expert paper 4 'Evidence to NICE PDG walking and cycling: experience from Bristol City Council and cycling city (2008–2011)': Ed Plowden, Bristol City Council

  • Expert paper 5 'Submission to the NICE programme development group on walking and cycling'

  • Expert paper 6 'Effectiveness of interventions to increase cycling'.

Economic modelling

Overall, all the interventions modelled were found to be highly cost effective, with each estimated to cost below £10,000 per quality-adjusted life years (QALYs) gained.

The economic model was constructed to incorporate, where possible, data from the reviews of effectiveness and cost effectiveness. In addition, it built on the relationship between:

  • physical activity and relative risk of mortality

  • levels of walking and cycling and overall physical activity

  • levels of walking and cycling and motorised travel (especially driving distance, but also driving time and number of trips).

Four interventions were modelled:

  • Two multi-component interventions (Cycling Demonstration Towns and SustainableTravel Towns).

  • Personalised travel advice (TravelSmart).

  • Use of pedometers.

  • Community-based led walks.

Health outcomes were expressed using QALYs gained and incremental net benefit (INB).

Wider impacts (environmental and traffic-related outcomes) were based on a limited selection of environmental outcomes and the value of a statistical life, expressed in terms of environmental benefit-cost ratios. The ratios were calculated based on the framework used by the Department for Transport.

It should be noted, that in the Department for Transport framework, most of the calculated benefits derived from health outcomes related to increased physical activity (up to 83%). However, health outcomes were excluded when calculating the environmental benefit-cost ratio, as they had been considered separately in a cost–utility analysis. Thus the results should be interpreted with caution.

A series of 'what if' analyses was undertaken to determine if the level of cost is justified for interventions producing a particular level of effect. In addition, the trade-off between narrow interventions with large effects per person were compared with wider interventions leading to smaller effects per person.

A number of assumptions were made which could underestimate or overestimate the cost effectiveness of the interventions. The results of the modeling were non-linear. The key factors influencing outcome were: threshold cost, level of effects, decay in effects and costs related to initial effects.

The specific scenarios considered and the full results can be found in 'Interventions to promote cycling and walking for recreational and travel purposes: Health economic and modelling report'.

  • National Institute for Health and Care Excellence (NICE)