Perceptions of activity-supportive environment and motorcycle use among urban Taiwanese adults

Cross-sectional survey data used in this study were obtained through a random-digit dialing telephone-based survey conducted over a period of one month from September to October 2015 by a telephone research company in three Taiwanese cities: Taipei City, the capital of Taiwan (area: 271.7 km²; population: 2,702,315); New Taipei City, the largest city in Taiwan (area: 2052.6 km²; population: 3,966,818); and Kaohsiung City, the second largest city in Taiwan (area: 2947.6 km²; population: 2,778,992). Those three cities have comprehensive transport infrastructure, including public bicycle system, public transportation, and pedestrian sidewalks. Respondents were selected through a stratified random sampling process. The eligible respondents were Mandarin-speaking adults aged 20–64 years with private household telephones. We categorized the respondents according to their cities of residence as Taipei City, New Taipei City, and Kaohsiung City. Furthermore, the respondents were stratified by age (20–29, 30–39, 40–49, 50–59, and 60–64 years) and gender. For each city, the telephone numbers were selected at random by a computer-assisted system. The interviewers, who received two days of training before the survey period, used a standardized questionnaire and were experienced in administering population-based telephonic surveys. For ensuring data reliability, the length of each interview was restricted to 20 min to maintain the respondent’s interest [23]. Of the 5333 adults contacted, 1069 completed the survey (response rate: 20.04%). After data cleaning, 1003 valid responses were obtained. Respondents were offered no rewards for participation in the survey. According to Research Ethics Committee (REC) of National Taiwan Normal University, telephone survey targeting adults has been considered to be minimal-risk for respondents and qualify for a verbal consent process. According to the approved procedure and template, verbal consent was obtained and recorded before the start of each interview. The study protocols were reviewed and approved by the REC of National Taiwan Normal University (REC number: 201504HM005).

The outcome variable was self-reported motorcycle use (riders and passengers were both included); data for this variable were acquired through the question “How many times have you used a motorcycle in the past 7 days?” The frequency distribution was skewed; hence, the variable of motorcycle use was dichotomized as indicators of use (yes) and nonuse (no) in the previous 7 days.

The variables comprising perceived neighborhood environmental factors were determined and measured using the Taiwanese version of the International Physical Activity Questionnaire Environmental Module (IPAQ-E). The IPAQ-E was developed as part of the International Physical Activity Prevalence Study to investigate the relationship between perceptions of the environment and walking or cycling in several countries [24‐26]. The Taiwanese version of the IPAQ-E was used in a previous study on the association between perceptions of the environment and active transportation [22]. The IPAQ-E questionnaire comprises three categories of items, with seven core items, four recommended items, and six optional items [27]. Of these 17 items, our study utilized the following seven, based on previous studies [12, 28]; the definitions of the items are provided within parentheses: (1) residential density (main type of housing in the neighborhood); (2) access to public transportation (10–15 min to a transit stop); (3) presence of sidewalks (presence of sidewalks on most streets); (4) safety from crime at night (crime rate, which may render night-time walks unsafe); (5) aesthetic experiences (the presence of interesting elements that make walking outside attractive); (6) street connectivity (the presence of numerous four-way intersections); and (7) destinations (the presence of numerous places of interest within easy walking distance). One personal item (vehicle ownership), three core items (access to shops, presence of bicycle lanes, access to recreational facilities), two recommended items (traffic safety, visibly active people), and four optional items (sidewalk maintenance, bicycle-lane maintenance, safety from crime during day time, and the safety of cyclists in traffic) were not included as environmental variables. Each question was answered on a 4-point Likert scale ranging from “strongly disagree” to “strongly agree.” An additional option, unknown/uncertain, was also provided; responses with this option were treated as missing data in the analysis. Following previous studies [22, 25], the seven perceived environmental factors were transformed into binary items. Residential density was dichotomized as “detached single-family housing” and “other” (e.g., townhouses, row houses, apartments, and condominiums); the other eight items were dichotomized as “agree” (“strongly agree” and “somewhat agree”) and “disagree” (“somewhat disagree” and “strongly disagree”).

The sociodemographic variables recorded in the survey were gender, age, city of residence, education level, occupation type, marital status, body mass index (BMI), and vehicle ownership (which included household motorcycle and car ownership). These variables were further categorized as follows: Age was divided into five categories: a) 20–29; b) 30–39; c) 40–49; d) 50–59; and e) 60–64 years. Education level was categorized into a) high school degree or lower (including elementary school, junior high school, high school, and vocational school degree) and b) university degree or higher. Occupation type was divided into two categories: a) part-time and b) full-time. Marital status was divided into a) married and b) unmarried (including widowed, separated, and divorced). BMI was calculated from the self-reported weight and height and was categorized as a) not overweight (<24 kg/m²), or b) overweight or obese (≥24 kg/m²), according to the threshold for Taiwan [29]. In addition, the number of vehicles owned was recorded; accordingly, household motorcycle and car ownership was dichotomized as “yes” (if >0) and “no” (if 0).

Table 1 presents the respondent characteristics (mean age: 44.5 ± 12.2 years). The distribution of the three covariates of gender, age, and city of residence in this sample were balanced. Among all respondents, 62.9% had a university degree or a higher level of education, 69.3% had a full-time job, 66.9% were married, 40.2% were overweight or obese, and 87.4% and 79.8% owned one or more motorcycles and cars, respectively. In addition, 63.1% of the respondents had used a motorcycle in the 7 days prior to the survey.

Logistic regression analysis of Model 1, which was unadjusted, demonstrated that four of the seven attributes of perceptions of the environment were significantly associated with motorcycle use (Table 2). Respondents who perceived favorable access to public transport (OR = 0.45; 95% CI: 0.25–0.82), the presence of sidewalks (OR = 0.67; 95% CI: 0.48–0.90), safety from crimes at night (OR = 0.64; 95% CI: 0.46–0.89), and convenient access to destinations (OR = 0.66; 95% CI: 0.47–0.95) were less likely to have used motorcycles in the past 7 days. After adjustment for potential confounders in Model 2, the same four attributes were significantly associated with motorcycle use. Respondents who perceived favorable access to public transport (OR = 0.46; 95% CI: 0.26–0.84), the presence of sidewalks (OR = 0.63; 95% CI: 0.46–0.87), safety from crimes at night (OR = 0.64; 95% CI: 0.45–0.89), and good access to destinations (OR = 0.65; 95% CI: 0.45–0.94) were associated with lower levels of motorcycle use in the 7 days prior to the survey.