Open Access Research

BMJ Open: first published as 10.1136/bmjopen-2015-007945 on 14 August 2015. Downloaded from http://bmjopen.bmj.com/ on June 2, 2023 at University College Chester. Protected by
What interventions increase commuter
cycling? A systematic review
Glenn Stewart, Nana Kwame Anokye, Subhash Pokhrel

To cite: Stewart G, ABSTRACT

Anokye NK, Pokhrel S. What Strengths and limitations of this study
Objective: To identify interventions that will increase
interventions increase
commuter cycling. ▪ Ours is the first study we know of to examine
commuter cycling? A
systematic review. BMJ Open Setting: All settings where commuter cycling might the evidence of interventions to increase com-
2015;5:e007945. take place. muter cycling, an activity that may potentially
doi:10.1136/bmjopen-2015- Participants: Adults (aged 18+) in any country. integrate physical activity into many people’s
007945 Interventions: Individual, group or environmental lives.
interventions including policies and infrastructure. ▪ Studies were included of any methodological
▸ Prepublication history Primary and secondary outcome measures: design with comparison groups and/or preinter-
and additional material is A wide range of ‘changes in commuter cycling’ vention and postintervention data but needed to
available. To view please visit indicators, including frequency of cycling, change in include cycle commuter outcomes.
the journal (http://dx.doi.org/ workforce commuting mode, change in commuting ▪ Despite wide variation between prevalence of
10.1136/bmjopen-2015- population transport mode, use of infrastructure by cycling between countries, robust evidence of
007945). what interventions will increase commuter
defined populations and population modal shift.
Results: 12 studies from 6 countries (6 from the UK, cycling is sparse.
Received 13 February 2015
Revised 11 July 2015 2 from Australia, 1 each from Sweden, Ireland, New
Accepted 20 July 2015 Zealand and the USA) met the inclusion criteria. Of
those, 2 studies were randomised control trials and the the global burden of coronary heart disease,
remainder preintervention and postintervention studies. 7% of type 2 diabetes, 10% of breast cancer

copyright.
The majority of studies (n=7) evaluated individual- and 10% of colon cancer.3 This is equivalent
based or group-based interventions and the rest to the global burden of mortality caused by
environmental interventions. Individual-based or group- tobacco.1 There is strong to moderate evi-
based interventions in 6/7 studies were found to dence for a 20–40% reduction in many long-
increase commuter cycling of which the effect was term conditions including cardiovascular
significant in only 3/6 studies. Environmental disease, coronary heart disease, stroke, type 2
interventions, however, had small but positive effects diabetes, osteoarthritis, breast and colon
in much larger but more difficult to define populations.
cancer and depression4 as the result of being
Almost all studies had substantial loss to follow-up.
physically active. Up to 70% of the National
Conclusions: Despite commuter cycling prevalence
Health Service (NHS) budget is estimated to
varying widely between countries, robust evidence of
what interventions will increase commuter cycling in be spent on long-term conditions.5
low cycling prevalence nations is sparse. Wider The WHO and at least 19 countries have
environmental interventions that make cycling issued physical activity guidelines.6 However,
conducive appear to reach out to hard to define but translating these guidelines into action has
larger populations. This could mean that environmental proven difficult; increasing technology and
interventions, despite their small positive effects, have proliferation of labour-saving devices includ-
greater public health significance than individual-based ing motor-vehicles have increasingly removed
or group-based measures because those interventions physical activity from everyday life, and elite
encourage a larger number of people to integrate sporting events such as the Olympics or the
physical activity into their everyday lives.
FIFA football world cup have not encouraged
mass participation.7
Environmental designs have been shown to
Health Economics Research
increase physical activity, particularly for stair
Group (HERG), Institute of INTRODUCTION use,8 playgrounds and transport to schools.9
Environment, Health and Physical activity is increasingly recognised as However, it is possibly unlikely that many will
Societies, Brunel University a fundamental prerequisite for maximal climb the stairs for bouts of 10 min as recom-
London, Uxbridge, UK health.1 Its opposite, physical inactivity, mended by guidelines4 or that travel to
Correspondence to
which has been described as a global pan- school will impact significantly on adult
Glenn Stewart; demic,2 was estimated to cause 9% of global behaviour, at least in the short term.
Glenn.Stewart@brunel.ac.uk premature mortality in 2008 as well as 6% of Organisations may have a financial interest

