Advertisement

Perceived environment attributes, residential location, and walking for particular purposes

      Abstract

      Background

      Identifying environmental factors that can influence physical activity is a public health priority. We examined associations of perceived environmental attributes with walking for four different purposes: general neighborhood walking, walking for exercise, walking for pleasure, and walking to get to and from places.

      Methods

      Participants (n =399; 57% women) were surveyed by mail. They reported place of residence, walking behaviors, and perceptions of neighborhood environmental attributes.

      Results

      Men with the most positive perceptions of neighborhood “aesthetics” were significantly more likely (odds ratio [OR]=7.4) to be in the highest category of neighborhood walking. Men who perceived the weather as not inhibiting their walking were much more likely (OR=4.7) to be high exercise walkers. Women who perceived the weather as not inhibiting their walking were significantly more likely to be high neighborhood walkers (OR=3.8) and those with moderate perceptions of “accessibility” were much more likely to do more walking for pleasure (OR=3.5).

      Conclusions

      Different environmental attributes were associated with different types of walking and these differed between men and women. Approaches to increasing physical activity might usefully focus on those attributes of the local environment that might influence particular subsets of walking behavior.

      Introduction

      Physical activity is now central to population strategies for chronic disease prevention.

      U.S. Department of Health and Human Services. Physical activity and health: a report of the Surgeon General. Atlanta GA: Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 1996

      ,

      Bauman A, Bellew B, Vita P, Brown W, Owen N. Getting Australia active: best practice for the promotion of physical activity. Melbourne, Australia: National Public Health Partnership, 2002

      Studies to identify factors that influence population levels of physical activity are needed to inform intervention development and public health policy.
      • Sallis J.F.
      • Owen N.
      Physical activity and behavioral medicine.
      To guide such studies, ecologic models of health behavior
      • Sallis J.F.
      • Owen N.
      Ecological models of health behavior.
      help to systematize the plethora of factors that are known to be associated with variations in physical activity habits in particular groups and in whole populations.
      • Trost S.G.
      • Owen N.
      • Bauman A.
      • Sallis J.F.
      • Brown W.
      Correlates of adults' participation in physical activity review and update.
      A range of studies has examined the variety of factors that can act to influence physical activity. Most often, these have been in the psychological and social domains.
      • Sallis J.F.
      • Owen N.
      Physical activity and behavioral medicine.
      ,
      • Trost S.G.
      • Owen N.
      • Bauman A.
      • Sallis J.F.
      • Brown W.
      Correlates of adults' participation in physical activity review and update.
      Broader physical environment, societal, and organizational influences are less well understood.
      • Sallis J.F.
      • Owen N.
      Physical activity and behavioral medicine.
      ,
      • Sallis J.F.
      • Owen N.
      Ecological models of health behavior.
      ,
      • Owen N.
      • Leslie E.
      • Salmon J.
      • Fotheringham M.J.
      Environmental determinants of physical activity and sedentary behavior.
      ,
      • Bauman A.
      • Sallis J.F.
      • Owen N.
      Environmental and policy measurement in physical activity research.
      A key feature of ecologic models of health behavior is the role ascribed to physical environment factors in the network of causality of physical activity participation.
      • Sallis J.F.
      • Owen N.
      Ecological models of health behavior.
      ,
      • Trost S.G.
      • Owen N.
      • Bauman A.
      • Sallis J.F.
      • Brown W.
      Correlates of adults' participation in physical activity review and update.
      ,
      • Owen N.
      • Leslie E.
      • Salmon J.
      • Fotheringham M.J.
      Environmental determinants of physical activity and sedentary behavior.
      ,
      • Bauman A.
      • Sallis J.F.
      • Owen N.
      Environmental and policy measurement in physical activity research.
      Specific environmental factors are likely to influence particular physical activity behaviors, given that different types of activity typically take place in distinct settings.
      • Sallis J.F.
      • Owen N.
      Ecological models of health behavior.
      ,
      • Owen N.
      • Leslie E.
      • Salmon J.
      • Fotheringham M.J.
      Environmental determinants of physical activity and sedentary behavior.
      For Australian adults, walking is the most common form of moderate-intensity activity reported in population surveys.

