Why some walk and others don't: exploring interactions of perceived safety and social neighborhood factors with psychosocial cognitions

Health Education Research, Apr 2013

Although physical activity is often believed to be influenced by both environmental and individual factors, little is known about their interaction. This study explores interactions of perceived safety and social neighborhood factors with psychosocial cognitions for leisure-time walking. Cross-sectional data were obtained from residents (age 25–75 years) of 212 neighborhoods in the South-East of the Netherlands, who participated in the Dutch GLOBE study in 2004 (N = 4395, survey response 64.4%). Direct associations of, and interactions between perceived neighborhood safety, social neighborhood factors (social cohesion, social network and feeling at home) and psychosocial cognitions (attitude, self-efficacy, social influence and intention) on two outcomes of leisure-time walking [yes versus no (binary), and among walkers: minutes per week (continuous)] were analyzed in multilevel regression models. The association between attitude and participating in leisure-time walking was stronger in those who felt less at home in their neighborhood. Social influence and attitude were stronger associated with participation in leisure-time walking in those who sometimes felt unsafe in their neighborhood. A positive intention was associated with more minutes walked in those who perceived their neighborhood as unsafe among those who walked. Only limited support was found for interactions between neighborhood perceptions and psychosocial cognitions for leisure-time walking.

A PDF file should load here. If you do not see its contents the file may be temporarily unavailable at the journal website or you do not have a PDF plug-in installed and enabled in your browser.

Alternatively, you can download the file locally and open with any standalone PDF reader:

https://academic.oup.com/her/article-pdf/28/2/220/1614863/cyt002.pdf

Why some walk and others don't: exploring interactions of perceived safety and social neighborhood factors with psychosocial cognitions

Why some walk and others don't: exploring interactions of perceived safety and social neighborhood factors with psychosocial cognitions Marie¨lle A. Beenackers 0 1 Carlijn B. M. Kamphuis 0 1 Johan P. Mackenbach 0 1 Alex Burdorf 0 1 Frank J. van Lenthe 0 1 0 The Author 2013. Published by Oxford University Press. All rights reserved. Downloaded from https://academic.oup.com/her/article-abstract/28/2/220/595870 by guest on 11 1 Department of Public Health, Erasmus University Medical Center , Rotterdam , The Netherlands Although physical activity is often believed to be influenced by both environmental and individual factors, little is known about their interaction. This study explores interactions of perceived safety and social neighborhood factors with psychosocial cognitions for leisure-time walking. Cross-sectional data were obtained from residents (age 25-75 years) of 212 neighborhoods in the South-East of the Netherlands, who participated in the Dutch GLOBE study in 2004 (N ¼ 4395, survey response 64.4%). Direct associations of, and interactions between perceived neighborhood safety, social neighborhood factors (social cohesion, social network and feeling at home) and psychosocial cognitions (attitude, self-efficacy, social influence and intention) on two outcomes of leisure-time walking [yes versus no (binary), and among walkers: minutes per week (continuous)] were analyzed in multilevel regression models. The association between attitude and participating in leisure-time walking was stronger in those who felt less at home in their neighborhood. Social influence and attitude were stronger associated with participation in leisure-time walking in those who sometimes felt unsafe in their neighborhood. A positive intention was associated with more minutes walked in those who perceived their neighborhood as unsafe among those who walked. Only limited support was found for interactions between neighborhood perceptions and psychosocial cognitions for leisure-time walking. Introduction Physical inactivity is among the most important and prevalent risk factors of many major diseases [ 1–4 ]. Understanding why people are physically inactive is therefore of key importance in developing strategies to reduce these major diseases. Walking is a relatively easy way to be physically active; it is accessible to most people because it does not require any financial means and it can be continued into old age. Known determinants of walking are individual psychosocial cognitions, such as attitude and self-efficacy [ 5, 6 ]. In the past decade, many studies also investigated possible environmental determinants of walking, such as safety, population density and access to facilities [ 7–11 ]. Thus far, many studies have looked at the relation between psychosocial cognitions and environmental factors with walking separately or have explored to what extent psychosocial cognitions mediated the influence of environmental factors on walking [ 12–16 ]. However, a social-ecological perspective suggests that there is an interplay between the individual and the environment. According to Emmons [ 17 ], improving the understanding of health behaviors in their social context implies that the role of individual factors for health behaviors may depend on the environmental context. One of the core principles of ecological models is that influences interact across levels [ 18 ]. So, although such interactions are implied in ecological models [ 18–20 ], these models do not provide specific hypotheses, and perhaps as a consequence, empirical studies into interaction effects are still scarce. Few studies that did investigate environment–individual interactions for walking have mainly focused on built environmental factors including connectivity of streets, availability of shopping and sports facilities and neighborhood aesthetics [ 21–23 ]. Other factors, such as social environmental factors [ 24–27 ] and safety concerns [ 28, 29 ], are also suggested to be of importance for walking behavior. Rhodes et al. [ 22 ] studied the interactions between safety and psychosocial cognitions with respect to walking behavior and found that low levels of perceived crime resulted in a larger influence of attitude on the intention to walk compared with people who perceived high levels of crime. To date, there are no studies that have looked at interactions of psychosocial cognitions with social neighborhood factors such as social cohesion and social network for walking. Therefore, it is the aim of this article to explore interactions of safety and social neighborhood perceptions (neighborhood social cohesion, neighborhood social network and feeling at home within your neighborhood) with psychosocial cognitions (attitude, self-efficacy, intention and social influence) for leisure-time walking. In general, two possible interaction mechanisms can be at play. The first mechanism proposes that the environment is less important for the decision to walk for those who have more positive psychosocial cognitions toward physical activity. When this interaction exists, people with less positive psychosocial cognitions would benefit more from a supportive environment. The other mechanism assumes a synergy between environmental factors and psychosocial cognitions; the environment is more important in the decision to walk for people with more positive cognitions. This means that the beneficial effects of having positive psychosocial cognitions and living in a stimulating environment on walking would strengthen each other. For example, among those who report to have a small social network in their neighborhood, one may expect that having a positive intention toward physical activity results in less walking than among those with a large social network, as having a smaller social network may be a barrier to putting one’s positive intentions into action. The aim of this article is to investigate interactions of perceived safety and social neighborhood perceptions (neighborhood social cohesion, neighborhood social network and feeling at home within your neighborhood) with psychosocial cognitions (attitude, self-efficacy, intention and social influence) for two outcomes of leisure-time walking; any versus no leisure-time walking and among walkers: minutes per week spent on leisure-time walking. Materials and methods Data collection Data for this study were collected among a stratified sample of the adult population of the city of Eindhoven and its surrounding municipalities in the Netherlands in 2004, as part of the Dutch GLOBE study. The baseline sample was stratified by age, degree of urbanization and socioeconomic status (SES). More detailed information on the objectives, study design and data collection of the Dutch GLOBE study can be found elsewhere [ 30, 31 ]. In short, the study started with a baseline survey in 1991. This baseline sample was stratified by age, degree of urbanization and SES. In 2004, a new subsample was added to the original cohort to restore population representativeness of the study sample. In this study, questionnaires from the cross-sectional sample of the fourth wave (October 2004) were used (N ¼ 4785, response 64.4%). The fourth wave was chosen because of its particular focus on neighborhood factors. The use of personal data in the GLOBE study is in compliance with the Dutch Personal Data Protection Act and the Municipal Database Act and has been registered with the Dutch Data Protection Authority (number 1248943). Respondents with a missing outcome (n ¼ 182) or who