key: cord-0427334-cz2tb10l
authors: Pearson, L. K.; Berkovic, D.; Reeder, S.; Gabbe, B. J.; Beck, B.
title: Adults' self-reported barriers and enablers to riding a bike for transport: a systematic review
date: 2022-04-22
journal: nan
DOI: 10.1101/2022.04.22.22274159
sha: 69f395c108229c731c238678fab3b408ea629edd
doc_id: 427334
cord_uid: cz2tb10l

Riding a bike for transport purposes is an effective way to improve population and environmental health. Despite this, participation levels in many countries are low. Identifying the barriers and enablers to riding a bike for transport is essential to developing interventions that encourage bike riding. In this mixed-methods systematic review, we aimed to identify the perceived barriers and enablers to adults riding a bike for transport in Organisation for Economic Development (OECD) countries. A systematic database search was conducted to identify relevant peer-reviewed and grey literature. Fifty-five papers/reports met eligibility criteria. There were 34 barriers and 21 enablers identified. The leading barriers related to riding on the road alongside motor vehicles. Other factors identified included the provision and quality of cycling infrastructure, personal factors such as physical fitness, attitudinal factors such as community perceptions of cyclists, and environmental factors. While this review highlights the complexity of factors that influence the uptake of riding a bike for transport, many of the leading factors could be overcome through the provision of high-quality protected infrastructure for bike riders. Other interventions to address other known barriers and enablers are needed to increase the uptake of bike riding.

The health benefits of riding a bike are vast and well established, including 40 improved cardiorespiratory fitness, cognitive function and mental wellbeing (Celis-41 Morales et al., 2017; Leyland et al., 2019; Nordengen et al., 2019) . Beyond this, an 42 increase in trips made by bike has the potential to reduce traffic congestion, 43 environmental pollutants, and offer economic benefits to both individuals and 44 governments (Litman, 2015; Pérez et al., 2017; Rabl & De Nazelle, 2012) . A large 45 proportion of trips made by car are a distance that could potentially be made by bike 46 (Beck, 2021; Department of Transport, 2021; Harper, 2021; Vagane, 2007) , 47 highlighting the potential for increased participation in bike riding for transport, in 48 addition to for recreational purposes. 49 50 Despite the benefits of bike riding, and increased accessibility from e-bikes 51 (Fishman & Cherry, 2016; Fyhri et al., 2017) , moving away from car-based transport 52 and towards bike riding has become a considerable cultural, behavioural and political 53 challenge (Agarwal & Gupta, 2021; Kent, 2015; Richards et al., 2011; Wilson & Mitra, 54 2020) . Recognising what facilitates, and what prevents people from choosing to cycle 55 for transport permits informed decisions for cycling infrastructure, promotion 56 campaigns and policy reform for government and advocacy organisations. 57

There are several studies that report the barriers and enablers of riding a bike for 59 transport in local populations, cited regularly throughout active transport-related 60 literature. However, these study findings are often extrapolated to populations with 61 potentially different circumstances. Further, there are a variety of measures and 62 populations reported, limiting the potential for comparison between studies. 63

There is, at present, no peer-reviewed systematic review that synthesises the 65 findings of, and appraises the quality of studies, from multiple contexts of the barriers 66 and enablers to riding a bike for transport. This review is important because it allows for 67 identification of potential issues in measurement of barriers and enablers, areas of 68 needed research, and consistencies and differences between countries. 69 70

The aim of this systematic review was to identify the perceived barriers and 71 enablers to adults' participation in riding a bike for transport. The inclusion of 72 qualitative research in this systematic review attempted to represent some of these lived 73 experiences while still maintaining large-scale quantitative studies to facilitate 74 generalisability and reproducibility. 75

Systematic searches of databased of peer-reviewed scientific and grey literature 77 were undertaken to identify the barriers and enablers of riding a bike for transport in 78

Organisation for Economic Co-operation and Development (OECD) countries (OECD, 79 2020) . Both quantitative and qualitative studies were included, and no publication year 80 limits were applied due to the limited published synthesis of data prior to this review. 81

Trips made by bike exclusively for transport purposes were chosen as they have likely 82 differing barriers and enablers than when trips are made for recreation. Transport 83 included any journey made to travel to a destination, such as commuting and to travel to 84 shops. Most OECD member countries are high-income and developed economies 85 (Canser et al., 2016) . Because of this, they share similar cycling infrastructure, laws and 86 volumes of traffic (Buehler & Pucher, 2017) . Due to these similarities, only OECD 87 categories of codes. Findings of qualitative and quantitative analyses were reported 185 alongside each other where findings overlapped. 186

Quantitative data 187

All quantitative data regarding study characteristics were extracted into a table. 188

Outcome data were extracted into two separate tables exclusively for either barriers or 189 enablers. If data were presented stratified into groups, overall cohort data were 190 extracted. 191

Quotes extracted from qualitative studies were analysed with an inductive approach of 193 thematic coding as described by Gibbs (Gibbs, 2007) . All quotes were retrieved from 194 included articles that met selection criteria. Quotes were coded with the same label 195 where they were an example of the same phenomenon, idea or explanation. Codes were 196 organised to provide meaning and structure to the data, leading to higher level 197 categories based on their relationships with each other. To describe each of these 198 categories, quotes were used in results. Analysis was conducted by two reviewers for 199 credibility and rigour. 200

There were 4,602 records screened, of which 60 studies met inclusion criteria ( Figure  202 1). To be included in the meta-analysis, study questions walking and cycling data being 203 aggregated, the study not reporting on barriers and/or enablers, and the study reporting 204 only on actual barriers and enablers that were calculated from self-reports and 205 infrastructure than surrounded the participant, or from self-reports and their frequency 206 of cycling. 207 208 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 Through quality assessment, 7 publications and 2 grey literature reports did not meet the 209 requirements. Reasons included lack of information about participants and recruitment, 210 and lack of clarity in the measurement and reporting of outcomes (see Supplementary  211 Materials A). 212 213 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 The final 45 final studies were mostly quantitative and cross-sectional in design 235 (n = 29), followed by qualitative (n = 8), mixed methods (n = 7) and one longitudinal 236 survey (see Table 1 ). Thirty-seven studies were identified from a bibliographic database 237 search, seven from grey literature sources and one from an integrated database (TRID). 238

Online surveys were the most common data collection method (n = 23). The majority of 239 studies reported on manual bikes (n = 43), and two on electric pedal-assisted bikes (e-240 bikes). A range of methods were used to assess barriers and enablers and quantify their 241 relative importance, including binary outcomes, Likert scales and other types of 242 proportional data. . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 Adults living in inner west Sydney between 50 and 75 who were willing to cycle two hours per week or more for a 12-week period Qualitative *quantitative outcome data extracted from cross-sectional survey only **barriers recorded for n = 295

There were 16 different outcome measures used across 36 quantitative studies of 251 barriers and enablers. Of these, most studies (n = 17) reported the proportion of people 252 who noted that the factor was a barrier or enabler to them as a binary outcome of "yes" 253 or "no". Remaining quantitative studies used Likert scales (n = 9) or other types of 254 proportional measures. Likert scales had further variety, ranging from -1 to 1, to 1 to 7 255 scales, measuring the importance of a barrier or enabler, level of agreement that it was a 256 barrier or enabler, or how much the factor affected their propensity to ride a bike. 257

Studies that used Likert scale measures reported the mean for the cohort. Other types of 258 measures used included the proportion of people who "somewhat agreed" or "strongly 259 agreed" that a particular factor was a barrier or enabler, the proportion of days not 260 commuted by bike due to a particular barrier, the proportion of people who selected a 261 particular factor as being the main barrier to them riding a bike, and the proportion that 262 reported an enabling factor as the most important improvement for them to start using a 263 bike. 264

Included studies were conducted in Europe (n = 21, 48%), North America (n = 12, 266 27%), Australia (n = 11, 23%) and New Zealand (n = 1, <1%). Quantitative study 267 sample sizes ranged from n = 89 to n = 13,700, with a median (quartile 1, quartile 3) of 268 665 (285, 1937) participants. Studies had varying proportions of women (range: 20-269 100%, median (quartile 1, quartile 3) = 50% (42%, 58%). Eleven of the 45 included 270 studies focussed on university student and/or staff populations exclusively. 271 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101/2022.04.22.22274159 doi: medRxiv preprint

There were 34 barriers to cycling for transport identified through quantitative and 273 qualitative methods (Table 2 ). The most commonly reported and measured barriers in 274 both quantitative and qualitative studies were bad weather, the time taken to a 275 destination being too great, a perceived lack of safety and high density of motor 276 vehicles. Barriers reported in exclusively in qualitative data that were not measured in 277 quantitative studies included a lack of connectivity of bike paths, negative non-rider 278 attitudes toward cyclists, and negative perceptions of e-bikes from people riding pedal 279 bikes. Barriers were categorised into eight themes; safety, trip factors, personal factors, 280 infrastructure, access, environmental factors, end of trip facilities and perceptions. 281

All barriers identified through data that could not be meta-analysed are shown in 282

Supplementary materials B. 283 284 285 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. End of trip facilities . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; Five barriers were reported across 31 studies that related to safety (see Table 3 ). These 251 included perceived risk of injury, traffic density, feeling of safety, fear of motorist 252 aggression and fear of crime. A fear of motorist aggression was reported by the largest 253 summary estimate relative to other barriers that were meta-analysed (summary estimate 254 = 70%; 95% CI: 60%; 80%). This result was supported by Likert scale data, where 255 participants rated barriers relating to safety as important and qualitative data where 256 participants commented on not feeling safe on a bike (see Supplementary Materials B). 257 "I mean the big resistance to cycling is…a lot of people just feel…don't feel safe 258 cycling" (Jones et al., 2016) 259 Feeling unsafe was another barrier relating to safety measured in 16 studies 260 (Table 3) . When asked about feelings of safety, some participants believed this was 261 regarding their safety on the road in relation to other road-users, or safety on the bike. 262

Many reported feeling unsafe riding on the road due to the presence of motor vehicle 263 traffic and the potential for injury. 264 "I don't feel safe at the moment, I know a few people, parents or whatever who 265 have been hit by or killed by cars in bike accidents… you are not protected like 266 you are in a car" (Jones et al., 2016) 267

Other participants reported feeling unsafe on a bike due to potential crime. 268 "It is dangerous to cycle in the city, except for certain areas" Male, non-cyclist 269 (Scott, 2009) 270 271 "I work late and I don't feel safe. In a car you can lock the doors, but on a bike 272 you just have to ride fast." (Scott, 2009) 273 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 Infrastructure 274 A lack of dedicated bike infrastructure had a higher summary estimate relative to other 275 barriers (summary estimate = 45%; 95% CI: 26%; 64%). This was supported by 276 participant reports across five qualitative studies (M. Daley et al., 2007; Jones et al., 277 2016; Mullan, 2013; Rérat, 2019; Salvo et al., 2018) , and quantitative studies reporting 278

proportional data not included in the meta-analyses (Jennifer Dill & McNeil, 2016; 279 Léchaud, 2016) . 280 281 Ten studies reported specifically on the poor condition of existing bike 282 infrastructure as a barrier (Table 3) , with a high summary estimate relative to all meta-283 analyses conducted (summary estimate = 61%; 95% CI: 37%; 85%). This was further 284 expanded in qualitative studies, where participants had varied concerns regarding the 285 condition of infrastructure, including; paths not being wide enough to accommodate 286 bikes (M. Daley et al., 2007) , bike lanes being adjacent to parked cars where they were 287 at risk of being hit by a door (Winters et al., 2015) , lanes being blocked by parked cars 288 (M. Daley et al., 2007) , and poor physical condition of bike infrastructure, including 289 from potholes, gravel and litter (Galway et al., 2021; Jones et al., 2016; Rérat, 2019) . 290

While not measured in quantitative studies, a lack of connectivity of existing bike 291 infrastructure was reported across four qualitative studies (M. Daley et al., 2007; Linden 292 et al., 2020; Mullan, 2013; Scott, 2009) . 293 "You are going along the road and all of a sudden there's a pedestrian crossing or 294 something like that, and the cycle lanes just ends, and then you have to work your 295 way back out into traffic." (Mullan, 2013) 296 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) greatly across meta-analysis results, Likert scale and qualitative data for other barriers 300 relating to trip factors. Of the five studies that used Likert scales to measure agreement 301 or importance of distance as a barrier, participants in three studies (Crawford et al., 302 2001; Delso et al., 2018; Fernández-Heredia et al., 2014) reported that this was 303 important or stopped them from cycling, while two studies (Manaugh et al., 2017; 304 Jennifer E. van Bekkum et al., 2011) reported that it was not important or disagreed. 305

Results varied between studies reporting having children or other people to 307 transport as a barrier (Table 3) . Of the two studies that used Likert scales, one reported 308 having people to transport as an important barrier (Crawford et al., 2001) 

Ten personal barriers were reported across 25 studies (Table 4) Owning clothes that are suitable to ride a bike in was reported by five cross-332 sectional studies, with differing findings (Table 4 ). Proportions of participants who 333 either slightly or strongly agreed that this was a barrier ranged from 19-24% (Léchaud, 334 2016; Rérat, 2019) . Between 5 and 7.6% of participants in two quantitative cross-335 sectional studies reported not owning suitable clothes as their main barrier to cycling 336 (Heart Foundation, 2013; Scott, 2009 ). However, on a Likert scale, a survey of women 337 at a university in Edinburgh reported this as not a discouraging factor (Jennifer E. van 338

Access 340 Access-related barriers were reported in 15 studies. The summary estimate for not 341 owning a bike being a barrier was lower relative to other meta-analyses (Figure 2 ). The 342 cost of a bike was reported in eight studies (Table 4) , where of three using Likert scales, 343 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101/2022.04.22.22274159 doi: medRxiv preprint two reported cost as not being a discouraging factor (Manaugh et al., 2017; Jennifer E. 344 van Bekkum et al., 2011) while one reported this as an important barrier (Crawford et 345 al., 2001) . 346 "My biggest barrier is the cost of a bike" (Swiers et al., 2017) 

Environmental barriers were reported in 30 studies. Bad weather was identified as a 349 barrier in in 22 papers (Table 4) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 Attitudes 370

All attitudinal barriers were identified through qualitative data only. Participants 371 reported the negative community perceptions of cyclists in five studies (Table 4) . 372

Reasons behind these perceptions differed, including the perception of status when 373 riding a bike, and the perception of cyclists on roads. Some participants reported that 374 when they cycled on the road, they felt that they were treated as a "second-class citizen" 375 (M. Daley et al., 2007) . Others reported that if someone were riding a bike, this may 376 lead them to thinking the person does not have a car. They reported having a car as 377 being a mark of socioeconomic status. 378 "There is that attitude that you only cycle because you can't afford a car" (Male, 379 cyclist) (Mullan, 2013) 380 Negative perceptions of e-bikes from pedal cyclists were reported in two qualitative 381 studies (Jones et al., 2016; . Participants reported riding an e-382 bike as being perceived as "cheating" by people who ride unpowered pedal bikes. 383 "I guess the initial reaction is that it's cheating, partly because I'm part of the 384 cycling group/culture and [they] think it's cheating" (Jones et al., 2016) 385

There were 22 enablers to cycling for transport identified (Table 5 ). Enablers were 387 categorised as either relating to infrastructure, resources and end of trip facilities, 388 personal improvement and enjoyment, motivations and practicality. Riding a bike being 389 a fun enjoyable activity was reported by the highest number of studies (n = 17), 390 followed by riding a bike being an environmental choice (n = 15) and an efficient mode 391 of transport (n = 17). All recorded qualitative and quantitative data that were not able to 392 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 be included in meta-analyses are shown in Supplementary Materials or available on 393 request. 394 395 396 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 Eight meta-analyses were possible where two or more studies reported 401 unweighted proportional data on the same outcome (Figure 3) . 402

. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 Infrastructure 407

The provision of protected bike lanes (bike lanes physically separated from traffic) had 408 the highest summary estimate relative to other meta-analyses of enablers (summary 409 estimate = 84%; 95% CI: 80%; 87%). This was further rated as important in two Likert 410 scale studies (Crawford et al., 2001; Winters et al., 2011) . Adequate lighting, signage 411 specific to people riding bikes, or having a mostly flat route were also rated as 412 important enablers (Crawford et al., 2001; Heart Foundation, 2013; Stangeby, 1997; 413 The Heart Foundation, 2011; Winters et al., 2011) . 414

Having access to changing rooms and showers was measured as an enabler in three 416 quantitative studies that were not included in meta-analyses (Crawford et al., 2001; 417 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 /2022 .04.22.22274159 doi: medRxiv preprint Heart Foundation, 2013 Whannell et al., 2012) , where in one survey of women, 4.9% 418 reported this as the main reason that would enable them to ride a bike (Heart 419 Foundation, 2013) . 420

Factors related to personal improvement and enjoyment from bike riding were widely 422 reported in qualitative data. A high summary estimate was reported for riding a bike 423 being a fun and enjoyable activity (summary estimate = 55%; 95% CI: 25%; 86%), 424 further supported by participants across eight studies ( Table 6 ) that commented on the 425 enjoyment of riding a bike. 426 "I always feel happy [when e-biking]. I liked it so much better than sitting in a car." 427 (Jones et al., 2016) 428 Some studies reported bike riding as a stress release, or to help a person's mental 429 health, as an enabling factor. In a survey conducted in the centre of Madrid (Muñoz et 430 al., 2013) participants reported a mean of 7.04 on a scale of 1 (strongly disagree) to 10 431 (strongly agree), for stress release being an enabling factor for bike riding. Participants 432 recalled bike riding as "my little bit of time alone" (M. Daley et al., 2007) . 433

Bike riding as a more sustainable and environmentally friendly option to travel was 435 reported as less of an enabling factor compared to summary estimates of other enablers 436 (summary estimate = 58%; 95% CI: 31%; 85%). However, this was rated as an 437 important enabling factor across all studies that measured this with a Likert scale 438 (Crawford et al., 2001; De Geus et al., 2008; Fernández-Heredia et al., 2014; Heinen et 439 al., 2011; Muñoz et al., 2013) , and one study of women who ride a bike as their primary 440 mode of transport (Cred Consulting, 2020). Participants in four qualitative studies 441 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101/2022.04.22.22274159 doi: medRxiv preprint (Galway et al., 2021; Samantha J Leger et al., 2019; D. Lois et al., 2016; J. E. van 442 Bekkum et al., 2011) reported riding a bike because it is "better for the planet" (Galway 443 et al., 2021) . 444 445 An enabling factor reported in four qualitative studies (Rérat, 2019; Salvo et al., 446 2018; Scott, 2009 Riding a bike being a more economically viable transport option had a high summary 458 estimate relative to other meta-analyses of enablers (summary estimate = 60%; 95% CI: 459 45%; 74%). Participants in four Likert scale studies rated this as an important enabling 460 factor (Crawford et al., 2001; Fernández-Heredia et al., 2014; Heinen et al., 2011; 461 Muñoz et al., 2013) , while participants in three qualitative studies (D. Lois et al., 2016; 462 Scott, 2009; reported that bike riding was a one-time cost 463 rather than ongoing costs, such as with a motor vehicle or public transport. 464 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. vehicles, such as protected bike lanes, were also commonly reported. There is potential 487 for these barriers, which are largely safety and infrastructure-related, to be overcome by 488 interventions that minimise interactions between bike riders and motor vehicle traffic. 489 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 The provision of high-quality protected infrastructure remains the gold-standard for 490 increasing participation in bike riding (Hull & O'Holleran, 2014; Stewart et al., 2015) . 491

However, there is difficulty in providing protected infrastructure across entire cities and 492 regions. Further interventions to minimise potentially unsafe interactions between bike 493 riders and motor vehicles could enable connections between protected infrastructure. 494

This may include areas of lower motor vehicle speeds (such as 30km/h urban speed 495 limits) or traffic calming infrastructure (Yang et al., 2010) . Despite the need for 496 protected infrastructure, on-road bike lanes remain a prominent form of bike 497 infrastructure in urban design (Beck, 2021) . As supported by our findings where "any 498 bike lane" had a lower summary estimate than specifically protected bike lanes, this 499 type of infrastructure does little to enhance participation in bike riding (Akar & Clifton, 500 2009; Michelle Daley & Rissel, 2011; J. Dill, 2009; Kristiann C Heesch et al., 2012; 501 Twaddle et al., 2010) , and places a person at a substantially higher risk of injury 502 compared to riding a bike in a protected bike lane (Haileyesus et al., 2007; Lusk et al., 503 2011; Mehan et al., 2009) . 504 505 A lack of storage on a bike was widely reported as a barrier to riding a bike for 506 transport. This included storage to carry children and objects required for work. This 507 suggests there is a greater need for availability of bikes with enhanced storage capacity, 508 and for environments supportive of such bikes. One way to account for this need could 509 be an increase in the accessibility of 'cargo bikes'. Cargo bikes have substantial 510 carrying capacity permitting the movement of children and belongings, however there 511 are issues in uptake due to current infrastructure often being too narrow to comfortably 512 ride and manoeuvre in (Liu et al., 2020) . Many urban environments are designed with 513 the premise of being used solely by upright two-wheeled bikes (Napper, 2020) . Newly 514 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 established infrastructure dedicated to bikes may benefit from input from cargo-bike 515 riders to ensure lanes are designed to enhance the capacity for bikes with storage 516 capabilities to ride on them. 517 518 When designing inclusive infrastructure, it is also important to incorporate the 519 needs of people riding bikes with children (Aldred et al., 2017; Clayton & Musselwhite, 520 2013) . It was evident through qualitative data that some adults would ride a bike for 521 transport if they were able to with their children. However, they reported not feeling 522 safe to do so with existing infrastructure. Other research indicates that parents often 523 avoid riding a bike for travel as sections of their bike network are perceived as unsafe 524

for children, such as sections with on-road lanes (McLaren, 2016) . While the barriers to 525 children riding a bike to commute to school have been thoroughly researched, there is 526 limited data on the barriers to adults riding alongside children for other trips. As many 527 trips with children are often within local areas (Yeung et al., 2008) , there is potential for 528 modal shifts to bike riding if the needs of this group are understood and met. 529 530 A range of measures were used in studies included in this review to identify 531 barriers or enablers, or quantify the importance of these barriers and enablers. Because 532 of this, many study findings were unable to be robustly compared or synthesised, 533 limiting discussion of differences in barriers and enablers between countries, samples 534 and study types. Measures included in meta-analyses used binary outcomes which were 535 useful for showing the size of the effect of a particular barrier or enabler. However, this 536 does not consider the relative strength that a barrier or enabler may have in influencing 537 an individual's bike riding behaviours. Other studies used Likert scales, permitting the 538 participant to choose if a factor was very much, or less so of a barrier or enabler to their 539 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 riding a bike. This approach highlights potentially influential factors, allowing for a 540 more focused approach. A downside of using Likert scales in this context is that studies 541 exploring barriers and enablers to bike riding often have a large list of factors for 542 inclusion. Where each factor requires a Likert scale answer, this can contribute to 543 respondent fatigue, threatening data validity (Cox III, 1980; Dolnicar et al., 2011) . To 544 allow for comparisons between, and synthesis of, study findings in future, an approach 545 that utilises the benefits of both binary and Likert scale measures is necessary. A two-546 step logic where participants first select all factors that are a barrier or enabler to them 547 riding a bike, followed by a Likert scale question displaying only those factors chosen 548 by the participant could be beneficial. This approach would identify both the proportion 549 of people affected by particular barriers and enablers, and how much they do so. 550 551 It was evident through qualitative findings that there were varying 552 interpretations of questions regarding feelings of safety while riding a bike. Some 553 people perceived questions around feelings of safety to mean feeling unsafe riding a 554 bike, while others perceived this as not feeling a sense of personal safety in, for 555 example, a particular area. The separation of safety due to traffic and road safety 556 compared to safety due to crime, theft or other environmental factors in future research 557 is imperative for the validity and reliability of data (Forman et al., 2008) . Pikora et al. 558 (2002) previously outlined potential questions and keywords to specify and reliably 559 measure safety in bike riding research, however this could benefit from updated review 560 (Pikora et al., 2002) . 561 562 Some enablers reported in this review were identified exclusively through 563 qualitative data. These included riding a bike for transport as a role model for children, 564 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 in a hope that they will see it as a viable choice of transport. This enabling factor is also 565 present in qualitative research of the enablers of physical activity (Caperchione et al., 566 2012; Stronach et al., 2016) . As it is well established that parents participating in 567 physical activity will have a positive influence on children's perception and 568 participation in physical activity (Cheung & Chow, 2010; Welk et al., 2003) , and there 569 is evidence that this may be an enabling factor for adults, further quantitative 570 measurement and exploration may be beneficial. An understanding of enabling factors 571 such as this have the potential to inform meaningful promotional campaigns and values-572 based messaging around bike riding. 573 574 Reviews in bike riding and active transport research are often limited to 575 published research, excluding potentially relevant grey literature conducted as part of 576 government or bicycle organisation research. This review included a systematic grey 577 literature search to incorporate this, while maintaining rigour by using study-specific 578 quality assessment tools. While consideration was taken to carefully interpret data from 579 grey literature sources, as they were not peer-reviewed, these sources may be subject to 580 selection and publication bias. 581 582 A further limitation extends from the synthesis of data across a diversity of 583 populations and locations. Locations had varying geographical areas, trip distances and 584 infrastructure. Populations varied, including groups of university students, and studies 585 of solely women. These variations and the substantial heterogeneity between samples 586 prevented robust comparisons being made between study designs, countries and 587 populations. Despite these variations and their likely influence on results, certain 588 summary estimates had smaller confidence intervals relative to others. This 589 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 demonstrates the consistency of reporting these barriers and enablers across diverse 590 geographical areas and population groups. In addition, not all barriers and enablers were 591 able to be meta-analysed due to the heterogeneity of outcome measures used. 592 593 This review only included studies conducted in OECD countries and published 594 in the English language. Findings are likely not reflective of the barriers and enablers 595 present for people in lower to middle income economies (Bank, 2019) , who encounter 596 vastly different road and infrastructure conditions (Forjuoh, 2003) . Further review is 597 required to understand the unique barriers and enablers faced by these areas. 598

There are other methods used to capture barriers and enablers to cycling than 600 what has been included in this review. Stated preference surveys have been used and 601 reviewed previously, however the majority report on focused categories of barriers 602 and/or enablers, such as infrastructure types (Aldred et al., 2017) or bike parking 603 (Heinen & Buehler, 2019 ). This review focused on self-reported barriers and enablers 604 from survey and qualitative data, which enabled the capture of the diversity and breadth 605 of barriers and enablers people experience for riding a bike, as well as comparisons in 606 frequency and distribution between categories. 607

Understanding the barriers and enablers to cycling for transport is integral to designing 609 environments supportive of current and potential bike riders, for promotional campaigns 610 to encourage participation in bike riding and to influence the development of supportive 611 policy. This systematic review identified a diverse range of barriers and enablers using 612 quantitative and qualitative data, highlighting the importance of safety and 613 infrastructure, and potential flaws in survey reliability. Further research should include 614 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 measures to show the strength of particular barriers and enablers. It is recommended 615 that future research of barriers to bike riding clearly specifies the meaning of "safety". 616 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 22, 2022. ; https://doi.org/10.1101 https://doi.org/10. /2022 

Gender gap generators for bicycle mode choice in Baltimore college campuses

Economic impact of cycling across various nations

Encouraging Bicycling among university students: Lessons from Queen's university

Influence of individual perceptions and bicycle infrastructure on decision to bike

Bicycling Choice and Gender Case Study: The Ohio State University

Cycling provision separated from motor traffic: a systematic review exploring whether stated preferences vary by gender and age

Spatial variation in bicycling: a retrospective review of travel survey data from Greater

Associations between individual characteristics, availability of bicycle infrastructure, and city-wide safety perceptions of bicycling: A cross-sectional survey of bicyclists in 6 Canadian and US cities

Trends in walking and cycling safety: recent evidence from high-income countries, with a focus on the United States and Germany

The similarities and differences analysis of OECD countries in terms of health system indicators

What a man wants: understanding the challenges and motivations to physical activity participation and healthy eating in middleaged Australian men

Association between active commuting and incident cardiovascular disease, cancer, and mortality: prospective cohort study

Central Goldfields Shire Council

Parental mediatory role in children's physical activity participation

An examination of cyclists' and non-cyclists' mode choice under a new cycle network

On the go: how women travel around our city: a case study on active transport across Sydney through a gender lens

Exploring changes to cycle infrastructure to improve the experience of cycling for families

The optimal number of response alternatives for a scale: A review

Employee attitudes towards active commuting

Private and public modes of bicycle commuting: A perspective on attitude and perception

Perspectives and images of cycling as a barrier or facilitator of cycling

All dressed up and nowhere to go? A qualitative research study of the barriers and enablers to cycling in inner Sydney

Psychosocial and environmental factors associated with cycling for transport among a working population

Potentially replaceable car trips: Assessment of potential modal change towards active transport modes in Vitoria-Gasteiz

National Travel Survey

Bicycling for transportation and health: the role of infrastructure

Revisiting the four types of cyclists: findings from a national survey

Quick, simple and reliable: Forced binary survey questions

Understanding cyclists' perceptions, keys for a successful bicycle promotion

E-bikes in the mainstream: reviewing a decade of research

Barriers and facilitators to public bicycle scheme use: A qualitative approach

Traffic-related injury prevention interventions for low-income countries

Reliability and validity of destination-specific barriers to walking and cycling for youth

A push to cyclingexploring the e-bike's role in overcoming barriers to bicycle use with a survey and an intervention study

Exploring socialecological influences on commuter cycling in a midsize northern city: A qualitative study in Thunder Bay

Thematic coding and categorizing. Analyzing qualitative data

Cyclists injured while sharing the road with motor vehicles

Environmental Impacts of Short Car Trip Replacement with Micromobility Modes

Gender differences in recreational and transport cycling: a cross-sectional mixed-methods comparison of cycling patterns, motivators, and constraints [Comparative Study Research Support

Gender differences in recreational and transport cycling: a cross-sectional mixed-methods comparison of cycling patterns, motivators, and constraints

Bicycle parking: a systematic review of scientific literature on parking behaviour, parking preferences, and their influence on cycling and travel behaviour

The role of attitudes toward characteristics of bicycle commuting on the choice to cycle to work over various distances

Bicycle infrastructure: can good design encourage cycling? Urban, Planning and Transport Research

Cycling in Warsaw, Poland -Perceived enablers and barriers according to cyclists and non-cyclists

Motives, perceptions and experiences of electric bicycle owners and implications for health, wellbeing and mobility

Still feeling the car-The role of comfort in sustaining private car use

The role of innovation on Bicycle Commuting in Portugal

If I had a regular bicycle, I wouldn't be out riding anymore": Perspectives on the potential of e-bikes to support active living and independent mobility among older adults in Waterloo

If I had a regular bicycle, I wouldn't be out riding anymore": Perspectives on the potential of e-bikes to support active living and independent mobility among older adults in Waterloo

The effect of cycling on cognitive function and well-being in older adults

Maybe next time: Perspectives on commuting from bike to work week non-participants

Evaluating active transport benefits and costs. Victoria Transport Policy Institute

What makes a good cargo bike route? Perspectives from users and planners

Qualitative analysis on cycle commuting in two cities with different cycling environments and policies

Qualitative analysis on cycle commuting in two cities with different cycling environments and policies

Risk of injury for bicycling on cycle tracks versus in the street

Overcoming barriers to cycling: understanding frequency of cycling in a University setting and the factors preventing commuters from cycling on a regular basis

Families and transportation: Moving towards multimodality and altermobility

Bicycle-related injuries among children and adolescents in the United States

Exercise, weather, safety, and public attitudes: A qualitative exploration of leisure cyclists' views on cycling for transport

The development of a critical appraisal tool for use in systematic reviews addressing questions of prevalence

Cycling habits and other psychological variables affecting commuting by bicycle in Madrid

What is a parked bicycle? Vehicle fleet characteristics in Australia

Understanding bike-sharing acceptability and expected usage patterns in the context of a small city novel to the concept: A story of 'Greek Drama' [Empirical Study

Transportation research part F: traffic psychology and behaviour

Cycling is associated with a lower incidence of cardiovascular diseases and death: Part 1-systematic review of cohort studies with meta-analysis

The health and economic benefits of active transport policies in Barcelona

Developing a reliable audit instrument to measure the physical environment for physical activity

Benefits of shift from car to active transport

Cycling to work: Meanings and experiences of a sustainable practice

Political activity for physical activity: health advocacy for active transport

RStudio: Integrated Development for R

Characteristics of utility cyclists in Queensland, Australia: an examination of the associations between individual, social, and environmental factors and utility cycling

A mixed methods study on the barriers and facilitators of physical activity associated with residential relocation

Improving university transportation sustainability: Reducing barriers to campus bus and bicycle commuting

Package 'meta'. The R foundation for statistical computing

To bike or not to bike: Seasonal factors for bicycle commuting

Attitudes towards walking and cycling instead of using a car

Cycling: A benefit for health or just a means of transport? Case study Brno (Czech Republic) and its surroundings

What interventions increase commuter cycling? A systematic review

Sistas' and Aunties: sport, physical activity, and Indigenous Australian women

A cross sectional survey of attitudes, behaviours, barriers and motivators to cycling in university students

Riding a bike for transport

Latent classes of daily mobility patterns: the relationship with attitudes towards modes

Latent bicycle commuting demand and effects of gender on commuter cycling and accident rates

Short car trips in Norway: is there a potential for modal shift?

Cycle commuting and perceptions of barriers: Stages of change, gender and occupation

Employees' perceptions of cycle commuting: A qualitative study

Parental influences on physical activity in children: An exploration of potential mechanisms

Tertiary student attitudes to bicycle commuting in a regional Australian university

Implementing cycling infrastructure in a politicized space: Lessons from Toronto

Motivators and deterrents of bicycling: Comparing influences on decisions to ride

I grew up on a bike": Cycling and older adults

Interventions to promote cycling: systematic review. bmj

Child transport practices and perceived barriers in active commuting to school

Joy, exercise, enjoyment, getting out: a qualitative study of older people's experience of cycling in