Atmospheric Environment 43 (2009) 5791–5795

Contents lists available at ScienceDirect

Atmospheric Environment journal homepage: www.elsevier.com/locate/atmosenv

Perceptions of air pollution during the work-related commute by adults in Queensland, Australia Hannah M. Badland a, *, Mitch J. Duncan b a

Centre for Physical Activity and Nutrition Research, Faculty of Health and Environmental Sciences, Auckland University of Technology, Private Bag 92006, Auckland 1142, New Zealand b Institute for Health and Social Science Research, CQUniversity Australia, Australia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 15 May 2009 Received in revised form 29 July 2009 Accepted 30 July 2009

Background: This study examines perceived health effect risks from air pollution during the work-related commute by socio-demographic and travel mode variables. Methods: Population-representative survey data were collected by telephone from 745 adults from Queensland, Australia. Variables assessed included socio-demographics, usual work travel mode, commute perceptions, and perceived health effects associated with air pollution during the work-related commute. Results: Approximately 45% of the sample perceived air pollution negatively affected health outcomes when commuting to/from work, yet only 13% recognised air pollution as a major barrier to walking or cycling to/from work. No relationships existed between usual travel mode to/from work and perceived health risks associated with air pollution. Those higher educated or living in major cities were more likely to recognise air pollution harmed their health during their work-related commute when compared to respective referent categories (p  0.05). Conclusions: Recognition of health risks from air pollution during the work-related commute was high, and awareness did not differ by travel mode. For the majority, air pollution was not a primary barrier for walking or cycling to/from work. Ó 2009 Elsevier Ltd. All rights reserved.

Keywords: Risk perception Air pollution Commute Travel mode Questionnaire

1. Introduction Air pollution is a serious threat to public health, with estimations of up to 1.6 billion people internationally being exposed to substandard air quality within urban environments (World Health Organization, 2005a). Adverse health effects resulting from air pollution exposure include increased risk of cardiovascular disease, respiratory illnesses, certain cancers, and mortality (Brook et al., 2004; Pope and Dockery, 2006). Within urban settings air quality is reducing, largely because of the increasing reliance on private automobiles for travel (Frank, 2000), and cars have been identified as the primary source of pollutants in on-road and roadside environments (Bickerstaff and Walker, 2001). Accordingly, individuals travelling by any transport mode either on or near roadways are exposed to levels of air pollution that may damage their health (Riediker et al., 2004; World Health Organization, 2005a), and those who travel by car and public transport are more likely to be adversely affected when compared with people who walk or cycle (Pope et al., 2002). * Corresponding author. Tel.: þ64 9 921 9999x7630; fax: þ64 9 921 9746. E-mail address: [email protected] (H.M. Badland). 1352-2310/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.atmosenv.2009.07.050

Recently there has been a movement within transport, built environment, and public health disciplines, as well as various government agencies to promote public and active transport (e.g., walking and cycling) modes. These transport modes present an opportunity to accumulate bouts of physical activity, in addition to emitting lower total levels of air pollution when compared with travelling by private automobile (Rissel, 2005). Although walking and cycling for transport commonly occurs along roadside environments where individuals are exposed to poorer air quality (World Health Organization, 2004). It has been suggested that the health benefits gained from participating in active transport modes in these settings may outweigh the detrimental effects from air pollution exposure (Rissel, 2005; Sharman et al., 2004). Accordingly, reducing private automobile use and increasing public and active transport modal shares for short journeys, such as commuting to work, is in an important step to improving overall air quality and health outcomes. Given efforts to promote sustainable modes of transport and the potential for individuals to be exposed to air pollution during commute journeys, it is important to examine public awareness of the health risks associated with poor air quality. Perceptions of the health risks from air pollution are framed within a subjective decision making process of affect that seeks to determine the quality

5792

H.M. Badland, M.J. Duncan / Atmospheric Environment 43 (2009) 5791–5795

(good or bad) of the event, with interpretation of the risk being calculated at the individual-level through both experiential responses and analytical deductions for the variable being examined. Level of risk is determined by the judgement of what the individual thinks and feels about the event (i.e., air pollution), and if the activity is viewed as having negative results (i.e., poor health outcomes) the risks are perceived as higher and benefits lower (Slovic and Peters, 2006). There is general awareness of the ill effects resulting from air pollution exposure (Bianco et al., 2008), with approximately one third of the population recognising the contribution poor air quality has on cardiovascular disease risk (Day, 2006). Yet, there is a paucity of information describing perceptions towards air pollution specific to travelling on or near roadside environments, and how this is associated with travel patterns of individuals (Bickerstaff and Walker, 2001). An important next stage for promoting active transport is to investigate attitudes towards air pollution, health outcomes, travel behaviours, and their interrelationships specific to the work-related commute journey. Understanding these perceptions will extend the active transport literature base and inform future work-related commuting interventions. As such, the objective of this paper is to investigate the health risk perceptions of air pollution during the journey to/from work with demographic and work-related travel mode variables in a population-representative sample of adults drawn from Queensland, Australia. 2. Methods 2.1. Design and sample Participants were adults aged 18 years and over residing in Queensland, Australia, and able to be contacted by a landline telephone. Data were collected in July–August 2008, using a computer-assistedtelephone-interview (CATI) survey conducted by the Population Research Laboratory at CQUniversity. Participants were selected at random from electronic databases of current landline telephone numbers held by the Population Research Laboratory. For sampling purposes the state of Queensland was separated into two separate areas, Brisbane and Moreton Statistical subdivisions and the rest of Queensland, with half of the sample being drawn from each site. A twostage selection process was applied to select participants: firstly, the household was selected at random, and secondly, a sex was randomly identified for each household and the corresponding individual who resided at the address was selected. If two eligible individuals resided in the household then the most recent birthday method was used to identify the potential participant. If interviewers were unable to establish contact with any individual on the first attempt the number was attempted a minimum of five times before the number was removed from the sample. This was a population-representative survey for the state of Queensland, Australia and in comparison to the most recently available Australian Census Data the current sample had a slight overrepresentation of older adults (Hanley and Mummery, 2008). CQUniversity’s Human Research Ethics Committee provided ethical clearance for the study. No data are available from those individuals who declined to participate in the current study. 2.2. Measures 2.2.1. Socio-demographics Participants provided information on socio-demographic items including sex, age, years of education, annual household income, and postcode of residential address. Postcode was used to further classify individuals according to their degree of urbanisation using the Australian Standard Geographical Classification (ASGC) system (Australian Institute of Health and Welfare, 2004). The ASGC has been used in research examining health outcomes and behaviours by

degree of urbanisation (Eakin et al., 2007; Lagace´ et al., 2007). The ASGC contains five categories: major cities, inner regional areas, outer regional areas, remote areas, and very remote areas. For the present study, these groups were collapsed into three categories: major cities, inner regional areas, and outer regional areas (including remote and very remote settings). 2.2.2. Travel behaviours A single item was used to determine usual mode of travel to/from a participant’s place of employment: ‘‘In the last seven days, what was your main mode of transport to and from your main place of employment?’’ Response options were: car, motorcycle/scooter, bus, train, ferry, bicycle, or walk. Car and motorcycle/scooter responses were collapsed into ‘private motorised’; bus, train, and ferry were collapsed in ‘public transport’; and walk and cycle were collapsed into ‘active transport’ modes. This question has been used in previous travel research (Badland et al., 2008). All participants were asked about their route selection during travel: ‘‘I take different travel routes so I can reduce the amount of air pollution in the area I travel through’’. Those who engaged in active transport modes were asked to comment on an additional statement, ‘‘The majority of my commute route to my main place of employment is performed on off-road pathways (separate from traffic) or low volume streets’’. Each item was answered using a five-point Likert scale from strongly agreed (¼1), neither agreed nor disagreed (¼3), to strongly disagreed (¼5). The responses were dichotomously collapsed into agreed (strongly agreed þ agreed) and disagreed (strongly disagreed þ disagreed) with the neutral category being excluded from further analysis. No reliability and validity tests were conducted with the travel behaviour or air pollution perception items in the questionnaire or with the interviewers administering the survey. The questionnaire is available from the authors on request. 2.2.3. Perceptions of air pollution Air pollution along the roadway during the work-related commute was assessed with regards to the perceived health effects and its presence as a barrier to active transport. Two items assessed the perceived health effects of air pollution: ‘‘Exposure to traffic exhaust and other pollution when travelling to and from work negatively affects my health’’ and ‘‘Exposure to air pollution when travelling increases my risk of cardiovascular disease’’. A third item ‘‘Exposure to traffic exhaust and other pollution is a major barrier to me walking and cycling to and from work’’ assessed the perceptions of air pollution as a barrier to engaging in active transport modes. Responses were assessed on a five-point Likert scale from strongly agreed (¼1), neither agreed nor disagreed (¼3), to strongly disagreed (¼5). The responses were dichotomously collapsed into agreed (strongly agreed þ agreed) and disagreed (strongly disagreed þ disagreed) with the disagreed and neutral categories being excluded from further analysis. 2.2.4. Statistical analysis Comparisons within groups were made using the Kruskall–Wallis (age, income, education, region, work transport mode variables) and the Mann–Whitney (sex, taking a different route to avoid air pollution, commute route being off road) tests to examine differences for the three air pollution items. Mutually adjusted odds ratios (OR) were derived by using multinomial logistic regression to examine relationships between the independent air pollution variables and dependent demographic and travel items. Statistical comparisons were made using SPSS version 16.0 and the analysis was performed in January 2009. An a ¼ 0.05 was used to define statistical significances. 3. Results A total of 1243 individuals completed the survey with an overall response rate of 37.1%. This response rate was similar to other

H.M. Badland, M.J. Duncan / Atmospheric Environment 43 (2009) 5791–5795

recently conducted CATI surveys (Badland and Schofield, 2008; Hutto et al., 2008). Initially, those who were not employed (n ¼ 498) were removed from the sample, and the remaining sample of working adults (n ¼ 745) was weighted by age to align with census data. Overall, the final sample of working adults consisted of slightly more men than women (56.1% versus 43.9%), the majority of respondents were between 18 and 34 years of age, 39.1% had at least 15 years of formal education, the modal household income was AUD  $100,000 per annum, 50.8% lived in a major city, and 82.5% of the sample used private automobiles to commute to work (Table 1). Approximately half of the sample recognised that air pollution exposure while travelling to work negatively affected their overall health (45.7%) and increased the risk of cardiovascular disease (44.2%). Substantially less people perceived air pollution as a major barrier to walking or cycling to work (13.3%) (Table 1). Significant differences in all outcome variables existed (p  0.001) regarding taking different work-related commute routes to avoid air pollution exposure. Significant relationships also existed across groups for years of education and degree of urbanisation for one or more of the air pollution items. Mutually adjusted logistic regression analyses identified those who were most educated were more likely to agree that air pollution

5793

exposure during the work-related commute negatively affected their health when compared to the least educated group (OR ¼ 1.87; 95% CI ¼ 1.18–2.97). Those who lived in inner (OR ¼ 0.55; 95% CI ¼ 0.39– 0.78) and outer (OR ¼ 0.48; 95% CI ¼ 0.31–0.75) regional areas were less likely to recognise air pollution was a health concern during the work-related commute in comparison with residents of major cities (Table 2). Poor air quality was also significantly less of a barrier to walking or cycling to work for those who lived in inner regional locations (OR ¼ 0.53; 95% CI ¼ 0.30–0.92) compared with respondents from major cities. When compared with the referent category, those who disagreed with selecting different routes to avoid air pollution were less likely to report that poor air quality was a barrier to walking or cycling to work (OR ¼ 0.08; 95% CI ¼ 0.01–0.89). No significant differences existed within groups for perceptions of air pollution exposure increasing the risk of cardiovascular disease. 4. Discussion In line with existing air pollution-health risk evidence (Bianco et al., 2008; Bickerstaff and Walker, 2001; Day, 2006), a large portion of the sample recognised that poor air quality had negative health consequences during the journey to and from work, and this

Table 1 Air pollution perceptions by response frequencies. Exposure to traffic exhaust and other pollution when travelling to and from work negatively affects my health Sample n

Agree %

p-value

n

%

340

45.7

Total sample (n ¼ 745) Sex Male Female

418 347

56.1 43.9

194 146

57.2 42.8

Age group 18–34 years 35–44 years 45–54 years 55 years

268 185 184 108

36.0 24.8 24.7 14.5

132 87 78 43

38.9 25.7 22.8 12.6

Annual household income (AUD) $26,000 51 $26,001–$52,000 77 $52,001–$100,000 198 $100,000 229 Prefer not to answer 190

6.8 10.4 26.6 30.7 25.5

23 37 89 105 86

6.9 11.0 26.3 31.0 24.8

Education