had more than 25% missing values on the variables used in the analyses (n ¼ 149) were omitted from the analyses. Respondents with a missing neighborhood identifier (n ¼ 59) were also excluded. Thus, a total of 4395 respondents were included. Remaining missing values were imputed (see Statistical Analyses section). The respondents resided in 212 administrative neighborhoods of Eindhoven and its surrounding municipalities (mean number of respondents per neighborhood n ¼ 21, interquartile range ¼ 6–27). Measures Leisure-time walking Leisure-time walking was assessed by the SQUASH, a validated Dutch questionnaire that measures different types of physical activity [ 32 ]. Within SQUASH, leisure-time walking (i.e. walking for recreational purposes, no transportation walking) was measured by asking the respondent how many days they walked during leisure-time in a usual week (frequency) and how much time they spend on this on those days (duration). Because many respondents did not walk at all during leisure time, the first outcome variable we analyzed was binary, namely any versus no leisure-time walking (‘yes, does walk during leisure-time’ versus ‘no, does not walk during leisure-time’). For those who indicated to do any leisure-time walking, total minutes of leisure-time walking per week were calculated using information on frequency and duration. Psychosocial cognitions Psychosocial cognitions were based on the Theory of Planned Behavior [ 33 ] and the Social Cognitive Theory [ 34 ]. All items were formulated toward ‘sufficient physical activity in line with recommended levels’ [ 35 ]. Attitude was measured with 11 items (Cronbach’s alpha ¼ 0.79) with a 5-point ordinal answering scale (1, very important to 5, not important at all). An example question was whether respondents found the argument ‘it takes too much time’ important in their decision to be sufficiently active. Self-efficacy was measured with two items (Cronbach’s alpha ¼ 0.77). The first item asked whether respondents thought it was easy or difficult 222 to be sufficiently physically active (1, very difficult to 5, very easy). The second item asked how sure they could be sufficiently physically active when they would want to (1, not sure at all to 5, very sure). Intention was measured with one item (‘do you plan to be sufficiently physically active?’; 1, no, not sure at all to 5, yes, for sure). Social influence was measured with three items (Cronbach’s alpha ¼ 0.73) that addressed whether persons important to the respondent would (i) think the respondent should be sufficiently active, (ii) stimulate the respondent to be physically active and (iii) are sufficiently active themselves. Answering categories ranged from 1, ‘not true’ to 3, ‘yes, true’. For all psychosocial cognitions (except intention), a mean score was calculated from the relevant items within each cognition. A higher score on each scale represented a more positive cognition. All items used to construct the scales can be found in Supplementary Table S1. Neighborhood perceptions—social Elements of the neighborhood social environment were measured using a 13-item scale (Cronbach’s alpha ¼ 0.87). All items were measured on a 5-point ordinal scale (1, totally disagree to 5, totally agree). A principal component analyses with Varimax rotation and Kaiser Normalization distinguished three factors. The first factor was labeled ‘social cohesion’, defined as ‘the extend of connectedness and solidarity among groups in society’ [ 36 ]. An item that had a high factor loading on this factor was ‘most people in this neighborhood can be trusted’. The second factor was labeled ‘social network’, defined as ‘the presence and nature of interpersonal relationships and interactions; extend to which one is interconnected and embedded in a community’ [ 27 ]. An item stat had a high factor loading on this factor was ‘I often visit my neighbors in their home’. The third factor was labeled ‘feeling at home in this neighborhood’. An item that had a high factor loading on this factor was ‘I move out of this neighborhood if I get the chance (recoded)’. For all three factors, a standardized factor score (mean ¼ 0, standard deviation of 1) was constructed using the factor loadings. The individual social neighborhood items, their means and standard deviations and the factor loadings can be found in Supplementary Table S2. Neighborhood perceptions—safety Perceived safety of the neighborhood was assessed with four items. The first three items assessed people’s fear of being home alone or of going out on the streets in their neighborhood in the daytime or at night. The items were dichotomized into ‘no, never feeling afraid’ (0) and ‘neutral/yes, sometimes feeling afraid’ (1). The fourth item asked the respondents whether they thought their neighborhood was unsafe (no ¼ 0, yes ¼ 1). These four dichotomous items were summed up to form a scale (Cronbach’s alpha ¼ 0.68). Respondents who did not agree with any of the items indicating an unsafe neighborhood were regarded as ‘high’ on perceived neighborhood safety; they felt safe. Respondents who agreed once or twice to a measure indicating an unsafe neighborhood were considered ‘medium’ on perceived neighborhood safety; they sometimes felt unsafe. Respondents who agreed to three or four of the items indicative of an unsafe surrounding were considered ‘low’ on perceived neighborhood safety; they often felt unsafe. Demographics Potential confounders included were gender, age, country of origin (the Netherlands and other country) and educational level ((i) no education or primary education; (ii) lower professional and intermediate general education; (iii) intermediate professional and higher general education; (iv) higher professional education and university or (v) missing). Educational level was included as an indicator for SES and has proven to be a good measure for SES in the Netherlands [ 37 ]. Statistical analyses Overall, missing values of questionnaire items varied from <1 to 3% per item, with only intention having 7% missing values. Because complete case analyses would result in a loss of 25% of the respondents, missing values for the predictors were imputed using the Expectation Maximization method [ 38 ] from PASW version 18.0. All the variables described in the method (psychosocial cognitions, neighborhood perceptions, demographics and leisure-time walking) were used in the imputation model. Weighted multilevel logistic regression (for participation in leisure-time walking) and linear regression (for total minutes walked in a usual week, within those who walked) models were used to explore the associations between the predictors and leisure-time walking of respondents (Level 1) nested within neighborhoods (Level 2). Associations among all neighbourhood predictors and between the neighbourhood predictors and the psychosocial cognitions were at best modest (correlation coefficients <0.3). Associations between the psychosocial cognitions were as expected somewhat higher (correlation coefficients 0.1; 0.5). Although multicollinearity is not expected to be a problem because of these modest correlations, all continuous variables were mean centered to prevent multicollinearity in the interaction models and to ease interpretation. All models were weighted (Level 1 weight) to reflect the source population in terms of gender, age and educational level. Model 1 contained all neighborhood perceptions. Model 2 contained all psychosocial cognitions. Model 3 combined neighborhood perceptions with psychosocial cognitions. Subsequently, interactions were explored whereby each neighborhood–individual interaction term was added separately to Model 3 (Model 4a–p). Interactions in a logistic regression model are tested for their departure from multiplicativity (the combined ‘effect’ of the two factors is larger or smaller that the product of the individual ‘effects’). Interactions in a linear regression model are tested for their departure from additivity (the combined ‘effect’ of the two factors is larger or smaller that the sum of the individual ‘effects’). Because additive interactions are considered more intuitive and more relevant to public health [ 39 ], and to increase comparability of the results for the two outcomes, the relative risk due to interaction (RERI), a measure to quantify interaction on an additive scale, was also calculated for all interactions departing from multiplicativity [ 40, 41 ]. The RERI is a measure of interaction between two parameters with a value further away from zero indicating a stronger interaction. The tool created by Knol and coworkers [ 40, 41 ] was used to calculate the RERI and the accompanying 95% confidence interval (CI). All multivariable models were adjusted for age, gender, educational level and country of origin. Significance was interpreted by using the 95% CI. All regression analyses were carried out in STATA 12 using GLLAMM [ 42 ] for the logistic regression analysis to study participation in leisure-time walking and using XTMIXED to study the amount of leisure-time walking within those who walked. Significant interactions have been visualized by simple slope analyses. Results Table I shows the characteristics of the sample. Approximately one-third (32.7%) of the respondents reported no leisure-time walking at all. Those who did walk spent on average 212 min per week on leisure-time walking. Crude analyses as presented in Table II show that females, higher educated and older respondents were more likely to participate in leisure-time walking. Among the walkers, minutes spent per week on leisure-time walking increased with age, but decreased with educational