      Australian Bureau of Statistics. Participation in sport and physical activities. Canberra: Australian Bureau of Statistics, 2000

      Walking can readily be incorporated into daily routines and is convenient and accessible to all segments of the community.
      • Morris J.
      • Hardman A.
      Walking to health.
      ,
      • Siegel P.
      • Brackbill R.
      • Heath G.W.
      The epidemiology of walking for exercise implications for promoting activity among sedentary groups.
      Research that identifies environmental factors associated with physical activity is accumulating.
      • Trost S.G.
      • Owen N.
      • Bauman A.
      • Sallis J.F.
      • Brown W.
      Correlates of adults' participation in physical activity review and update.
      ,
      • French S.
      • Story M.
      • Jeffery R.
      Environmental influences on eating and physical activity.
      ,
      • Humpel N.
      • Owen N.
      • Leslie E.
      Environmental factors associated with adults' participation in physical activity a review.
      ,
      • Giles-Corti B.
      • Donovan R.J.
      The relative influence of individual, social and physical environment determinants of physical activity.
      ,
      • Saelens B.
      • Sallis J.F.
      • Frank L.
      Environmental correlates of walking and cycling findings from the transportation, urban design, and planning literatures.
      There are also studies identifying environmental factors that are specifically related to walking.
      • Hovell M.F.
      • Sallis J.F.
      • Hofstetter C.R.
      • Spry V.M.
      • Faucher P.
      • Caspersen C.J.
      Identifying correlates of walking for exercise an epidemiologic prerequisite for physical activity promotion.
      ,
      • Ross C.E.
      Walking, exercising, and smoking does neighborhood matter?.
      ,
      • Brownson R.C.
      • Houseman R.A.
      • Brown D.R.
      • et al.
      Promoting physical activity in rural communities walking trail access, use, and effects.
      ,
      • Ball K.
      • Bauman A.
      • Leslie E.
      • Owen N.
      Perceived environmental aesthetics and convenience, and company are associated with walking for exercise among Australian adults.
      ,
      • Carnegie M.A.
      • Bauman A.
      • Marshall A.
      • Mohsin M.
      • Westley-Wise V.
      • Booth M.L.
      Perceptions of the physical environment, stage of change for physical activity and walking among Australian adults.
      A recent review
      • Humpel N.
      • Owen N.
      • Leslie E.
      Environmental factors associated with adults' participation in physical activity a review.
      found consistent evidence that perceived aesthetic attributes, accessibility of facilities, and opportunities were associated with physical activity. Safety and the weather had less strong associations. In an Australian study,

      Humpel N, Owen N, Leslie E, Marshall A, Bauman A, Sallis JF. Associations of location and perceived environmental attributes with walking in neighborhoods. Am J Health Promotion 2004. In press

      the specific behavior of neighborhood walking demonstrated stronger associations with environmental attributes than did measures of total walking or of total physical activity. Aesthetics, convenience of facilities, and access to services were positively associated with the neighborhood walking for men; convenience was associated with neighborhood walking for women.
      This study reports the patterns of association of an extended range of perceived neighborhood attributes with walking for particular purposes. The primary aim was to examine whether the attributes of local environments found to be associated with general neighborhood walking

      Humpel N, Owen N, Leslie E, Marshall A, Bauman A, Sallis JF. Associations of location and perceived environmental attributes with walking in neighborhoods. Am J Health Promotion 2004. In press

      are associated also with more-specific walking purposes: walking for exercise, walking for pleasure, and walking to get to and from places. Given that consistent gender differences were found in previous studies,

      Humpel N, Owen N, Leslie E, Marshall A, Bauman A, Sallis JF. Associations of location and perceived environmental attributes with walking in neighborhoods. Am J Health Promotion 2004. In press

      ,

      Humpel N, Marshall A, Leslie E, Bauman A, Owen N. Changes in neighborhood walking are related to changes in perceptions of environmental attributes. Ann Behav Med 2004. In press

      associations of environmental attributes with walking were examined separately for men and women.

      Methods

      Participants

      Participants were clients aged >40 years from a health insurance organization that also provides telephone-delivered prevention and disease-management services. They were drawn from a regional coastal Australian city (population 180,000) and its surrounding suburbs. The mail-out survey was sent in the spring. From 982 potential respondents, the initial response was 282 (29%). A reminder letter was then sent out to participants who had not responded within 2 weeks. A total of 429 (43%) returned the survey. Of these, 30 surveys were incomplete, resulting in a final sample of 399.

      Measures

      The survey included questions on age, educational attainment, gender, walking behaviors, perceptions of the neighborhood environment, and participants' postal code. Table 1 provides sample characteristics.
      Table 1Characteristics of sample and distribution by location
      Men % (n)Women % (n)
      Gender42.6 (170)57.4 (229)
      Age
      40–59 years44.6 (75)56.6 (128)
      ≥60 years55.4 (93)43.4 (98)
      Education
      ≤12 years27.3 (45)56.5 (122)
      Trade/technical46.7 (77)23.6 (51)
      University26.1 (43)19.9 (43)
      Location
      Noncoastal28.7 (48)26.0 (59)
      Coastal71.3 (119)74.0 (168)

      Walking

      For neighborhood walking, participants were asked: “How many times a week do you go for a walk for any reason in and around your neighborhood?” “How much time would you usually spend when you do go for a walk in and around your neighborhood?” For walking for exercise, participants were asked: “What is the average number of times per week you spend walking in your neighborhood or elsewhere, for exercise for at least 10 minutes at a time?” “What is the average number of minutes you spend walking each time?” This was repeated for walking for pleasure and for walking to get to and from places. Reported frequency of walking was multiplied by the number of usual minutes, to give an index of estimated minutes of walking each week, for each specific walking type. Reliability of the neighborhood walking item had been examined previously.

      Humpel N, Owen N, Leslie E, Marshall A, Bauman A, Sallis JF. Associations of location and perceived environmental attributes with walking in neighborhoods. Am J Health Promotion 2004. In press

      ,

      Humpel N, Marshall A, Leslie E, Bauman A, Owen N. Changes in neighborhood walking are related to changes in perceptions of environmental attributes. Ann Behav Med 2004. In press

      Neighborhood walking included all walking that occurred in the neighborhood for any purpose; walking for exercise, pleasure, and to get to and from places could have occurred in the neighborhood or elsewhere.

      Perceived environmental attributes

      There were 24 neighborhood environment attribute items, based on findings from a review of studies that assessed relationships between environment attributes and physical activity behaviors,
      • Humpel N.
      • Owen N.
      • Leslie E.
      Environmental factors associated with adults' participation in physical activity a review.
      previous studies,

      Humpel N, Owen N, Leslie E, Marshall A, Bauman A, Sallis JF. Associations of location and perceived environmental attributes with walking in neighborhoods. Am J Health Promotion 2004. In press

      ,

      Humpel N, Marshall A, Leslie E, Bauman A, Owen N. Changes in neighborhood walking are related to changes in perceptions of environmental attributes. Ann Behav Med 2004. In press

      and also items from the Neighborhood Environment Walkability Scale.

      Saelens B, Sallis JF, Black J, Chen D. Preliminary evaluation of the Neighborhood Environment Walkability Scale and neighborhood walking differences in physical activity. Am J Public Health 2003;93:1152–8

      Items were based on semantic differential principles,
      • Brinton J.E.
      Deriving an attitude scale from semantic differential data.
      where the anchors were the most negative and the most positive result for that situation. Participants were asked to check the most appropriate value on a ten-point scale (see Appendix for a full list of study items, factor groupings, and response options). Previously, test–retest reliability had been examined for eight of the environmental attribute items.

      Humpel N, Marshall A, Leslie E, Bauman A, Owen N. Changes in neighborhood walking are related to changes in perceptions of environmental attributes. Ann Behav Med 2004. In press

      Intra-class correlations (absolute agreement) for these items ranged from 0.73 to 0.91.
      Principal component analysis using varimax rotation was used to identify groups of related environmental attributes. The analysis identified four factors accounting for 56% of the variance. Four items were excluded from further analysis as they cross-loaded across several of the factors or did not fit with the interpretation of the factors. One factor (and its eigenvalues) was interpreted as “accessibility” of facilities for walking (eight items, 6.1). Factors interpreted as influences on walking included “aesthetics” (four items, 2.3); “safety” (four items, 1.36); and “weather” (four items, 2.0). Loadings ranged from 0.49 to 0.89 across the four factors. Cronbach's alpha coefficient of internal consistency was calculated for each subscale. All scores were above the 0.70 recommended level: aesthetics=0.81; accessibility=0.88; safety=0.73; and weather=0.77.

      Location by postal code

      Previous Australian studies have found physical activity to be higher among coastal residents, after adjusting for educational attainment and other demographic factors.

      Humpel N, Owen N, Leslie E, Marshall A, Bauman A, Sallis JF. Associations of location and perceived environmental attributes with walking in neighborhoods. Am J Health Promotion 2004. In press

      ,

      Humpel N, Marshall A, Leslie E, Bauman A, Owen N. Changes in neighborhood walking are related to changes in perceptions of environmental attributes. Ann Behav Med 2004. In press

      ,
      • Bauman A.
      • Smith B.
      • Stoker L.
      • Bellew B.
      • Booth M.L.
      Geographical influences upon physical activity participation evidence of a “coastal effect”.
      In Australia, a postal code district is a mail delivery area identified by four digits, used functionally in the same way as are ZIP codes in the United States. A structured query language (SQL) function used Australian Bureau of Statistics 1996 Census data to identify postal areas that intersect the coastline. This variable was coded into noncoastal (27%) and coastal (73%) location. The high proportion of participants with a coastal location is due to the survey area being a long narrow strip between mountains and the ocean.

      Data analysis

      Items in each factor were summed to provide a total score for each category of environmental attribute. These summed scores were then divided by the number of items in each category. This facilitated comparison across the categories, with all having a final score out of ten (Table 2). The scores of aesthetics, accessibility, safety, and weather were transformed into categorical variables with three levels: low (a less positive perception of the environment); moderate; or high (a highly positive perception of the environment). A high score for weather meant that the weather did not inhibit walking. The cutoff points used for these levels were those that most closely approximated the tertiles of the distributions.
      Table 2Mean minutes of walking, by type and by scores on environmental attribute categories
      Men Mean (SD)Women Mean (SD)
      Walking
      Neighborhood187 (181)171 (128)
      Exercise124 (124)137 (113)
      Pleasure32 (58)31 (62)
      To get to places32 (56)29 (48)
      Environment
      The summed environmental scores were divided by the number of items in each category to facilitate comparison across the categories.
      Aesthetics8.19 (1.6)8.33 (1.7)
      Accessibility6.37 (1.9)6.45 (2.0)
      Safety7.88 (1.7)7.72 (1.8)
      Weather6.28 (2.0)6.06 (2.2)
      SD, standard deviation.
      a The summed environmental scores were divided by the number of items in each category to facilitate comparison across the categories.
      All four walking outcome measures were dichotomized at the median score and analyzed separately. The median score for neighborhood walking was 150 minutes per week, and for exercise walking, 120 minutes per week. For walking for pleasure and to get to and from places, the median score was 0 minutes resulting in any, or no walking categories for these two outcomes. Because the study asked about walking for different purposes separately, and an overlap in reporting was a likely consequence, no analyses were conducted using a total measure of walking.
      A series of logistic regression models was used to examine the association between location and the four perceived environment categories, with the four walking outcome variables. All models were adjusted for age and education level. As past studies have found that physical activity differs for men and women,

      U.S. Department of Health and Human Services. Physical activity and health: a report of the Surgeon General. Atlanta GA: Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 1996

      ,
      • Sallis J.F.
      • Owen N.
      Physical activity and behavioral medicine.
      ,

      Humpel N, Owen N, Leslie E, Marshall A, Bauman A, Sallis JF. Associations of location and perceived environmental attributes with walking in neighborhoods. Am J Health Promotion 2004. In press

      all models were stratified by gender.

      Results

      Characteristics of participants are in Table 1. The mean age of the sample was 60 (standard deviation [SD]=11) years with 57.4% being women. The participants were overall a relatively active group, with the mean minutes of neighborhood walking for both men and women being above the recommended “sufficient” activity for health benefits of 150 minutes per week.

      U.S. Department of Health and Human Services. Physical activity and health: a report of the Surgeon General. Atlanta GA: Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 1996

      Table 2 shows mean minutes of types of walking for men and women. There were no statistically significant gender differences.
      A significant difference between the age groups was found for walking for pleasure. The proportion of those aged >60 years (28.8%) who walked for pleasure was significantly less than for those <60 years (46.2%) (χ2=12.24, p <0.001).
      A higher proportion of those with the most positive perceptions for all four environmental perception categories reported more neighborhood walking. Higher proportions of neighborhood walkers were found among those with high perceptions for aesthetics (66.7%; χ2=17.08, p <0.001) and among those reporting that weather is not an influence (67.4%; χ2=27.98, p <0.001). Significantly higher proportions of those walking for exercise were found among those with the most positive perceptions for all four environmental perception categories. For example, 74.8% (χ2=39.85, p <0.001) of those reporting that weather did not inhibit walking were high exercise walkers. A higher proportion of those with the most positive perceptions for accessibility reported more walking for pleasure (45.2%; χ2=7.28, p <0.05). No significant differences in proportions were found for walking to get from place to place.

      Place of residence

      No significant gender or education differences were found for specific coastal versus noncoastal location identified by postal code. The proportion of participants aged >60 years living in a coastal location (78.0%) was significantly higher than for those aged <60 years (68.7%; χ2=4.37, p <0.05). A higher proportion of coastal than noncoastal residents were in the higher-level group for each type of walking. This difference was significant for neighborhood walking, with 57.5% coastal participants and 38.5% noncoastal participants reporting a high level of neighborhood walking (χ2=11.01, p <0.001).
      Participants living in coastal locations (mean [M]=189 minutes) walked significantly more minutes in their neighborhood (F(1,382)=5.10, p <0.05) than did participants in noncoastal locations (M=149 minutes). Coastal residents (M=139 minutes) reported more minutes of walking for exercise (F(1,385)=5.10, p <0.05) compared to noncoastal (M=109 minutes). Differences in location for mean scores for the environmental attributes of aesthetics, accessibility, safety, and weather were nonsignificant. Since coastal location implies that a beach is nearby, the item asking about walking distance to a beach or lake was analyzed separately (see Appendix, item 3). Participants reporting that a beach/lake was within easy walking distance reported significantly more neighborhood walking minutes (M=224) than did those reporting a beach/lake was not within walking distance (M=139; F(2,379)=11.00, p <0.000); significantly more exercise walking (M=163 compared to M=100 minutes; F(2,382)=9.72, p <0.00); and significantly more walking for pleasure compared to those perceiving that a beach/lake is not within walking distance (M=33 and M=21, respectively; F(2,380)=3.88, p <0.02).

      Multivariate models of walking behaviors

      Neighborhood walking

      Men with the most positive perceptions about the aesthetic nature of the environment were more than seven times more likely to be high neighborhood walkers (Table 3). Those men who reported that the weather was not inhibiting their walking habits were more than four times as likely to be high neighborhood walkers. Accessibility of facilities for walking demonstrated a negative relationship with neighborhood walking for men. Women who reported that the weather was not inhibiting their walking habits were more than three times as likely to be in the high neighborhood walking category (Table 4). Women living in a coastal location were three times more likely to be high neighborhood walkers. Men evidenced the same pattern regarding location, although the odds ratio was nonsignificant. No evidence of a relationship between safety and neighborhood walking was found for men or women.
      Table 3Odds ratios (95% CI) for environmental variables and likelihood of men being high walkers
      Neighborhood walkingExercise walkingPleasure (social) walkingWalking to get to and from places
      Location
      Noncoast1.001.001.001.00
      Coast1.69 (0.69–4.18)1.72 (0.70–4.19)1.59 (0.64–3.95)0.94 (0.40–2.19)
      Aesthetics
      Low1.001.001.001.00
      Moderate1.92 (0.63–5.86)2.06 (0.68–6.26)1.47 (0.78–4.57)0.94 (0.33–2.64)
      High7.43 (1.92–28.82)
      p < 0.01.
      3.86 (1.03–14.46)
      p < 0.05.
      1.45 (0.38–5.49)0.64 (0.19–2.17)
      Accessibility
      Low1.001.001.001.00
      Moderate1.14 (0.39–3.29)2.57 (0.89–7.46)0.91 (0.33–2.51)1.10 (0.43–2.85)
      High0.30 (0.09–0.91)
      p < 0.05.
      0.70 (0.25–2.01)2.02 (0.68–5.98)1.40 (0.50–3.87)
      Safety
      Low1.001.001.001.00
      Moderate0.98 (0.31–3.01)1.23 (0.40–3.82)0.54 (0.18–1.65)0.58 (0.20–1.68)
      High0.91 (0.27–3.06)1.04 (0.31–3.58)0.22 (0.06–0.78)
      p < 0.05.
      0.54 (0.17–1.67)
      Weather
      Strong influence1.001.001.001.00
      Moderate4.09 (1.44–11.66)
      p < 0.01.
      4.08 (1.42–11.74)
      p < 0.01.
      1.37 (0.51–3.68)0.92 (0.59–4.29)
      Not an influence4.71 (1.60–13.91)
      p < 0.01.
      5.48 (1.83–16.38)
      p < 0.01.
      0.58 (0.20–1.69)1.60 (0.40–2.19)
      Age
      40–59 years1.001.001.001.00
      ≥60 years1.59 (0.66–3.87)0.95 (0.40–2.25)0.33 (0.14–0.81)
      p < 0.05.
      1.90 (0.84–4.31)
      CI, confidence interval.
      * p < 0.05.
      ** p < 0.01.
      Table 4Odds ratios (95% CI) for environmental variables and likelihood of women being high walkers
      Neighborhood walkingExercise walkingPleasure (social) walkingWalking to get to and from places
      Location
      Noncoast1.001.001.001.00
      Coast3.32 (1.51–7.29)
      p < 0.01.
      1.40 (0.62–3.18)1.65 (0.72–3.82)1.83 (0.87–3.85)
      Aesthetics
      Low1.001.001.001.00
      Moderate1.74 (0.65–4.62)1.78 (0.63–5.02)0.80 (0.28–2.24)0.59 (0.23–1.48)
      High1.12 (0.41–3.12)0.75 (0.25–2.26)0.84 (0.29–2.42)0.84 (0.31–2.25)
      Accessibility
      Low1.001.001.001.00
      Moderate1.12 (0.47–2.66)1.31 (0.51–3.42)3.51 (1.64–9.15)
      p < 0.05.
      1.36 (0.58–3.19)
      High1.76 (0.70–4.47)1.42 (0.52–3.88)2.61 (0.97–6.97)1.61 (0.67–3.86)
      Safety
      Low1.001.001.001.00
      Moderate0.66 (0.26–1.63)0.69 (0.26–1.84)1.68 (0.65–4.33)0.74 (0.31–1.76)
      High1.09 (0.40–2.96)2.64 (0.90–7.82)1.13 (0.40–3.19)0.56 (0.22–1.45)
      Weather
      Strong influence1.001.001.001.00
      Moderate1.93 (0.86–4.36)1.03 (0.45–2.36)0.62 (0.65–4.33)0.77 (0.35–1.72)
      Not an influence3.84 (1.68–8.77)
      p < 0.01.
      7.68 (3.03–19.46)
      p < 0.001.
      0.93 (0.41–2.07)0.73 (0.34–1.57)
      Age
      40–59 years1.001.001.001.00
      ≥60 years1.06 (0.52–2.16)0.45 (0.21–1.00)0.41 (0.19–0.87)
      p < 0.05.
      1.00 (0.51–1.96)
      CI, confidence interval.
      * p < 0.05.
      ** p < 0.01.
      *** p < 0.001.

      Walking for exercise

      Men with a high score on aesthetics were nearly four times as likely, and those with the highest scores for weather (weather did not inhibit) were nearly six times more likely to walk for exercise (Table 3). For women, those with the highest score for weather were over seven times more likely to walk for exercise (Table 4). Whether men or women lived in a coastal location was not associated with more walking for exercise.

      Walking for pleasure and to get to and from places

      Men who perceived their environment as highly safe for walking were less likely to walk for pleasure (odds ratio [OR]=0.22, 95% confidence interval [CI], 0.06–0.78, p <0.05; Table 3). Age was also found to be associated with walking for men. Men aged ≥60 years were less likely to walk for pleasure (OR=0.33, CI, 0.14–0.80, p <0.05). Women (Table 4) with moderately positive perceptions about accessibility were more than three times more likely to walk for pleasure (OR=3.51, CI, 1.64–9.15, p <0.01), and women aged ≥60 were less likely to walk for pleasure (OR=0.41, CI, 0.19–0.87, p <0.05). These findings on women's walking for pleasure need to be viewed with caution as the logistic regression model reported borderline significance (χ2=20.87, p =0.052). The logistic regression model found neither perceived environmental attributes nor location to be associated with walking to get to and from places.

      Discussion

      Ecologic models of health behavior suggest that different environmental attributes may be associated with different physical activity behaviors.
      • Sallis J.F.
      • Owen N.
      Ecological models of health behavior.
      ,
      • Owen N.
      • Leslie E.
      • Salmon J.
      • Fotheringham M.J.
      Environmental determinants of physical activity and sedentary behavior.
      This may apply to the same behavior performed for different purposes, such as walking.
      • Saelens B.
      • Sallis J.F.
      • Frank L.
      Environmental correlates of walking and cycling findings from the transportation, urban design, and planning literatures.
      In this study, different environmental attributes were related to the different types of walking. Weather, aesthetics, accessibility, and location were associated with neighborhood walking. Weather and aesthetics were found to be associated with walking for exercise. Safety and accessibility were associated with walking for pleasure. This study supports the need to focus the broad framework of ecologic models into models for specific behaviors. By exploring behavior as specifically as possible, and not using total or generic measures of activity, more can be learned about the environment–behavior relationship. Further studies are needed on the development of appropriate measures for specific environmental attributes to build our understanding of these relationships.
      Safety did not prove to be an important influence on neighborhood or exercise walking for this sample of adults. This could be attributable to participants perceiving that they live in a low-crime area. An unexpected finding was that, of the four perceived environment factors, weather demonstrated the strongest association with walking for both men and women. Those participants who reported the weather as not inhibiting their walking were most likely to participate in high levels of neighborhood and exercise walking. Although the weather is often reported as a barrier to physical activity, few studies have examined this variable using multivariate models. If the strong association of perceptions of the influence of the weather can be found in other samples, this may have some implications for the promotion of physical activity. This potential determinant is not modifiable, although individual perceptions about the influence of weather may be.
      The strong relationship of coastal location with neighborhood walking for women is puzzling, particularly because of the weak associations found for perceived environmental attributes. The location attribute is objectively measured and the environmental attributes are perceptions; this could perhaps be an important difference. In a previous study,

      Humpel N, Owen N, Leslie E, Marshall A, Bauman A, Sallis JF. Associations of location and perceived environmental attributes with walking in neighborhoods. Am J Health Promotion 2004. In press

      the location association was found for men. The gender difference in the findings of the two studies may be related to the age differences of the two samples. In the previous study the mean age was 43 years with very few participants aged >60 years, whereas in this study, the mean age was 60 years. In this sample, a larger proportion of participants aged >60 years live in a coastal location as compared to those aged <60 years. Environmental influences on walking may change for men and women with increasing age and retirement from work.
      Although not the main focus of this study, significant association of age with walking for pleasure was found for both men and women. But, contrary to might be expected, those aged >60 years were less likely to walk for pleasure. It may be that around retirement age and older, walking may be perceived as being more for the purpose of exercise for health reasons, as it is often the advice given by general practitioners to reduce risk factors for disease.
      Australian adults have been found to be more active in coastal locations, but the reasons for this are unknown. The categories of environmental attributes used in this study did not explain the variations in walking by coastal and noncoastal participants. The specific beach or lake within easy walking distance item was found to be associated with increased neighborhood, exercise, and pleasure walking. This relationship of access to water and increased walking adds further evidence of the coastal hypothesis,
      • Bauman A.
      • Smith B.
      • Stoker L.
      • Bellew B.
      • Booth M.L.
      Geographical influences upon physical activity participation evidence of a “coastal effect”.
      although it is not clear if this is the result of better aesthetics rather than better access.
      Limitations of this study are its cross-sectional design, reliance on self-report of walking and perceptions of environmental attributes, the relatively low response rate, and the potential for cognitive overlap in the reporting of walking for different purposes. Participants may also have unintentionally overstated their levels of walking with this multiple measurement of walking format. Even with these limitations, consistent associations in this and other studies, particularly for aesthetics, implies the possibility of a causal relationship. A further limitation may have been the primarily coastal region. The geographical nature of this district is that of a long narrow strip between the mountains and the coastline, resulting in limited variation in the environment and possibly a generally more aesthetic appeal overall. While this may have limited the variation in the sample, at this stage of research into environment–physical activity relationships, these findings add to the growing knowledge base.
      Future studies could include measures of perceptions of the environment, and additionally, objectively assessed measures derived from geographic information systems (e.g., housing density, road data).
      • Humpel N.
      • Owen N.
      • Leslie E.
      Environmental factors associated with adults' participation in physical activity a review.
      ,
      • Saelens B.
      • Sallis J.F.
      • Frank L.
      Environmental correlates of walking and cycling findings from the transportation, urban design, and planning literatures.
      Measuring both perceived and objective variables may shed light on whether it is the actual supports for physical activity that are accessible in the local environment or perceptions of the environment for physical activity that are most likely to influence behavior.
      Although modest, the associations identified in this study do add to the body of data on the influence of perceived environmental attributes on walking. The gender differences observed emphasize that different aspects of the environment may be of differing importance for men and women. Prospective studies are needed to determine whether the cross-sectional associations of environmental attributes with physical activity that have so far been reported are likely to be cause–effect relationships.
      • Bauman A.E.
      • Sallis J.F.
      • Dzewaltowski D.A.
      • Owen N.
      Toward a better understanding of the influences on physical activity the role of determinants, correlates, causal variables, mediators, moderators, and confounders.
      Public health approaches to increasing physical activity might consider strategies that focus on the importance of particular attributes of local environments that may impact on particular physical activity behaviors such as walking. Innovative strategies are particularly needed in countries like Australia, where adults' physical activity levels are declining.
      • Bauman A.
      • Armstrong T.
      • Davies J.
      • et al.
      Trends in physical activity participation and the impact of integrated campaigns among Australian adults, 1997–1999.

      Acknowledgements

      This study was supported by Carelink, a division of the Australian Health Management Group, a registered health benefits organization.

      Appendix.

      Tabled 1
      Perceived environmental attributes measured in the study, factor groupings, and response options
      Figure thumbnail FX1

      References

      1. U.S. Department of Health and Human Services. Physical activity and health: a report of the Surgeon General. Atlanta GA: Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, 1996

      2. Bauman A, Bellew B, Vita P, Brown W, Owen N. Getting Australia active: best practice for the promotion of physical activity. Melbourne, Australia: National Public Health Partnership, 2002

        • Sallis J.F.
        • Owen N.
        Physical activity and behavioral medicine.
        Sage Publications, Thousand Oaks CA1999
        • Sallis J.F.
        • Owen N.
        Ecological models of health behavior.
        in: Glanz K. Lewis F.M. Rimer B.K. Health behavior and health education: theory, research, and practice. 3rd ed. Jossey-Bass, San Francisco2002: 462-484
        • Trost S.G.
        • Owen N.
        • Bauman A.
        • Sallis J.F.
        • Brown W.
        Correlates of adults' participation in physical activity.
        Med Sci Sports Exerc. 2002; 34: 1996-2001
        • Owen N.
        • Leslie E.
        • Salmon J.
        • Fotheringham M.J.
        Environmental determinants of physical activity and sedentary behavior.
        Exerc Sports Sci Rev. 2000; 28: 153-158
        • Bauman A.
        • Sallis J.F.
        • Owen N.
        Environmental and policy measurement in physical activity research.
        in: Welk G. Physical activity assessment for health: related research. Human Kinetics, Champaign IL2002: 241-251
      3. Australian Bureau of Statistics. Participation in sport and physical activities. Canberra: Australian Bureau of Statistics, 2000

        • Morris J.
        • Hardman A.
        Walking to health.
        Sports Med. 1997; 23: 306-322
        • Siegel P.
        • Brackbill R.
        • Heath G.W.
        The epidemiology of walking for exercise.
        Am J Public Health. 1995; 85: 706-710
        • French S.
        • Story M.
        • Jeffery R.
        Environmental influences on eating and physical activity.
        Annu Rev Public Health. 2001; 22: 309-335
        • Humpel N.
        • Owen N.
        • Leslie E.
        Environmental factors associated with adults' participation in physical activity.
        Am J Prev Med. 2002; 22: 188-199
        • Giles-Corti B.
        • Donovan R.J.
        The relative influence of individual, social and physical environment determinants of physical activity.
        Soc Sci Med. 2002; 54: 1793-1812
        • Saelens B.
        • Sallis J.F.
        • Frank L.
        Environmental correlates of walking and cycling.
        Ann Behav Med. 2003; 25: 80-91
        • Hovell M.F.
        • Sallis J.F.
        • Hofstetter C.R.
        • Spry V.M.
        • Faucher P.
        • Caspersen C.J.
        Identifying correlates of walking for exercise.
        Prev Med. 1989; 18: 856-866
        • Ross C.E.
        Walking, exercising, and smoking.
        Soc Sci Med. 2000; 51: 265-274
        • Brownson R.C.
        • Houseman R.A.
        • Brown D.R.
        • et al.
        Promoting physical activity in rural communities.
        Am J Prev Med. 2000; 18: 235-241
        • Ball K.
        • Bauman A.
        • Leslie E.
        • Owen N.
        Perceived environmental aesthetics and convenience, and company are associated with walking for exercise among Australian adults.
        Prev Med. 2001; 33: 434-440
        • Carnegie M.A.
        • Bauman A.
        • Marshall A.
        • Mohsin M.
        • Westley-Wise V.
        • Booth M.L.
        Perceptions of the physical environment, stage of change for physical activity and walking among Australian adults.
        Res Q Exerc Sport. 2002; 73: 146-155
      4. Humpel N, Owen N, Leslie E, Marshall A, Bauman A, Sallis JF. Associations of location and perceived environmental attributes with walking in neighborhoods. Am J Health Promotion 2004. In press

      5. Humpel N, Marshall A, Leslie E, Bauman A, Owen N. Changes in neighborhood walking are related to changes in perceptions of environmental attributes. Ann Behav Med 2004. In press

      6. Saelens B, Sallis JF, Black J, Chen D. Preliminary evaluation of the Neighborhood Environment Walkability Scale and neighborhood walking differences in physical activity. Am J Public Health 2003;93:1152–8

        • Brinton J.E.
        Deriving an attitude scale from semantic differential data.
        in: Snider J. Osgood C. Semantic differential technique. Aldine Publishing, Chicago1976: 467-473
        • Bauman A.
        • Smith B.
        • Stoker L.
        • Bellew B.
        • Booth M.L.
        Geographical influences upon physical activity participation.
        Aust N Z J Public Health. 1999; 23: 322-324
        • Bauman A.E.
        • Sallis J.F.
        • Dzewaltowski D.A.
        • Owen N.
        Toward a better understanding of the influences on physical activity.
        Am J Prev Med. 2002; 23: 5-14
        • Bauman A.
        • Armstrong T.
        • Davies J.
        • et al.
        Trends in physical activity participation and the impact of integrated campaigns among Australian adults, 1997–1999.
        Aust N Z J Public Health. 2003; 27: 76-79