Stewart G, et al. BMJ Open 2015;5:e007945. doi:10.1136/bmjopen-2015-007945 1

Open Access

BMJ Open: first published as 10.1136/bmjopen-2015-007945 on 14 August 2015. Downloaded from http://bmjopen.bmj.com/ on June 2, 2023 at University College Chester. Protected by
in protecting and promoting the health of their employ- motorised transport to arrive at a location where sport/
ees, particularly if an intervention is free, but these are leisure cycling is desirable. The aim of this review there-
likely to only include those employees in organisations fore was to add to the article by Yang et al by identifying
that are either large enough and/or are inclined to and analysing interventions designed to increase com-
develop such plans.10 muter cycling, for example, that was intended for travel
As a year-round activity that is integrated into everyday from ‘place to place’. The specific review question was
life, active commuting would appear to be one means by therefore: what interventions increase commuter
which physical activity might be increased and main- cycling?
tained. In the UK, there are some 21.5 million people
who undertake a ‘regular commute to a fixed onshore
location’ for whom 64.2% have journeys under 10 km METHODS
and 42.9% have journeys under 5 km,11 a cycling dis- Data sources and search strategy
tance considerably less than the 8 km cited by the In October and November 2014, eight databases were
British Medical Association (BMA)12 that a ‘person can searched including Scopus, ERIC, CINAHL, the
easily cover’. While active commuting is largely defined Cochrane library, Digital Dissertations, Sports Discus,
to include both walking and cycling, it is cycling that has PsycINFO and Web of Science. Scopus is the largest ever
been found to be of sufficient intensity to meet the cri- bibliographic database and indexes over 20 000 titles
teria for Health-Enhancing Physical Activity (HEPA),13 from science, technology, medicine and the social
to require the 4 MJ(=955 kcal) weekly energy expend- sciences, is updated daily and contains both the Medline
iture needed to reduce all-cause and cardiovascular mor- and EMBASE databases.24
tality14 and to improve performance in untrained men In each database, a set of core key words were used:
and women. Despite this, levels of cycling vary widely cycl* OR bik* OR cycle AND hire OR active and
from approximately 1% in Australia, the UK and the commut* OR active AND travel* OR green AND
USA to 27% in the Netherlands.15 commut* OR green AND transport* OR green AND
A number of reviews have therefore sought to travel* OR ecological AND commut* OR ecological
examine the evidence of how the prevalence of physical AND transport* OR non-motor* OR non-auto. On the
activity can be increased through cycling.16–20 However, returned titles, database-specific filters were then
where these reviews seek to identify how to increase the applied to narrow the search. Table 1 presents the ori-

copyright.
prevalence of cycling, in general this may include ginal hits in each database on core keywords search and
cycling for transport, leisure, recreation, health or sport. how attrition happened as the database-specific filters
It may be that commuter cycling—defined following the were applied.
European Network for Cycling Expertise as journeying The final set of titles after applying relevant filters
for the sake of completing a journey as opposed to a (n=9825) were then imported to RefWorks reference
journey that is an end in itself21—is more likely to be manager software. Duplicates (n=492) were removed to
continued and to have sustainable health gains. obtain 9333 titles ready for screening.
Increased commuter cycling may also have the add-
itional benefits of avoiding the external costs of Study attributes required for inclusion ( population,
motorised transport including injuries, pollution intervention, comparators and study designs)
(including noise pollution) and community severance.22 Outcome variable
In an age of austerity, commuter cycling has further Commuter cycling may be affected by a wide range of
monetary implications for the individual and their factors such as urban planning, congestion, safety and
family. perception of safety, pollution, petrol, pricing, etc.25
The Government recognises that increasing physical This can lead to the ‘inverse evidence law’ whereby that
activity requires ‘weaving incidental activity into our which may have the most effect on health is precisely
daily lives’ including using bicycles for transport.23 This that which is least measured due to methodological con-
was echoed in the recent NHS Five Year Forward View siderations.26 The application of strict inclusion criteria
that there needs to be a ‘radical upgrade in prevention adopted by Ogilvie et al17 or Yang et al20 was therefore
and public health’.5 Policymakers will find implementa- precluded in favour of the more pragmatic approach
tion of this ambition difficult unless there is good-quality adopted by Pucher et al.16 To be included, studies
evidence of interventions that have examined evidence needed to report specifically on commuter cycling as a
of interventions to increase commuter cycling. The dependent variable (rather than walking and cycling
nearest such review appears to have been by Yang et al.20 combined) and indicate quantifiable changes in com-
While Yang et al sought to determine what interventions muter cycling rather than modelling or stated prefer-
are successful in promoting cycling for any purpose, this ence studies or other variables that may influence levels
may not be ideal from a public health perspective. of cycling but are not direct measurements such as per-
Cycling for sport or leisure need not become habitual in ceived safety, opinion, confidence on a bicycle or atti-
the way that commuting is largely a necessary activity tude towards active transport. This, however, included a
and indeed may increase external costs as people use wide range of outcome measure such as changes at an

2 Stewart G, et al. BMJ Open 2015;5:e007945. doi:10.1136/bmjopen-2015-007945

 Open Access

BMJ Open: first published as 10.1136/bmjopen-2015-007945 on 14 August 2015. Downloaded from http://bmjopen.bmj.com/ on June 2, 2023 at University College Chester. Protected by
Table 1 Search strategy and results
Initial number of hits
with core keywords Final number of hits
Database search Attrition as database-specific filters was applied after applying all filters
Scopus 845 Filter not used 845
ERIC 11 847 Limit to: academic journals—4879 1789
Limit to: higher education, postsecondary education,
case studies, intervention, program effectiveness—
1789
CINAHL plus 43 324 Limit to: academic journals—37 745 1359
Limit to: cycling—2414
Limit to: adult—1359
Cochrane library 39 218 Limit to: reviews—377 377
Digital Just over 1 million Limit to: scholarly journals—326 607 2938
Dissertations Limit to: American J of PH, Social Research, Health
Affairs—2938
PsycINFO 56 448 Limit to: academic journals—47 213 196
Population—human—11 366
Subject—health, physical activity—196
Sports Discus 95 808 Limit to: academic journals—23 166 615
Subject thesaurus term—cycling, exercise, physical
fitness, cyclists, prevention—1494
Subject—males, comparative studies, young adults,
evaluation, adulthood, women, teenagers, research,
middle age, case studies—615
Web of Science Just over 1 million Limit to: engineering, behavioural sciences, public 827
environmental occupational health, sports sciences,
healthcare sciences services, sociology—371 085
Research domains—restrict to behavioural

copyright.
sciences, public environmental occupational health,
sports sciences, healthcare sciences services,
sociology, life sciences, biomedicine other topics,
social issues, social sciences other types—194 675
Research areas—restrict to behavioural sciences,
public environmental occupational health,
engineering, social sciences other topics, urban
studies, transportation—144 459
Limit to: articles—125, 612
English—114 955
Limit to: transportation, urban studies—827
Total databases Total initial hits=over 2 Total after applying filters=9825 Total imported to
searched=8 million RefWorks=9825

aggregate population level, number of days cycling, dis- Identification of studies

tance, time taken cycling and time at follow-up. In level 1 screening, the titles, abstracts and keywords of
9333 non-duplicates were screened according to the cri-
Eligibility criteria teria set out in table 2 by the lead author (GS). This
To be included in the review, studies needed to include screening excluded a large number of studies (n=9267),
comparison groups and/or preintervention and postinter- leaving only 66 studies to be retrieved for full-text
vention data, include adults rather than schoolchildren, screening. In level 2 screening, all 66 full texts were
include data on cycling rather than aggregated data of screened independently by two authors (GS and SP) by
walking and cycling, include data relating to commuting applying the eligibility criteria. Disagreement (n=2/66)
to work and be written in English (table 2). All study was settled by asking the opinion of the third author
designs (except correlation studies identifying determi- (NKA) and reaching a consensus thereafter. This screen-
nants of commuter cycling) were eligible, as the intention ing process led to 54 studies not being eligible, and
was to capture wider public health interventions, noting therefore they were excluded from further review. The
that often such evaluations are not limited to robust study reasons for exclusion were: correlation studies (n=16);
designs such as the randomised controlled trials. no pre–post data (n=11); did not provide the outcome

Stewart G, et al. BMJ Open 2015;5:e007945. doi:10.1136/bmjopen-2015-007945 3

Open Access

BMJ Open: first published as 10.1136/bmjopen-2015-007945 on 14 August 2015. Downloaded from http://bmjopen.bmj.com/ on June 2, 2023 at University College Chester. Protected by
Table 2 Eligibility criteria
Inclusion criteria Exclusion criteria
▸ Evaluation studies with comparison groups and/or ▸ Correlation studies (identifying determinants of commuter
preintervention and postintervention data cycling)
▸ Adults rather than schoolchildren ▸ No comparison groups or pre–post data available
▸ Data relating to commuting to work ▸ Did not provide outcome data in the format needed for this
▸ Written in English review
– Indiscriminate data (eg, has only aggregated data of
walking and cycling)
– Irrelevant data (has data only on other forms of commuting,
eg, walking only or cycling for recreation)
▸ Non-evaluation (eg, editorials, commentaries, opinion pieces)
▸ Others, eg
– Temporal/trend analysis of cycling behaviour
– Reviews of correlation studies
– School children
– Written in a language other than English

data in the format needed for this review (n=14); generalisability of included evidence: (1) the population
non-evaluation articles (n=4); and others (n=9). size where the claimed effect was observed/measured;
A total of 12 studies were thus identified as eligible for (2) the robustness of the comparator; (3) the extent to
full review. Figure 1 depicts a PRISMA diagram of the which the interventions being evaluated were able to
study identification and inclusion process. increase commuter cycling prevalence and (4) the
robustness of the study design. The information
Quality appraisal retrieved through data extraction tables, coupled with
The first author (GS) extracted data on the 12 included assessments in the quality appraisal checklist, was used
to inform each of the above four domains. As interven-

copyright.
studies using a predefined data extraction table, informed
by the Centre for Reviews and Dissemination guidelines27 tions were a mixture of environmental and individual-
and National Institute for Health and Care Excellence focused/group-focused measures, no meta-analysis of
(NICE)28 (see online supplementary appendix 1). The data was attempted. As such, we chose to present the
first author (GS) developed the data extraction table and findings as a narrative synthesis.
piloted it by extracting data on one study. This was
reviewed by the other two authors before finalising it. RESULTS
Two authors (GS and NA) then independently applied Study characteristics
a quality checklist extracted from NICE’s public health Of the 12 included studies, 6 were from the UK, 2 from
guidance methods manual28 on 25% (n=3/12) of Australia, 1 each from Sweden, Ireland, New Zealand
included studies. This checklist covered several questions and the USA. Of those, two studies were randomised
enabling the reviewers to judge the quality of each study. control trials (RCT)29 31 and the remainder preinterven-
A quality rating was also given for all included studies as tion and postintervention studies. The majority of
per the checklist guidelines. There was no disagreement studies (n=7) evaluated individual-based or group-based
in overall quality ratings between the two reviewers on interventions and the rest environmental interventions.
those 3/12 studies. However, minor disagreement Table 3 includes the characteristics of included studies.
emerged in answering some questions in the checklist—
which was settled by seeking the opinion of the third Individual or group interventions
author and reaching a consensus. Then the first author The two RCTs included in this review evaluated individ-
completed quality assessment on the remaining studies. ual interventions that were based on provision of written
Of the 12 included studies, 1 study29 was given a “++” information or advice and a bicycle by health profes-
rating (ie, study designed to minimise risk); 3 studies30–32 sionals to encourage cycling. Mutrie et al,31 a Scottish
were given a “+” rating (ie, potential sources of bias not study, aimed to increase active commuting among 295
addressed in the study or not clear from the way the study employees at three workplaces in Glasgow.
was reported); and 8 studies were given a “−” rating (ie, Hemmingsson et al,29 a Swedish study, intended to
study with significant sources of bias). increase levels of physical activity through a support pro-
gramme involving three aims (awareness raising, coun-
Data analysis tering and helping relationships) in obese women. The
The data were analysed qualitatively, considering four Scottish study was based on the transtheoretical model
key attributes underlying the robustness and of behaviour change providing self-help materials to

4 Stewart G, et al. BMJ Open 2015;5:e007945. doi:10.1136/bmjopen-2015-007945

 Open Access

BMJ Open: first published as 10.1136/bmjopen-2015-007945 on 14 August 2015. Downloaded from http://bmjopen.bmj.com/ on June 2, 2023 at University College Chester. Protected by
copyright.
Figure 1 PRISMA diagram of the study identification and inclusion process (Digital Dissert, Digital Dissertations).

people either thinking about active commuting (con- physical activity prescriptions, group counselling and
templators) or irregular commuters ( preparers). bicycles compared with the control group which
Follow-up at 6 months found that while 18% more parti- received low-intensity group support and pedometers at
cipants in the intervention group (compared with the 6 months. The mean proportion of intervention group
control group) moved to a higher stage of active com- participants (n=60) commuting with a bicycle at least
muting behaviour, the significant difference was largely once a week during months 2–18 was 29.4% compared
attributed to walking and not cycle commuting. Only 18 with 8% in the control group.
people (of the 295) were cycling with no difference Workplace travel plans that seek to encourage active
between the intervention and control groups. The travel programmes have been promoted as having advan-
Swedish study, however, reported an effective interven- tages to the employee (health) and employer (lower
tion. They compared two different support programmes absenteeism). These may include ‘bike to work’ (BTW)
to increase physical activity through active commuting. initiatives. Three studies examined the effects of such
The intervention group received physician meetings, programmes in England (Bristol),33 Australia34 and New

Stewart G, et al. BMJ Open 2015;5:e007945. doi:10.1136/bmjopen-2015-007945 5

 6

Open Access
Table 3 Characteristics of and results from included studies
Serial
number Study Country/setting Intervention Study design Time period Sample size Effect
1 Brockman and England (University of Workplace travel plan(s) Preintervention and Surveys from 1950–2829 Percentage usually cycling to work
fox33 Bristol) postintervention 1998 to 2007 increased from 7% to 12% but was not
significant
2 Hemmingsson Sweden/community Support programme RCT 18 months 120 Proportion of participants cycling >2 km/
et al29 (awareness raising, day was 38.7% (OR 7.8)
countering and helping
relationships)
3 Mutrie et al31 Scotland/Hospital Workplace; self-help pack RCT 6 months 295 people No effect
Trust, Health Board including maps, activity identified as
and University diary, safety accessories thinking about
active travel
4 Telfer et al36 Australia (Sydney) Cycle proficiency training Preintervention and 2 months 113 No difference in mean frequency or
postintervention duration of cycle trips
5 O’Fallon35 New Zealand— Number of workplace Preintervention and 12 months 3825 675 respondents to cycle question—112
number of workplaces interventions postintervention cycled less (16.6%), 347 (51.4%) about
the same and 216 (32.0%) more
Stewart G, et al. BMJ Open 2015;5:e007945. doi:10.1136/bmjopen-2015-007945

6 Johnson and London—community Cycle training Preintervention and 12 months 130 Number of days cycled to work in the last
Margolis37 setting postintervention week increased from 0.66 to 1.33
7 Caulfield39 Ireland—Dublin Whole city approach Preintervention and 5 years Dublin population Percentage of cyclists increased from 4%
postintervention 1.2 million to 5% (20 588 to 26 670)
8 Rose and Australia (Victoria) Ride To Work Day Preintervention and 5 months 5577 27% of first-time riders still cycling to
Marfurt34 postintervention work after 5 months
9 McCartney Scotland—Glasgow Building a bridge Preintervention and 2007–2010 216 897 people 47.5% increase in the number of cyclists
et al38 postintervention living south of city (n=approximately 400)
centre
10 Goodman et al30 England—Cycling Whole city approaches Preintervention and 2008–2012 1 266 337 0.69 percentage point increase in cycling
Cities and Towns postintervention to work in intervention towns, compared
initiative (12 locations) with matched towns
11 Goodman et al32 England (3 cities/ Changes in walking and Preintervention and 2010–2012 22 500 (residents At 2 year follow-up, 18% of people who
towns) cycle infrastructure postintervention within 5 km of knew about project reported transport
projects cycling compared with 7% of full sample
12 Krizek et al40 US—Minneapolis Changes in cycle Preintervention and 1990–2000 4855 0.493 percentage point increase in
infrastructure postintervention bicycle modal share
RCT, randomised control trial.

copyright.
BMJ Open: first published as 10.1136/bmjopen-2015-007945 on 14 August 2015. Downloaded from http://bmjopen.bmj.com/ on June 2, 2023 at University College Chester. Protected by
 Open Access

BMJ Open: first published as 10.1136/bmjopen-2015-007945 on 14 August 2015. Downloaded from http://bmjopen.bmj.com/ on June 2, 2023 at University College Chester. Protected by
Zealand.35 Results were again mixed; in Bristol, the analysis indicated that compared with matched towns,
University travel survey indicated a non-significant rise in cycling to work in the intervention towns increased by
the cycle commuting modal share from 7% in 1998 to 0.69 percentage points with larger differences in differ-
11.8% following the implementation of a workplace ences compared with the unfunded comparison group
travel plan. In Australia, in 2004, 5577 people registered (1.02 percentage points) and the national comparator
for the Ride to Work Day event of whom baseline data was (1.23 percentage points). Although seemingly small in
received from 1952 (35%). Of these, 17% indicated that effect, the size of the sample population in the 12 towns
they had not cycled to work before the event. At (1 266 337 in the 2011 census) indicates that the abso-
5 months postevent, 27% of first-timers were still cycling lute percentage change in cycle commuting (0.97%)
to work (defined as at least once a week) compared with may have a large public health significance.30
67% of those who had been cycling to work before the In Ireland, the Department of Transport set targets of
event. In New Zealand, 40 organisations were originally increasing cycling from 2% of journeys in 2009 to 10%
recruited to the ‘Bike Now’ programme of which 27 by 2020. In Dublin, commuter cycling was hypothesised
(675 workers) remained in the programme at 1 year. Of to result from financial incentives (tax-free loans to pur-
these, 112 (16.6%) of 675 respondents indicated that chase cycles), infrastructure change (traffic calming,
they were cycling less, 347 (51.4%) about the same and cycle lanes including segregated lanes), promotional
216 (32.0%) more. None of the above included a events such as Bike week (family rides, removing traffic
control group. from streets, repair clinics and promotion talks), a
Two studies examined the effect of cycling training on shared bike scheme and publication of the first design
cycling to work. Results were not consistent; in Sydney, a standards for cycling in Ireland.39 Census data indicated
telephone 1-week recall interviews found no difference in that results were equivocal; cycle modal share fell from
either duration or frequency of cycling at 2 months 6% in 1996 to 4% in 2002 and 2006 but had risen to 5%
(including number of days cycled to work) following a in 2011. In Cork, cycle modal share fell from 2% in
cycling proficiency training programme (n=110) 1996 to 1% from 2002 onwards, whereas in Galway it fell
although statistically significant increases in those who from 3% in 1996 to 2% from 2002. However, it is not
did not cycle before the course were found.36 In London, clear as to what extent the 2008 financial crisis in
3-month postintervention questionnaires found that the Ireland might have affected the results.
mean number of days cycled to work increased from 0.66 In three UK cities/towns, traffic-free infrastructure (a

copyright.
to 1.33 in the past week.37 Neither study included a main project plus feeder routes) were evaluated for
control group. Loss to follow-up from the London study their effects on residents living within 5 km of the
was high (104 responses from 471 participants). respective projects.32 A total of 22 500 survey packs were
distributed to which 3516 people replied of whom 53%
Environmental interventions and 43% provided data at the 1-year and 2-year follow-
Environmental interventions were either a relatively ups, respectively (excluding those who had moved
small single intervention (eg, the construction of a house). Respondents were asked if they had cycled on
bridge38) or a larger programme such as the English the infrastructure for six journey purposes including
Cycling Cities and Towns (CCT) initiative that targeted commuter cycling. At the 2-year follow-up, 18% of
12 cities and towns with some 2.7 million residents over people who knew about the project reported transport
3 years30 or several policies taken together on cycle com- cycling compared with 7% of the full sample. However,
muter prevalence.39 the multivariate statistical analysis presented in the study
The opening of a bridge in Glasgow was associated provided data on infrastructure use for any purpose
with a 47.5% increase in the number of cyclists entering rather than commuter cycling.
the city centre from the South with almost no change in One US study40 assessed the effects of transport/cycle
numbers of cyclists crossing other bridges. Some of this infrastructure on cycle commuting. Cycle commuter
change may have been accounted for by road works modal share increased in central Minnesota (from 2.8%
associated with the construction of the M74 which was to 3.3% at the University of Minnesota (n=4855)) and
not controlled for.38 Minneapolis (from 0.788 to 0.841, n=21 111) where
The English CCT programme aimed to increase cycle facilities had been implemented or improved, com-
cycling through capital and revenue investment provided pared with the suburbs where cycle commute share fell
through competitive tendering to the respective CCTs. from 0.335% to 0.279% (n=9016). This study, however,
Changes in cycle commuting between 2001 and 2011 in was not immune to other external influences and, as
the CCTs were compared with changes in matched acknowledged, the ‘Lance Armstrong effect’ may have
towns using a ‘difference in difference’ analysis. been present at the time.
Controls were either statistically matched towns, towns
that had applied unsuccessfully for funding or a
non-London national comparison group (all non- DISCUSSION
intervention urban areas outside London with a popula- This review improves our understanding as to what inter-
tion of over 30 000). The ‘difference in difference’ ventions are likely to increase commuter cycling, an

Stewart G, et al. BMJ Open 2015;5:e007945. doi:10.1136/bmjopen-2015-007945 7

Open Access

BMJ Open: first published as 10.1136/bmjopen-2015-007945 on 14 August 2015. Downloaded from http://bmjopen.bmj.com/ on June 2, 2023 at University College Chester. Protected by
activity that may potentially integrate physical activity This review has its own limitations which stem largely
into many people’s lives. There is wide agreement that from major weaknesses of the included studies.
increased population levels of physical activity would Evaluation of real-world interventions where variables
make substantial improvements to health. Intuitively, the cannot be controlled by researchers can be challenging,
potential for commuter cycling to at least partially meet as found in the case of all included studies. Studies suf-
this need is large. In Europe, most car journeys are less fered from high dropout rates (almost all studies had
than 5 km, most people can cycle, the financial costs are substantial loss to follow-up) and used numerous mea-
small, and for the individuals, financial savings would be sures of outcome variables (commuter cycling) which
made against the cost of motorised or public transport. were difficult to compare meaningfully. Most studies suf-
There is also evidence of where this has been achieved fered from the lack of robust comparison groups,
—in countries such as Holland, Germany and Denmark, leading to a less robust alternative of measuring out-
a substantial proportion of journeys are by bicycle, comes preintervention and postintervention instead.
including those by people aged 65+.15 Despite all this, Significant aspects of potential bias in those studies
this review highlights how little robust research evidence therefore cannot be ruled out. Put together, this review
exists on what may increase commuter cycle prevalence was limited to a narrative synthesis of evidence rather
in low cycling nations. than a more robust quantitative meta-analysis.
Of the two RCTs included in the review, only one Finally, the paucity of high-quality evidence found in
found evidence of effectiveness but in a select population this review may highlight an important methodological
of obese women—which may therefore lack external val- issue related to the review itself. As noted by Pucher
idity for the wider population. The second RCT found no et al,16 interventions that might affect cycling prevalence
evidence of effectiveness even in people assessed, are many and varied but few may be published. By
through the transtheoretical model of behaviour change, restricting its focus to commuter cycling, this review may
as at least ready to begin to change. Other individual- have excluded a number of interventions that increased
based or group-based studies that assessed the effective- general and commuter cycling. Given the potential for
ness of interventions on populations opting into health gain, the lack of robust evidence on effective
programmes found either no or small effect sizes, again interventions may be disappointing for policymakers.
indicating a lack of evidence for their applicability at a More research is therefore needed to fill in this import-
population level. This review also found that many studies ant gap as well as to further our understanding as to

copyright.
did not include relevant control groups and had high how lessons learnt in high cycling-prevalent countries
rates of loss to follow-up, indicating a high risk of bias and can be applied to other countries.
that the effect of external factors cannot be discounted.
Wider environmental interventions perhaps have the
potential to have a greater though more dilute effect
CONCLUSION
over a greater proportion of the population. Evaluation
Despite its potential to increase health, there is little
of environmental interventions includes several meth-
robust evidence of effective interventions to increase
odological issues that have been well documented,41
commuter cycling even at a subpopulation level. Many
including defining what may be described as the denom-
studies lack appropriate controls, their external validity
inator population (working population, working popula-
to the wider population remains unclear, and they have
tion from a particular area or the whole population). As
high rates of loss to follow-up—all indicating a high risk
included studies did have some of these issues present
of bias. Wider environmental interventions that make
in their design, there is a clear paucity of evidence of
cycling conducive appear to reach out to hard to define
effectiveness. NICE guidance acknowledges that a range
but larger populations. This could mean that environ-
of factors may be important in helping or restricting
mental interventions, despite their small positive effects,
people from cycling42 and that may reflect the challenge
have greater public health significance than individual-
of rigorous evaluation in this field. However, it is also
based or group-based measures because those interven-
noted that even small changes at a population level can
tions encourage a larger number of people to integrate
have significant effects and therefore important implica-
physical activity into their everyday lives. More research
tions for population health. For example, the 0.69 per-
is needed to establish how prevalence of commuter
centage point differential effect of an environmental
cycling can be increased.
intervention in England implies that over 8000 people
started commuter cycling, following the implementation Acknowledgements The authors would like to thank Professor Christina
of CCT. Therefore, environmental interventions, despite Victor and Dr Geraldine Barrett for their critical comments earlier in the review
process. They hugely benefited from the comments made by the two referees
showing small effect sizes, appear to have more public
who reviewed the manuscript on first submission.
health significance than individual-based/group-based
Contributors This study is a part of GS doctoral thesis. GS conceptualised,
interventions, as they will reach out to many more
designed and executed the study with significant inputs from NKA and SP
people (though often harder to define populations) to throughout. GS prepared the first draft of this manuscript, which was
encourage integration of physical activity into everyday commented on and revised by NKA and SP. All authors have read and
life via commuter cycling. approved the manuscript.

8 Stewart G, et al. BMJ Open 2015;5:e007945. doi:10.1136/bmjopen-2015-007945

 Open Access

BMJ Open: first published as 10.1136/bmjopen-2015-007945 on 14 August 2015. Downloaded from http://bmjopen.bmj.com/ on June 2, 2023 at University College Chester. Protected by
Funding This research received no specific grant from any funding agency in 20. Yang L, Sahlqvist S, McMinn A, et al. Interventions to promote
the public, commercial or not-for-profit sectors. cycling: systematic review. BMJ 2010;341:c5293.
21. Aldred R. A Matter of Utility? Rationalising Cycling, Cycling
Competing interests None declared. Rationalities. Mobilities. 2014. doi:10.1080/17450101.2014.935149
22. Mindell JS, Watkins SJ, Cohen JM. Health on the Move 2. Policies
Provenance and peer review Not commissioned; externally peer reviewed. for health promoting transport. Stockport: Transport and Health
Study Group, 2011.
Data sharing statement No additional data are available.
23. Public Health England. Everybody active, every day. An evidence
Open Access This is an Open Access article distributed in accordance with based approach to physical activity. Public Health England, 2014.
24. Brunel University. Databases Webpage. Site accessed 1st
the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license,
November 2014.
which permits others to distribute, remix, adapt, build upon this work non- 25. Heinen E, van Bee B, Kees M. Commuting by bicycle: an overview
commercially, and license their derivative works on different terms, provided of the literature. Transport Rev 2010;30:59–96.
the original work is properly cited and the use is non-commercial. See: http:// 26. Ogilvie D, Egan M, Hamilton V, et al. Systematic reviews of health
creativecommons.org/licenses/by-nc/4.0/ effects of social interventions: 2. Best available evidence: how low
should you go? J Epidemiol Community Health 2005;59:886–92.
27. Centre for Review and Dissemination. Systematic Reviews: CRD’s
Guidance for Undertaking Reviews in Health care. 2009.
REFERENCES 28. NICE. Methods for the development of NICE public health guidance
1. Das P, Horton R. Rethinking our approach to physical activity. (third edition): Appendix F Quality Appraisal Checklist—quantitative
Lancet 2012;380:189–90. intervention studies. 3rd edn. National Institute for Health and Care
2. Kohl HW, Craig CL, Lambert EV, et al. The pandemic of physical Excellence, 2012.
inactivity: global action for public health. Lancet 2012;380:294–305. 29. Hemmingsson E, Uddén J, Neovius M, et al. Increased physical activity
3. Lee IM, Shiroma EJ, Lobelo F, et al. Effect of physical inactivity on in abdominally obese women through support for changed commuting
major non-communicable diseases worldwide: an analysis of burden habits: a randomized clinical trial. Int J Obes 2009;33:645–52.
of disease and life expectancy. Lancet 2012;380:219–29. 30. Goodman A, Panter J, Sharp SJ, et al. Effectiveness and equity
4. Department of Health. Start Active, Stay Active. A report on physical impacts of town-wide cycling initiatives in England: a longitudinal,
activity for health from the four home countries’ Chief Medical controlled natural experimental study. Soc Sci Med 2013;97:
Officers. 2011. 228–37.
5. National Health Service. Five Year Forward plan. 2014. 31. Mutrie N, Carney C, Blamey A, et al. “Walk in to Work Out”: a
6. IHRSA. Physical Activity Guidelines International. http://download. randomised controlled trial of a self help intervention to promote
ihrsa.org/publicpolicy/Physical_Activity_Guidelines_INTL.pdf active commuting. J Epidemiol Community Health 2002;56:407–12.
7. Hallal PC, Bauman AE, Heath GW, et al. Physical activity: more of 32. Goodman A, Sahlqvist S, Ogilvie D. Who uses new walking and
the same is not enough. Lancet 2012;380:190–1. cycling infrastructure and how? Longitudinal results from the UK
8. Foster C, Hillsdon M. Changing the environment to promote iConnect study. Prev Med 2013;57:518–24.
health-enhancing physical activity. J Sports Sci 2004;22:755–69. 33. Brockman R, Fox KR. Physical activity by stealth? The potential
9. Reynolds R, McKenzie S, Allender S, et al. Systematic review of health benefits of a workplace transport plan. Public Health
incidental physical activity community interventions. Prev Med 2011;125:210–16.
2014;67:46–64. 34. Rose G, Marfurt H. Travel behaviour change impacts of a major ride

copyright.
10. Hosking J, Macmillan A, Connor J, et al. Organisational travel plans for to work day event. Transportation Res Part A 2007;41:351–64.
improving health. Cochrane Database Syst Rev 2010;3:CD005575. 35. O’Fallon C. Bike Now: Exploring methods of building sustained
11. Office for National Statistics. 2011 Census Analysis—Distance participation in cycle commuting in New Zealand. Road and
Travelled to Work, 2014. Transport Research 2010;19(2).
12. British Medical Association. Healthy Transport=Healthy Lives. 2010. 36. Telfer B, Rissel C, Bindon J, et al. Encouraging cycling through a
13. Oja P, Vuori I, Paronen O. Daily walking and cycling to work: their pilot cycling proficiency training program among adults in central
utility as health-enhancing physical activity. Patient Educ Couns Sydney. J Sci Med Sport 2006;9:151–6.
1998;33:S87–94. 37. Johnson R, Margolis S. A review of the effectiveness of adult cycle
14. Shephard RJ. Is active commuting the answer to population health? training in Tower Hamlets, London. Transport Policy
Sports Med 2008;38:751. 2013;30:254–61.
15. Pucher J, Buehler R. Making cycling irresistible: lessons from the 38. McCartney G, Whyte B, Livingston M, et al. Building a bridge,
Netherlands, Denmark, and Germany. Transport Rev 2008;28. transport infrastructure and population characteristics: explaining
16. Pucher J, Dill J, Handy S. Infrastructure, programs, and policies to active travel into Glasgow. Transport Policy 2012;21:119–25.
increase bicycling: an international review. Prev Med 2010;50(Suppl 39. Caulfield B. Re-cycling a city—examining the growth of cycling in
1):S106–25. Dublin. Transportation Res Part A 2014;61:216–226.
17. Ogilvie D, Egan M, Hamilton V, et al. Promoting walking and cycling 40. Krizek K, Barnes G, Thompson K. Analyzing the effect of bicycle
as an alternative to using cars: systematic review. BMJ facilities on commute mode share over time. J Urban Plann Dev
2004;329:763–6. 2009;135:66–73.
18. Fraser SD, Lock K. Cycling for transport and public health: 41. Petticrew M, Cummins S, Ferrella C, et al. Natural experiments:
a systematic review of the effect of the environment on cycling. Eur J an underused tool for public health? Public Health 2005;119:751–7.
Public Health 2011;21:738–43. 42. National Institute for Health and Clinical Excellence (NICE). Walking
19. Krizek KJ, Forsyth A, Baum L. Walking and Cycling International and cycling: local measures to promote walking and cycling as forms
Literature Review. Final Report. 2009. of travel or recreation 2012.

Stewart G, et al. BMJ Open 2015;5:e007945. doi:10.1136/bmjopen-2015-007945 9