level. Crude analyses also showed that a positive attitude [odds ratio (OR) 1.67, 95% CI 1.42–1.95], a strong self-efficacy (OR 1.20, 95% CI 1.11–1.29), a positive social influence (OR 1.39, 95% CI 1.22– 1.57) and a strong intention toward physical activity (OR 1.37, 95% CI 1.27–1.47) were positively associated with participating in leisure-time walking (Table II). Those with a larger social network in the neighborhood (OR 1.16, 95% CI 1.08–1.24) were also more likely to walk in leisure time. A positive attitude (b 33.77, 95% CI 14.19–53.34), strong self-efficacy (b 39.12, 95% CI 29.05–49.20) and a positive intention toward physical activity (b 16.06, 95% CI 7.40–24.73) were also associated 224 with more walking in those who walked during leisure time (Table II). None of the neighbourhood perceptions were significantly associated with minutes walked. Adjusted for potential demographic confounders and the other neighborhood perceptions, individuals with a larger social network (OR 1.14, 95% CI 1.07–1.22) were more likely to engage in walking in leisure time (Model 1, Table III). The association remained significant after additional adjustment for the psychosocial cognitions (Model 3, Table III). Of the psychosocial cognitions, all but self-efficacy remained a significant predictor of leisure-time walking after adjusting for the potential demographic confounders and the other psychosocial cognitions (Model 2, Table III). After additional adjustment for the neighborhood perceptions, the associations between social influence and leisure-time walking were no longer significant [although there was only little change in the point estimate (Model 3, Table III)], whereas attitude and intention remained significant. In those who walked during leisure time, a strong self-efficacy was associated with longer total duration of walking during leisure time, also in the fully adjusted model (b 38.31, 95% CI 27.37– 49.25) (Model 3, Table IV). In Model 3, there was also a significant inverse association between perceived social cohesion in the neighbourhood and minutes walked (b 11.69, 95%CI 21.00 to 2.38]) (Model 3, Table IV). Interactions Additional inclusion of the interaction terms resulted in three significant interactions for participation in leisure-time walking in the regression models. The calculated RERIs basically followed the results of the multiplicative interactions. Safety interacted significantly with both attitude and social influence. The association between attitude and participation in leisure-time walking in people who sometimes felt unsafe was 1.59 times as high compared with those who never felt unsafe (95% CI 1.10–2.31) (as visualized in Fig. 1). This pattern was not observed for those who often felt unsafe aThe numbers and percentages presented are unweighted and are therefore a representation of the actual numbers in the dataset. bPercentages are presented, unless otherwise stated. cSocial neighbourhood factors (‘social cohesion’, ‘social network’ and ‘feeling at home’) were not included in this table because they were standardized factor scores (mean ¼ 0, standard deviation of 1). The mean and standard deviations for the individual items that were used to construct the factor scores can be found in Supplementary Table S2. in their neighborhood (OR 1.04, 95% CI 0.59– 1.83]). The interaction between social influence and safety was similar in such a way that the association between social influence and participation in leisure-time walking was 1.36 as high in those who felt sometimes unsafe compared with those who never felt unsafe in their neighborhood (as visualized in Fig. 2). This pattern was not observed for those who often felt unsafe (OR 0.85, 95% CI 0.53– 1.35). The third interaction was between feeling at home in your neighborhood and attitude (OR 0.87, 95% CI 0.75–1.00); among those feeling more at home, attitude had a weaker association with participation in leisure-time walking, than among those feeling less at home in their neighborhood (visualized in Fig. 3). Among those who walked during leisure time, one significant interaction was observed for total minutes walked per week. In those who felt unsafe, a positive intention was associated with over 30 min more walking during leisure time compared with those who did not feel unsafe and had a positive intention toward physical activity. This interaction has been visualized in Fig. 4. Discussion This study is among the first to evaluate interactions between elements of the social environmental and safety in neighborhoods and psychosocial cognitions toward leisure-time walking. Several interactions were found but no clear pattern could be detected. Our finding of an association among attitudes, self-efficacy, social influence, intention and leisure-time walking is in line with both theory and previous empirical research [ 5, 6 ]. Interestingly, a positive social influence was associated with participating in leisure-time walking but not with minutes walked. Also, our finding that a large social network was positively related to participating in leisure-time walking has been found in previous studies [ 24–27 ]. The negative association between neighborhood social cohesion and minutes walked among the walkers was unexpectedly and without a plausible explanation. This study extends on previous research by exploring environment–individual interactions. Three interactions were found with perceived neighborhood safety. For participation in leisure-time 225 aBold figures indicate statistical significance (P < 0.05), *P < 0.05, **P < 0.01, ***P < 0.001. walking, perceived neighborhood safety interacted with attitude and social influence: in those who sometimes felt unsafe, a positive attitude and a positive social influence were significantly stronger associated with any leisure-time walking. This pattern was not observed for those who often felt unsafe. This finding was different from findings by Rhodes et al. [ 22 ] who found that low levels of perceived crime resulted in a larger influence of attitude on the intention to walk compared with people who perceived high levels of crime. For our second outcome, minutes walked among those persons who engaged in leisure-time walking, also an interaction with safety was found: those who perceived feelings of unsafety but had a positive intention to walk in leisure time walked 30 min per week more than persons who felt safe in their neighborhood and 226 persons who lacked intention to walk. Although these unexpected interactions with safety are hard to interpret, a possible explanation may be found in the association between safety and walking itself. Although we were primarily interested in the influence of neighborhood safety on leisure-time walking, the cross-sectional nature of this study cannot preclude the direction of association. Therefore, it is possible that those who walk in their neighborhood are more likely to report feelings of unsafety because they are more exposed to their neighborhood. This inverse association between neighborhood safety and physical activity has been observed before [ 43, 44 ]. The final interaction observed was between feeling at home and attitude, whereby feeling at home in your neighborhood was stronger associated with * 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 fi o n i I R E R r o f s l e a r e t n i e l b a i r a v i t l u ,1 ly e e s c .0 n o n 0 o h e t r e < re n f a e h r * s t 5 an re n 0 c f n i < se d io t i y n . P* teh l b g s a o r ,) ,r to c to 5 e n t c .0 v ah fa 0 e e t w b e . w R ic l m n M io to s t ig f c e a b w n i s ch in d e e e a n v b t e a o l i o ld c d ith it v u i n n o in w g i s re tl e ti o h y c n s o s u h h i m g t n r g i g e fi ls ro o t f c n ld ry s e io o a R th t B v c O r a o r . an e f t e in c th s in ig ) , R r s ) e c d O h o li – e t f a a o it (4 th fo , y n y l s r i t e r t ( e d te n f u 3 a o e o s m m c l a e d e r nd od oo th a p r R n h o e fo O it 227 * 8 1 4 3 2 5 5 5 .4 .3 .2 .6 6 .0 .2 11 .48 9 2 1 .7 7 9 . 1 7 1 1 2 4 0 1 3 7 1 2 8 4 8 5 4 2 1 3 1 0 5 9 5 2 2 2 2 6 4 5 3 1 9 0 8 2 4 2 s i C C l e v e l i t l u m e l b a 3 7 1 2 8 4 8 5 4 2 1 3 1 tirrsceod irfaebhohogdo irfsceeeaevdP ifseadumM fsaenU iliscaceoonohS iltracenookSw liteeeanghoFm iilsscccaohoyogncittt)(eud5–1A lf-f(eeccay–1fiS ilicaeeconnuSflctitI()enon–51n l:ea4dop–odM .tttfsaaea4y+ ifseaedumM fsaenU .tfssaeeb4y+ ifseaedumM fsaenU .tfssace4yo+ ifseaedumM fsaenU .tifsaed4yn+ ifseaedumM fsaenU .isceeo4onh+ .ifsceo4onh+ .iscegoh4on+ .iscehoh4on+ .itren4ko+w .jtren4ko+w .trekn4ko+w .ltren4ko+w .lfee4ohmm+ .lfeen4ho+m .lfeeo4ho+m .lfeep4ho+m alllseodAwm <P..**0100* .illiaynnomm trrsseeeepohn trfreeeceehn P N P M 228 engaging in leisure-time walking in those with a below average attitude. This interaction could indicate the existence of the first mechanism proposed in the Introduction: those who have negative psychosocial cognitions benefit more from a positive neighborhood environment than those with more positive psychosocial cognitions toward physical activity. Or, stated the other way around, not feeling at home in your neighborhood may not be a barrier for walking among those with a positive attitude toward physical activity, as this positive attitude makes them more likely to be active anyway. Overall, we found limited empirical support for interactions, and neither of the proposed mechanisms was clearly favored in our results although the interaction between feeling at home and attitude hints at the first mechanism in which those with negative psychosocial cognitions benefit most from a positive neighborhood environment. The recent study by Carlson et al. also found a limited number of interactions [ 21 ]. In their article, they studied the interactions among walkability, parks and recreation facilities, aesthetics and walking facilities within the neighborhood with social support, self-efficacy and perceived barriers on leisuretime walking. They found one significant interaction between walking facilities and self-efficacy in which self-efficacy was only associated with leisure-time walking in neighborhoods with few walking facilities. This interaction also supports the first proposed mechanism in which positive psychosocial cognitions can help to overcome neighborhood barriers. Although methodological reasons, including lack of statistical power and measurement error in environmental and (to a lesser extent) individual factors, may have contributed to this finding, it is also possible that walking behavior mainly is a result of a combination of environmental and individual factors, in which only few interactions are involved, which have little implications for public health practice. However, the strong theoretical support for environment–individual interactions in ecological models prompts for more research that indentifies and quantifies these interactions. Study limitations and strengths Several limitations need to be considered in the interpretation of the findings of this study. First, the cross-sectional design restricts interpretation on causality and direction of the associations. This is particularly relevant because of the increasing recognition of a dynamic interrelation in which individuals change places and places change people [ 45 ]. Second, our psychosocial cognitions were measured with regard to ‘sufficient physical activity in line with recommended levels’ where it would have better preferred to ask this specifically for leisure-time walking. This may have resulted in an underestimation of associations with leisure-time walking. Third, self-reported physical activity data are known for overestimations. In addition, the SQUASH questionnaire was validated for total physical activity but not for the underlying specific activities such as leisure-time walking. Because this study used a robust dichotomous measure, it is expected to be of little influence although we can not exclude some bias in the associations. Finally, the results of this study should be interpreted in the context of a medium-sized city in the Netherlands. The situation in Dutch urban areas may not be representative for other urban areas in the world. Conclusion This study explored interactions between neighborhood factors and psychosocial cognitions for explaining leisure-time walking in adults and found limited evidence for these interactions. The relationship between neighborhood and individual determinants of walking and environment–individual interactions remains complex, and more studies are needed that incorporate these interactions to strengthen these results. Supplementary data Supplementary data are available at HEALED online. Acknowledgements The GLOBE study is carried out by the Department of Public Health of the Erasmus University Medical Centre in Rotterdam, in collaboration with the Public Health Services of the city of Eindhoven and region South-East Brabant. 231 Funding The Netherlands organization for health research and development (ZonMw; 122000003). Conflict of interest statement None declared. 232 1. Centers for Disease Control and Prevention . 1996 . Physical Activity and Health: A Report of the Surgeon General . Washington DC: U.S. Department of Health and Human services, 1996 . 2. Centers for Disease Control, National Center for Health Statistics, Division of Health Promotion Statistics. 2008. Data 2010 . . . The Healthy People 2010 Database - Focus area: 22 - Physical Activity and Fitness. Available at: http:// www.healthypeople.gov/Data/data2010.htm. Accessed: 29 April 2011 . 3. Warburton DE , Nicol CW , Bredin SS . Health benefits of physical activity: the evidence . Can Med Assoc J 2006 ; 174 : 801 - 9 . 4. Bauman AE . Updating the evidence that physical activity is good for health: an epidemiological review 2000-2003 . J Sci Med Sport 2004 ; 7 : 6 - 19 . 5. Trost SG , Owen N , Bauman AE et al. Correlates of adults' participation in physical activity: review and update . Med Sci Sports Exerc 2002 ; 34 : 1996 - 2001 . 6. Sallis JF , Owen N. Physical Activity and Behavioral Medicine . Thousand Oaks , CA: Sage Publications, 1999 . 7. Owen N , Humpel N , Leslie E et al. Understanding environmental influences on walking: review and research agenda . Am J Prev Med 2004 ; 27 : 67 - 76 . 8. Saelens BE , Sallis JF , Frank LD . Environmental correlates of walking and cycling: findings from the transportation, urban design, and planning literatures . Ann Behav Med 2003 ; 25 : 80 - 91 . 9. Saelens BE , Handy SL . Built environment correlates of walking: a review . Med Sci Sports Exerc 2008 ; 40 : S550 - 66 . 10. McCormack G , Giles-Corti B , Lange A et al. An update of recent evidence of the relationship between objective and self-report measures of the physical environment and physical activity behaviours . J Sci Med Sport 2004 ; 7 : 81 - 92 . 11. Van Holle V , Deforche B , Van Cauwenberg J et al. Relationship between the physical environment and different domains of physical activity in European adults: a systematic review . BMC Public Health 2012 ; 12 : 807 . 12. Ball K , Timperio A , Salmon J et al. Personal, social and environmental determinants of educational inequalities in walking: a multilevel study . J Epidemiol Community Health 2007 ; 61 : 108 - 14 . 13. De Bourdeaudhuij I , Teixeira PJ , Cardon G et al. Environmental and psychosocial correlates of physical activity in Portuguese and Belgian adults . Public Health Nutr 2005 ; 8 : 886 - 95 . 14. Giles-Corti B , Donovan RJ . Relative influences of individual, social environmental, and physical environmental correlates of walking . Am J Public Health 2003 ; 93 : 1583 - 9 . 15. Granner ML , Sharpe PA , Hutto B et al. Perceived individual, social, and environmental factors for physical activity and walking . J Phys Act Health 2007 ; 4 : 278 - 93 . 16. Kamphuis CB , van Lenthe FJ , Giskes K et al. Socioeconomic differences in lack of recreational walking among older adults: the role of neighbourhood and individual factors . Int J Behav Nutr Phys Act 2009 ; 6 : 1 . 17. Emmons KM . Health behaviors in a social context . In: Berkman LF , Kawachi I (eds). Social Epidemiology . New York: Oxford University Press, 2000 . 18. Sallis JF , Owen N , Fisher EB . Ecological models of health behavior . In: Glanz K , Rimer BK , Viswanath K (eds). Health Behavior and Health Educaton Theory, Research, and Practice . San Fransisco, CA: Jossey-Bass , 2008 , 465 - 86 . 19. Kremers S , de Bruijn G-J , Visscher T et al. Environmental influences on energy balance-related behaviors: a dual-process view . Int J Behav Nutr Phys Act 2006 ; 3 : 9 . 20. Bauman AE , Sallis JF , Dzewaltowski DA et al. Toward a better understanding of the influences on physical activity: the role of determinants, correlates, causal variables, mediators, moderators, and confounders . Am J Prev Med 2002 ; 23 : 5 - 14 . 21. Carlson JA , Sallis JF , Conway TL et al. Interactions between psychosocial and built environment factors in explaining older adults' physical activity . Prev Med 2012 ; 54 : 68 - 73 . 22. Rhodes RE , Brown SG , McIntyre CA . Integrating the perceived neighborhood environment and the theory of planned behavior when predicting walking in a Canadian adult sample . Am J Health Promot 2006 ; 21 : 110 - 8 . 23. Cerin E , Vandelanotte C , Leslie E et al. Recreational facilities and leisure-time physical activity: an analysis of moderators and self-efficacy as a mediator . Health Psychol 2008 ; 27 : 26 - 35 . 24. Addy CL , Wilson DK , Kirtland KA et al. Associations of perceived social and physical environmental supports with physical activity and walking behavior . Am J Public Health 2004 ; 94 : 440 - 3 . 25. Lindstro¨m M , Hanson BS , O¨ stergren PO. Socioeconomic differences in leisure-time physical activity: the role of social participation and social capital in shaping health related behaviour . Soc Sci Med 2001 ; 52 : 441 - 51 . 26. Cleland VJ , Timperio A , Crawford D. Are perceptions of the physical and social environment associated with mothers' walking for leisure and for transport? A longitudinal study . Prev Med 2008 ; 47 : 188 - 93 . 27. McNeill LH , Kreuter MW , Subramanian SV . Social environment and physical activity: a review of concepts and evidence . Soc Sci Med 2006 ; 63 : 1011 - 22 . 28. Pikora T , Giles-Corti B , Bull F et al. Developing a framework for assessment of the environmental determinants of walking and cycling . Soc Sci Med 2003 ; 56 : 1693 - 703 . 29. Loukaitou-Sideris A , Eck JE . Crime prevention and active living . Am J Health Promot 2007 ; 21 : 380 - 9 , iii. 30. van Lenthe FJ , Schrijvers CTM , Droomers M et al. Investigating explanations of socio-economic inequalities in health: the Dutch GLOBE study . Eur J Public Health 2004 ; 14 : 63 - 70 . 31. Mackenbach JP , van de Mheen H , Stronks K. A prospective cohort study investigating the explanation of socio-economic inequalities in health in the Netherlands . Soc Sci Med 1994 ; 38 : 299 - 308 . 32. Wendel-Vos GCW , Schuit AJ , Saris WHM et al. Reproducibility and relative validity of the short questionnaire to assess health-enhancing physical activity . J Clin Epidemiol 2003 ; 56 : 1163 - 9 . 33. Ajzen I. The theory of planned behavior . Organ Behav Hum Decis Process 1991 ; 50 : 179 - 211 . 34. Bandura A. Social Foundations of Thought and Action: A Social Cognitive Theory . