1. Introduction
Individuals who are physically active are healthier across a broad spectrum, as compared to those who are not physically active.1 Despite the known benefits of physical activity, 48% of United States adults do not meet Physical Activity Guidelines for Americans for aerobic activity.2,3 Evidence suggests that bicycling is a feasible way for individuals to meet the physical activity guidelines.4-6 Indeed, compared to both active and inactive non-bicyclists, regular bicyclists are more active, more fit, less likely to develop cardiovascular diseases and cancers, and live longer lives.7,8
Bicycling behavior can be divided into two domains – recreation, i.e., riding solely for the purpose of pleasure or exercise, and transportation, i.e., utilitarian riding to get from place to place. When considering bicycling for transportation, estimates from the 2009 United States (US) National Household Travel Survey indicate that individuals who bicycle for transportation engage in approximately 28 minutes more physical activity per day, as compared to automobile commuters.9 Bicycling for transportation has itself been associated with significant health benefits, despite a potential increased risk of injury or air pollution exposure when bicycling on roadways.10 Economic and environmental benefits have been associated with transportation bicycling as well, through a reduction in health care costs, fuel expenditure, motor vehicle traffic, and air pollution.7,11
Bicycling to work (i.e. bicycle commuting) is a type of active transportation that provides regular physical activity to individuals, in that it can be built into a regular weekday schedule as a means of getting to and from their place of employment. Bicycle commuting in particular has been associated with health benefits such as higher mental wellbeing, as compared to individuals who do not bicycle commute.12 Additionally, bicycle commuting has been found to be associated with reduced sickness absence at work, as compared to individuals who do not bicycle commute.12,13 This reduction in workplace absence would produce substantial economic benefit, by limiting lost wages and loss of productivity in the workplace.
Although the benefits of bicycling for transportation, and for bicycle commuting in particular, are clear, only 0.5% of the US population was estimated to bicycle commute in 2017.14 A number of studies have identified factors across multiple levels of the ecological model that influence bicycling for transportation.15-19 City infrastructure, such as bicycle lanes, is associated with bicycling behaviors,17 but when considering bicycle commuting, the workplace environment and policy or regulatory supports may be influential as well. Limited research has been conducted to examine what factors at the institutional and policy level, specifically within the workplace, influence bicycle commuting. Most studies that have examined the influence of the workplace on active travel have not examined walking and bicycling independently.20-23 The few studies that have examined workplace correlates of bicycling to work specifically have found mixed results, with some finding an association between workplace supports such as bicycle racks, showers, and policies and bicycle commuting,24,25 whereas others have found no association.26-28
More research is needed to understand what factors within a workplace may influence an individual’s travel to work. The objective of this study was to examine the association between workplace facilities and policies with bicycle commuting, and to test interactions by gender, among a convenience sample of adult bicyclists from Austin, TX, and Birmingham, AL.
2. Methods
2.1. Study Sample
An analysis of cross-sectional data from a sample of adults living in Travis County, Texas and Jefferson County, Alabama was conducted in 2018. Data collection procedures for this study have been previously described.29 Briefly, an observational study was conducted to assess multiple level ecological factors (intrapersonal, interpersonal, perceived social environment, perceived built environment, policy, and actual built environment) and their association with frequency of recreation and transportation bicycling behavior. Travis County, Texas and Jefferson County, Alabama were selected as study sites to include a wide range of environmental and social attributes that may be associated with bicycling rates. Austin, located in Travis County, Texas has been defined as one of the best cities for bicycling in the US, while Birmingham, located in Jefferson County, Alabama, is considered to be much less bicycle friendly.30 Recruitment was conducted via the Internet, posted advertisements, and word of mouth, from October 2016 to January 2017. Persons were eligible to participate if they lived in the study area, were aged 18-65 years, and reported having ridden a bicycle at least one time in the past year. Individuals who participated were put into a drawing to win a gift card.
Participants responded to an Internet-based questionnaire designed specifically for this study to assess proposed ecological factors of bicycling behavior. Study data were collected and managed using REDCap electronic data capture tools.31 The The University of Texas Health Science Center at Houston Institutional Review Board deemed this pilot study exempt from review.
2.2. Measures
2.2.1. Bicycle commuting
Bicycle commuting, a dichotomous (yes/no) variable, was the main outcome of interest for the current analysis. Participants were asked if they have ridden a bicycle specifically for transportation in the past year. Those who answered “yes” were then asked what the purposes of their trips were when they rode a bicycle for transportation.32 Options included commuting to/from work, commuting to/from school, personal errands, required for my job, drop off/pick up someone, visit a friend or relative, and other. Those who responded that a purpose of their trips was commuting to/from work were categorized as a bicycle commuter. All other participants were categorized as a non-bicycle commuter. This analysis was limited to participants who reported being employed full- or part-time.
2.2.2. Workplace facilities and policies
Six workplace facilities and policies were examined as independent variables, along with one summary variable of total reported workplace facilities and policies for bicycle commuting. As part of the questionnaire, participants who self-reported that they were employed full- or part-time were asked if their workplace had policies that supported bicycling to work. Next, employed participants were asked if any of the following facilities were provided at their workplace for bicycle commuters: bicycle lockers, locked rooms/cage, clothes storage, bike racks, and showers.33 For all questions, response options included yes, no, and I don’t know. Finally, a summary score ordinal variable was derived by adding the total number of reported policies/facilities per participant. Values of this summary variable ranged from 0 to 6.
2.2.3. Sociodemographic Characteristics
The following sociodemographic characteristics were included in this analysis as covariates: age in years, gender (male, female, neither/both/other), race/ethnicity (White, Black, Hispanic/Latino, other), education (less than high school/high school diploma or equivalent, some college/associates degree, 4-year college degree, graduate degree), household income (less than $30,000, $30,000 to less than $75,000, $75,000 and above), number of licensed motor vehicles available for use (0-4 or more), married or living with partner (yes, no), and having children under age 18 living at the participant’s home at least 3 days per week (yes, no).
2.3. Statistical Analysis
For the purposes of this analysis, the workplace facilities and policies summary score values 4, 5, and 6 were combined into a “4 or more” category due to small sample size. Mean and standard deviation were utilized to describe the study population for continuous variables, while count and percentage were utilized for categorical variables. Logistic regression models were used to assess the association between each of the workplace facility/policy variables (three categories: yes, no [referent], I don’t now), as well as the summary score variable (0 [referent], 1, 2, 3, 4+), and being a bicycle commuter (yes vs. no [referent]). Six independent bivariate logistic regression models were first conducted to assess the unadjusted association between each of the workplace variables and being a bicycle commuter. Second, six independent multivariable logistic regression models were conducted to identify the association between each of the workplace facility/policy variables and being a bicycle commuter, while adjusting for sociodemographic characteristics. Finally, a multivariable logistic regression model was conducted to identify the association between the summary score variable for “total workplace facilitators” and being a bicycle commuter, while adjusting for sociodemographic characteristics. A post-hoc test for linear trend was then conducted.
Interactions were tested between gender and each of the workplace facilitator/policy and summary score variables. Significant interactions were defined as a p-value <0.15 with the Wald test.34 If a significant interaction was identified, the estimates in the adjusted models were reported as a linear combination by each level of the gender. Clustering by study site was accounted for in all models. Collinearity of all variables was assessed via the variance inflation factor (VIF). A VIF for any variable greater than 10 indicated collinearity.35 Sensitivity analyses were conducted. First, associations were explored stratified rather than clustered by study site. Second, “frequent” bicycle commuters were examined, defined as individuals who reported that a purpose of their trips was commuting to/from work as well as reported bicycling for transportation 4 or more times a month. Each multivariable logistic regression model was run with the outcome of frequent bicycle commuters (yes, no [referent], I don’t know). All analyses were conducted using STATA 15.1 (STATA Corporation, College Station, TX, USA).
3. Results
A total of 998 participants completed the questionnaire. After excluding those participants who were not employed full- or part-time (n=145, 14.5%), those with missing data (n=75, 8.8%), unrealistic ages (n=11, 1.3%), and those who reported gender as neither/both/other (exclusion due to small sample size; n=7, 0.9%) the final analytic sample was of 760 participants. Descriptive characteristics of the sample are reported in Table 1. In this sample of bicyclists, 45% of participants reported bicycling to work in the past year. As compared to participants who did not bicycle to work, participants who bicycled to work were slightly younger (mean age 37 vs. 39), more likely to be male, less likely to be Black, more likely to be Hispanic/Latino, lower income, less likely to have multiple motor vehicles, and less likely to have children living at home. In this sample, 71% of bicycle commuters were from Travis County, whereas 66% of non-bicycle commuters were from Jefferson County. Access to workplace facilities/policies varied by type, from 7% for bike lockers to 50% for bike racks. Thirty-one percent of participants reported no workplace supports, whereas 16% of participants reported 4 or more supports; these distributions differed based on bicycle commuter status.
Table 1.
Descriptive characteristics of adult bicyclists residing in Travis County, Texas and Jefferson County, Alabama, October 2016-January 2017
| Variable | Full Sample n=760 |
Bicycle commuter n=342 |
Non-bicycle commuter n=418 |
|---|---|---|---|
| Age [mean (SD)] | 38.1 (10.6) | 36.6 (10.1) | 39.3 (10.9) |
| Gender [n(%)] | |||
| Male | 430 (56.6) | 223 (65.2) | 207 (49.5) |
| Female | 330 (43.4) | 119 (34.8) | 211 (50.5) |
| Race [n(%)] | |||
| White | 649 (85.4) | 295 (86.3) | 354 (84.7) |
| Black | 32 (4.2) | 4 (1.2) | 28 (6.7) |
| Hispanic | 37 (4.9) | 24 (7.0) | 13 (3.1) |
| Other | 42 (5.5) | 19 (5.6) | 23 (5.5) |
| Education [n(%)] | |||
| Less than high school/high school graduate or equivalent | 26 (3.4) | 8 (2.3) | 18 (4.3) |
| Some college/associates degree | 116 (15.3) | 55 (16.1) | 61 (14.6) |
| 4-year college degree | 350 (46.1) | 159 (46.5) | 191 (45.7) |
| Graduate degree | 268 (35.3) | 120 (35.1) | 148 (35.4) |
| Household Income [n(%)] | |||
| Less than $30,000 | 100 (13.2) | 64 (18.7) | 36 (8.6) |
| $30,000 to less than $75,000 | 261 (34.3) | 134 (39.2) | 127 (30.4) |
| $75,000 or more | 399 (52.5) | 144 (42.1) | 255 (61.0) |
| Number of licensed motor vehicles available for use [n(%)] | |||
| 0 | 33 (4.3) | 28 (8.2) | 5 (1.2) |
| 1 | 243 (32.0) | 137 (40.1) | 106 (25.4) |
| 2 | 339 (44.6) | 128 (37.4) | 211 (50.5) |
| 3 | 99 (13.0) | 33 (9.7) | 66 (15.8) |
| 4 or more | 46 (6.1) | 16 (4.7) | 30 (7.2) |
| Married or living with partner [n(%)] | |||
| No | 265 (34.9) | 133 (38.9) | 132 (31.6) |
| Yes | 495 (65.1) | 209 (61.1) | 286 (68.4) |
| Children under the age of 18 living at home at least 3 days per week [n(%)] | |||
| No | 547 (72.0) | 263 (76.9) | 284 (67.9) |
| Yes | 213 (28.0) | 79 (23.1) | 134 (32.0) |
| Site [n(%)] | |||
| Jefferson | 372 (49.0) | 98 (28.7) | 274 (65.6) |
| Travis | 388 (51.0) | 244 (71.3) | 144 (34.4) |
| Workplace facilitators for bicycling to work | |||
| Bike Lockers [n(%)] | |||
| Yes | 56 (7.4) | 44 (12.9) | 12 (2.9) |
| No | 668 (87.9) | 291 (85.1) | 377 (90.2) |
| I don’t know | 36 (4.7) | 7 (2.1) | 29 (6.9) |
| Locked Rooms/Cage [n(%)] | |||
| Yes | 103 (13.6) | 61 (17.8) | 42 (10.1) |
| No | 616 (81.0) | 266 (77.8) | 350 (83.7) |
| I don’t know | 41 (5.4) | 15 (4.4) | 26 (6.2) |
| Clothes Storage [n(%)] | |||
| Yes | 208 (27.4) | 115 (33.6) | 93 (22.3) |
| No | 514 (67.6) | 214 (62.6) | 300 (71.8) |
| I don’t know | 38 (5.0) | 13 (3.8) | 25 (6.0) |
| Bike Racks [n(%)] | |||
| Yes | 381 (50.1) | 228 (66.7) | 153 (36.6) |
| No | 353 (46.5) | 110 (32.2) | 243 (58.1) |
| I don’t know | 26 (3.4) | 4 (1.2) | 22 (5.3) |
| Showers [n(%)] | |||
| Yes | 270 (35.5) | 143 (41.8) | 127 (30.4) |
| No | 463 (60.9) | 189 (55.3) | 274 (65.6) |
| I don’t know | 27 (3.6) | 10 (2.9) | 17 (4.1) |
| Bike Policy [n(%)] | |||
| Yes | 279 (36.7) | 178 (52.1) | 101 (24.2) |
| No | 360 (47.4) | 123 (36.0) | 237 (56.7) |
| I don’t know | 123 (15.9) | 41 (12.0) | 80 (19.1) |
| Workplace Facilitators [n(%)] | |||
| 0 | 233 (30.7) | 57 (16.7) | 176 (42.1) |
| 1 | 164 (21.6) | 72 (21.1) | 92 (22.0) |
| 2 | 138 (18.2) | 70 (20.5) | 68 (16.3) |
| 3 | 102 (13.4) | 60 (17.5) | 42 (10.1) |
| 4 or more | 123 (16.2) | 83 (24.3) | 40 (9.6) |
All six workplace facilitator variables were significantly directly associated with being a bicycle commuter in the bivariate logistic regression models, as compared to non-bicycle commuters (Table 2). In the fully adjusted multivariable logistic regression models (adjusting for age, gender, race, education, income, number of motor vehicles, relationship status, and children in household), having bike lockers was associated with a significant 4.93 times higher odds of being a bicycle commuter (95% CI 4.40, 5.52), having locked rooms or cages was associated with a significant 2.04 times higher odds of being a bicycle commuter (95% CI 1.35, 3.09), having clothes storage was associated with a significant 1.94 times higher odds of being a bicycle commuter (95% CI 1.40, 2.72), having bike racks was associated with a significant 3.05 times higher odds of being a bicycle commuter (95% CI 2.58, 3.61), having showers was associated with a significant 2.04 times higher odds of being a bicycle commuter (95% CI 1.90, 2.18), and having workplace policies for bicycling to work was associated with a significant 3.13 times higher odds of being a bicycle commuter (95% CI 2.40, 4.09). As compared to bicyclists who reported “no”, individuals who reported “I don’t know” to having bike lockers, clothes storage, bike racks, and showers had significantly lower odds of being a bicycle commuter (results not reported).
Table 2.
Bivariate and multivariable logistic regression models of the association between workplace facilities/policies and bicycle commuting, adult bicyclists (n=760) residing in Travis County, Texas and Jefferson County, Alabama, October 2016-January 2017
| Workplace Variable | Bivariate OR (95% CI) |
Multivariablea OR (95%CI) |
|---|---|---|
| Bike Lockers | 4.75 (3.26, 6.93)*** | 4.93 (4.40, 5.52)*** |
| Locked Rooms/Cage | 1.91 (1.55, 2.35)*** | 2.04 (1.35, 3.09)*** |
| Clothes Storage | 1.77 (1.35, 2.33)** | 1.94 (1.40, 2.72)*** |
| Bike Racks | 3.29 (2.63, 4.12)*** | 3.05 (2.58, 3.61)*** |
| Bike Policy | 3.40 (2.94, 3.92)*** | 3.13 (2.40, 4.09)*** |
| Showers | 1.63 (1.52, 1.75)*** | 2.04 (1.90, 2.18)*** |
p<0.05,
p<0.01,
p<0.001
Adjusted for age, race, education, income, number of motor vehicles, relationship status, children in household, and gender.
The association between number of workplace facilitators and bicycle commuting is depicted in Figure 1. As compared to no facilitators, having 1 facilitator was associated with 2.06 times higher odds of being a bicycle commuter (95% CI 1.63, 2.59), having 2 facilitators was associated with 2.79 times higher odds of being a bicycle commuter (95% CI 2.00, 3.88), having 3 facilitators was associated with 4.16 times higher odds of being a bicycle commuter (95% CI 3.80, 4.56), and having 4 or more facilitators was associated with 7.55 times higher odds of being a bicycle commuter (95% CI 7.36, 7.75). Testing for linear trend indicated a significant direct linear association between number of facilitators and odds of being a bicycle commuter (p<0.001). Examination of the VIF indicated that collinearity could be discarded as a concern for all models. No significant interactions between each of the workplace facilitators and gender were observed. Sensitivity analyses examining the association between facilities and policies stratified by study site resulted in similar findings, although the strength of some associations were attenuated and no longer significant, likely due to smaller sample sizes (Supplemental Table 1). The dose response association between number of facilitators and bicycling to work was evident in both study sites. Similarly, sensitivity analyses examining the outcome of frequent bicycle commuting as compared to all other bicyclists in the sample resulted in similar findings for all models with the strengths of the associations slightly attenuated, but continued evidence of the dose response association between facilitators and frequent bicycle commuting (Supplemental Table 2).
Figure 1.
Multivariable logistic regression models of the association between number of workplace facilities/policies and bicycle commuting, adult bicyclists (n=760) residing in Travis County, Texas and Jefferson County, Alabama, October 2016-January 2017
Note: Adjusted for age, race, education, income, number of motor vehicles, relationship status, children in household, and gender. p-trend <0.001
4. Discussion
This study assessed the influence of workplace bicycle facilities and policies on bicycling to work in a convenience sample of adult bicyclists residing in Travis County, Texas, and Jefferson County, Alabama. Results indicate that all facilitators explored were significantly associated with bicycling to work, including bicycle lockers, locked rooms/cages, clothes storage, bicycle racks, showers, and policies. Further, this study identified a dose response relationship between the number of workplace facilitators and odds of bicycling to work, such that as the number of facilitators increased, the odds of bicycling to work increased. These findings were not modified by gender. When frequent bicycle commuting was examined, results were comparable, with all workplace facilitators significantly associated with bicycle commuting among those individuals who rode a bicycle for transportation 4 or more times a month. The dose response relationship was also evident when examined by study site.
Individuals who reported they did not know if a particular facility or policies were available at their workplace had lower odds of bicycling to work, as compared to individuals who reported “no” to the presence that facility or policy. These individuals may or may not work in places that do have supportive bicycling facilities or policies, but perhaps due to lack of interest they are unaware. These individuals can be conceptualized using the Transtheoretical Model of Health Behavior Change, a behavioral theory that posits that individuals move through six stages during behavior change - precontemplation, contemplation, preparation, action, maintenance, and termination.38 In the precontemplation stage, individuals have not yet conceived of or have no interest in the behavior.38 It can be difficult to influence an individual in precontemplation to change their behavior if they have no interest or if there are seemingly insurmountable barriers. In the context of bicycling to work, for example, the presence of absence of workplace facilitators may be irrelevant to individuals that don’t like to bicycle, or have physical or family barriers that they are unable to overcome. However, for those in the contemplation stage, where they intend to change in the next 6 months,38 awareness of facilitators may increase and, further, the availability of facilitators may be a meaningful motivator for changing from contemplation to preparation, action, and eventually maintenance.
Previous findings on this topic have been variable; the results reported here support some of the previous work and refute others. From an international perspective, research reports from Flanders, Belgium, showed that the presence of facilities for bicyclists at the workplace, such as showers, shelters, and safe bicycle parking, were significantly associated with bicycling to work.24 Looking at work in the United States, one study conducted in Washington, DC, found that presence of bicycle parking, showers, and lockers was significantly associated with commuting to work by bicycle.25 Handy and Xing,28 in contrast, found that bicycle facilities such as racks and showers close to work were not significantly associated with bicycle commuting in six small US cities located in California and Colorado. Another study in Manhattan, Kansas also found no significant associations between bicycle parking or policies and bicycling to work at least one time per week.27.
The way the bicycling behavior was operationalized may in part explain these differences - Bopp and colleagues27 and Handy and Xing28 both examined bicycle commuting in a typical week. Bicycle commuting as frequently as one time a week or more may not be as influenced by accessible workplaces as compared to any reported bicycle commuting activity. In an attempt to explore more frequent bicycle commuting in the current study, a sensitivity analysis was conducted to assess the association between workplace facilitators and what was defined as “frequent” bicycle commuting, defined as reported bicycling to work and bicycling for transportation 4 or more times in the past month. This is an imperfect approximation of weekly bicycle commuting to work, however, as there was no way to identify the purpose or destination for transportation bicycling frequency in this study. Despite this limitation, the results of this sensitivity analysis were very similar to what was identified in the more basic bicycle commuter analysis, in that workplace facilities are significantly associated with frequent bicycle commuting. Nonetheless, further exploration to determine how much workplace facilities influence the frequency of the behavior over and above other environmental and sociodemographic factors.
Another explanation of the differing findings from similar studies may be in the built environments of the cities where these studies were conducted. Washington, DC is a large metropolitan area with a dense urban core, whereas the cities explored by Bopp and colleagues27 and Handy and Xing28 are smaller university towns with other factors influencing bicycle commuting, such as parking costs.28 The counties in our studies are not necessarily comparable to either Washington, DC or Manhattan, Kansas, but do have diverse urbanicity that may contribute to the differing findings. Indeed, it is known that environmental context is important for urban-based physical activity behaviors such as transport-based and recreational walking.39-41 Therefore, when considering how environments and policy influence physical activity behaviors, it is essential to also understand the context of communities in which the behaviors take place.
Nehme and colleagues26 examined the association between the introduction of new worksite showers and physical activity behaviors, including bicycling to work. Due to small sample size, they were unable to identify if the introduction of showers was associated with changes in bicycle commuting, but they found no effect with more general employee physical activity.26 A simulation study out of Edmonton, Canada found that bicyclists considered showers for cyclists at destinations to be attractive, but determined that secure bicycle parking was much more important.42 The current study did find a significant association between presence of showers and bicycle commuting, but out of all the facilitators examined, showers was the least strongly associated with commuting by bicycle. Similar to the results of Hunt and Abraham,41 parking facilities, such as bicycle lockers and bicycle racks, were the facilities most strongly associated with bicycle commuting, along with policies. However, barriers to bicycling for travel to work or during the workday have been previously explored in populations, and lack of facilities such as showers have been identified as important in encouraging individuals to bicycle to work or during the workday,43, 44 as is a lack bike racks or other method of secure bicycle storage.44.
Somewhat surprisingly, these associations were not modified by gender. Previous studies have shown substantial differences by gender in bicycling prevalence,36 in which factors influence physical activity behaviors,37 and what workplace supports are associated with active commuting,20 which led to an expectation of differences in the present analysis. However, evidence from this study suggests that, among this sample of bicyclists, workplace supports for bicycling are important for both men and women to participate in bicycle commuting. More research is needed to understand if these findings apply across population groups and in differing contexts.
Overall, it is evident that, in this non-representative sample, workplace facilities and policies that support employees to bicycle to work are important for bicycle commuters. Further, the findings indicate that there is not necessarily one particular workplace intervention that is the most important for bicycle commuting – the more bicycle commuting facilitators present, the more likely it will be that an employee bicycles to work. The cost of implementing many of these facilitators would be negligible, particularly the development and implementation of supportive policy and the installation of bicycle racks. Therefore, companies that desire to increase active transportation, and specifically bicycle commuting, among their employees should consider how their infrastructure and workplace culture can be changed to support this behavior. Employees who are supported to engage in active transportation such as bicycling may improve physical and mental health outcomes, and in turn reduce health care costs and absenteeism for employers.10-13
4.1. Strengths and Limitations
This study addresses a gap in the literature for understanding how the presence of workplace facilities and policies are associated with bicycle commuting behavior among employees. This work stands apart from previous research on bicycle commuting to work,24-27 in that the individual facilities were examined independently, as well as via a summary variable, and further, that these results were examined for an interaction by gender. In addition, this study examined bicycling as an independent activity, rather than considering “active commuting”, i.e. bicycling and walking together. This information provides an increased level of detail for helping to understand what facilitators may be of particular value to individuals who are interested in commuting by bicycle.
This study also has limitations. Due to the cross-sectional nature of these data, reverse causality is possible, such that individuals who are interested in bicycling to work may have chosen workplaces that support that behavior. Further, facilitators may have been introduced into the workplace to support bicycle commuters already performing the behavior. In this case, however, one could argue that though one or more bicycle commuters were the catalyst for environmental or policy change, the addition of facilitators that support bicycling to work may in turn influence other employees to take up the behavior as well. Longitudinal studies are warranted to elucidate these potential causal mechanisms. Distance between home and workplace is a potential moderator of this association as well, in that commuting by bicycle may not be a feasible option for individuals that live too far from their workplace location. Hence, the effectiveness of workplace policies to promote bicycling may depend on the proportion of employees that reside close to the given workplace location. Although in this study participants were asked to report their home and work addresses, there was a high amount of missing data for these variables, precluding us from calculating home-to-work distances and including this variable in the analysis. In addition, employer name or industry sector was not assessed. Future research should examine how distance, as well as other barriers such as family responsibilities, home and work neighborhood environments, employment industry sector, and social norms may modify these associations.
The variables under study were assessed via self-report, which may introduce error due to misreporting and response bias. Previous research on this topic have also utilized self-report to examine the association between the workplace environment and bicycling to work,24,25,27,28 indicating a need for research on the association between objectively measured workplace environments and bicycling to work. The use of a convenience sample introduces selection bias that prevents these data from being generalizable to a larger population and may violate some of the assumptions of hypothesis testing. In particular, this sample was majority white, highly educated, high income, and from a limited age range. Comparing to 2017 estimates from the US Census to the study sample from Jefferson County, Alabama, the population were 52.1% white (88.2% in study sample), had a median household income of $49,321 (54.5% had income above $75,000 in study sample), and 20.6% had a bachelor’s degree (43.6% in study sample).45 Data from the 2017 US Census estimates were more comparable to the study population for Travis County, Texas – the population were 74.9% white (79.3% in study sample), had a median household income of $72,884 (45% had a household income above $75,00 in study sample), and 31.6% had a bachelor’s degree (45.7% in study sample).45 However, it is difficult to gauge how comparable these results are to the population of bicyclists, or of transportation and recreation bicyclists as population representative data are not available. In the US, transportation bicyclists are less likely to be high school educated as compared to non-bicyclists,36 indicating a need-based motivation for bicycling for transportation that may not be true of the present sample. Moving forward, more and better research is warranted examining the effect of workplace facilitators among low-income commuter bicyclists, who remain understudied.
4.2. Conclusions
In this sample of bicyclists, workplace facilities and policies were shown to be strongly associated with bicycle commuting, and further the number of facilitators was associated with bicycle commuting in a dose-response fashion. Although other factors certainly influence an individual’s decision to be a bicycle commuter, such as sociodemographic factors and the perceived and actual built environment in their neighborhood and along the commute route,29,36,46 the workplace environment clearly is an important consideration. Workplaces that wish to prioritize active transportation by bicycle should consider providing facilities for safe bicycle storage, as well as accommodations employees to prepare for the workday. In addition, developing and implementing policies that outline this support for active transportation are warranted.
Supplementary Material
Highlights:
All workplace facilitators were significantly associated with bicycle commuting.
Dose-response relationship evident between number of facilitators and bicycle commuting.
Employers should consider infrastructure and policies supporting active commuting.
Acknowledgments
Financial Disclosure
This study was funded by the Michael and Susan Dell Foundation through resources provided at the Michael & Susan Dell Center for Healthy Living, The University of Texas Health Science Center at Houston School of Public Health in Austin. The Michael and Susan Dell Foundation was not involved in any conduct of this study. AKP was supported by a National Research Service Award T32 post-doctoral research fellowship, funded by the NIH, National Heart, Lung, and Blood Institute (#T32HL007055).The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Warburton DER, Bredin SSD. Health benefits of physical activity: a systematic review of current systematic reviews. Curr Opin Cardiol. 2017;32(5):541–556. doi: 10/1097/HCO.0000000000000437. [DOI] [PubMed] [Google Scholar]
- 2.Clarke TC, Norris T, Schiller JS. Early release of selected estimates based on data from 2016 National Health Interview Survey. National Center for Health Statistics. May 2017. Available from: http://www.cdc.gov/nchs/nhis.htm [Google Scholar]
- 3.U.S. Departmetn of Heatlh and Human Services. Physical Activity Guidelines for Americans. 2nd ed. Washington, DC: U.S. Departmetn of Heatlh and Human Services; 2018. [Google Scholar]
- 4.Chaix B, Kestens Y, Duncan S, et al. Active transportation and public transportation use to achieve physical activity recommendations? A combined GPS, accelerometer, and mobility survey study. Int J Behav Nutr Phys Act. 2014; 11:124. doi: 10.1186/s12966-014-0124-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Sahlqvist S, Song Y, Ogilvie D. Is active travel associated with greater physical activity? The contribution of commuting and non-commuting active travel to total physical activity in adults. Prev Med. 2012;55(3):206–211. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Titze S, Merom D, Rissel C, Bauman A. Epidemiology of cycling for exercise, recreation or sport in Australia and its contribution to health-enhancing physical activity. J Sci Med Sport. 2014;17(5):485–490. [DOI] [PubMed] [Google Scholar]
- 7.Oja P, Titze S, Bauman A, et al. Health benefits of cycling: a systematic review. Scand J Med Sci Sports. 2011. ;21(4):496–509. [DOI] [PubMed] [Google Scholar]
- 8.Andersen LB, Schnohr P, Schroll M, Hein HO. All-cause mortality associated with physical activity during leisure time, work, sports, and cycling to work. Arch Intern Med. 2000;160(11):1621–1628. [DOI] [PubMed] [Google Scholar]
- 9.Mansfield TJ, Gibson JM. Estimating Active Transportation Behaviors to Support Health Impact Assessment in the United States. Front Public Health. 2016;4:63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Mueller N, Rojas-Rueda D, Cole-Hunter T, et al. Health impact assessment of active transportation: A systematic review. Prev Med. 2015;76:103–114. [DOI] [PubMed] [Google Scholar]
- 11.Gotschi T Costs and benefits of bicycling investments in Portland, Oregon. J Phys Act Health. 2011;8 Suppl 1:S49–58. [DOI] [PubMed] [Google Scholar]
- 12.Mytton OT, Panter J, Ogilvie D. Longitudinal associations of active commuting with wellbeing and sickness absence. Prev Med. 2016;84:19–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Hendriksen IJ, Simons M, Garre FG, Hildebrandt VH. The association between commuter cycling and sickness absence. Prev Med. 2010;51(2):132–135. [DOI] [PubMed] [Google Scholar]
- 14.United States Census Bureau. 2017. American Community Survey 1-Year Estimates, Table S0801 - commuting characteristics by sex. https://factfinder.census.gov/faces/tableservices/jsf/pages/productview.xhtml?pid=ACS_17_1YR_S0801&prodType=table. Accessed May 9, 2019
- 15.Nehme EK, Pérez A, Ranjit N, Amick BC, Kohl HW. Sociodemographic Factors, Population Density, and Bicycling for Transportation in the United States. J Phys Act Health. 2016;13(1):36–43. [DOI] [PubMed] [Google Scholar]
- 16.Zlot AI, Schmid TL. Relationships among community characteristics and walking and bicycling for transportation or recreation. American Journal of Health Promotion. 2005;19(4):314–317. [DOI] [PubMed] [Google Scholar]
- 17.Dill J. Bicycling for transportation and health: the role of infrastructure. J Public Health Policy. 2009;30 Suppl 1:S95–110. [DOI] [PubMed] [Google Scholar]
- 18.Fraser SD, Lock K. Cycling for transport and public health: a systematic review of the effect of the environment on cycling. Eur J Public Health. December 2011;21(6):738–743. [DOI] [PubMed] [Google Scholar]
- 19.Adlakha D, Hipp AJ, Marx C, et al. Home and workplace built environment supports for physical activity. Am J Prev Med. 2015;48(1): 104–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Kaczynski AT, Bopp MJ, Wittman P. Association of workplace supports with active commuting. Prev Chronic Dis. 2010;7(6):A127. [PMC free article] [PubMed] [Google Scholar]
- 21.Petrunoff N, Wen LM, Rissel C. Effects of a workplace travel plan intervention encouraging active travel to work: outcomes from a three-year time-series study. Public Health. 2016;135:38–47. [DOI] [PubMed] [Google Scholar]
- 22.Bopp M, Hastmann TJ, Norton AN. Active commuting among K-12 educators: a study examining walking and biking to work. J Environ Public Health. 2013;2013:162731. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Bopp M, Child S, Campbell M. Factors associated with active commuting to work among women. Women Health. 2014;54(3):212–231. [DOI] [PubMed] [Google Scholar]
- 24.de Geus B, De Bourdeaudhuij I, Jannes C, Meeusen R. Psychosocial and environmental factors associated with cycling for transport among a working population. Health Educ Res. 2008;23(4):697–708. [DOI] [PubMed] [Google Scholar]
- 25.Buehler R. Determinants of bicycle commuting in the Washington D.C. region: The role of bicycle parking, cyclist showers, and free car parking at work. Transportation Research Part D: Transport and Environment. 2012;17(7):525–531. [Google Scholar]
- 26.Nehme EK, Pérez A, Ranjit N, Amick BC, Kohl HW. The Effect of New Shower Facilities on Physical Activity Behaviors of Employees: A Quasi-experiment. J Phys Act Health. 2017;14(2):98–107. [DOI] [PubMed] [Google Scholar]
- 27.Bopp M, Kaczynski AT, Besenyi G. Active commuting influences among adults. Prev Med. 2012;54(3-4):237–241. [DOI] [PubMed] [Google Scholar]
- 28.Handy SL, Xing Y. Factors correlated with bicycle commuting: A study in six small U.S. cities. International Journal of Sustainable Transportation. 2011;5:91–110. [Google Scholar]
- 29.Porter AK, Kohl HW, Pérez A, Reininger B, Pettee Gabriel K, Salvo D. Perceived Social and Built Environment Correlates of Transportation and Recreation-Only Bicycling Among Adults. Prev Chronic Dis. 2018;15:E135. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Places for Bikes. The best cities for bikes. https://cityratings.peopleforbikes.org/ Accessed May 9, 2019.
- 31.Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Schroeder P, Wilbur M. 2012 National survey of bicyclist and pedestrian attitudes and behavior, volume 3: Methodology report. Vol Report No. DOT HS 811 841 C. Washington, DC: National Highway Traffic Safety Administration; 2013, October. [Google Scholar]
- 33.Moritz WE. Survey of North American bicycle commuters: Design and aggregate results. Transportation Research Record. 1997;1578:91–101. [Google Scholar]
- 34.Hosmer DW, Lemeshow S, Sturdivant RX. Applied logistic regression. 3rd ed. Hoboken, NJ: Wiley; 2013. [Google Scholar]
- 35.Kutner MH, Nachtsheim CJ, Neter J, Li W. Applied linear statistical models. 5th ed. New York, NY: McGraw-Hill/Irwin; 2004. [Google Scholar]
- 36.Porter AK, Salvo D, Pérez A, Reininger B, Kohl HW. Intrapersonal and Environmental Correlates of Bicycling in U.S. Adults. Am J Prev Med. March 2018;54(3):413–418. doi: 10.1016/j.amepre.2017.10.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Valson JS, Kutty VR. Gender differences in the relationship between built environment and non-communicable diseases: A systematic review. J Public Health Res. 2018;7(1):1239. doi: 10.4081/jphr.2018.1239 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Prochaska JO, Velicer WF. The transtheoretical model of health behavior change. Am J Health Promot. 1997, 12:38–38. [DOI] [PubMed] [Google Scholar]
- 39.Salvo D, Reis RS, Stein AD, Rivera J, Martorell R, Pratt M. Characteristics of the built environment in relation to objectively measured physical activity among Mexican adults, 2011. Prev Chronic Dis. 2014;11:E147. doi: 10.5888/pcd11.140047 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Salvo D, Sarmiento OL, Reis RS, et al. Where Latin Americans are physically active, and why does it matter? Findings from the IPEN-adult study in Bogota, Colombia; Cuernavaca, Mexico; and Curitiba, Brazil. Prev Med. 10 2017; 103S:S27–S33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Feuillet T, Charreire H, Menai M, et al. Spatial heterogeneity of the relationships between environmental characteristics and active commuting: towards a locally varying social ecological model. Int J Health Geogr. 2015;14:12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Hunt JD, Abraham JE. Influences on bicycle use. Transportation. 2007;34(4):453. [Google Scholar]
- 43.Tin Tin S, Woodward A, Thornley S, Langley J, Rodgers A, Ameratunga S. Cyclists' attitudes toward policies encouraging bicycle travel: findings from the Taupo Bicycle Study in New Zealand. Health Promot Int. 2010;25(1):54–62. [DOI] [PubMed] [Google Scholar]
- 44.Webster KM, Cunningham CJ. Preparing for bike-sharing: insight from focus groups and surveys, Chattanooga, Tennessee, 2010. Health Promot Pract. 2013;14(1):62–68. [DOI] [PubMed] [Google Scholar]
- 45.United States Census Bureau. 2013-2017 American Community Survey 5-Year Estimates. https://factfinder.census.gov/faces/nav/jsf/pages/community_facts.xhtml. Accessed July 1, 2019.
- 46.Wahlgren L, Schantz P. Exploring bikeability in a metropolitan setting: stimulating and hindering factors in commuting route environments. BMC Public Health. 2012;12:168. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.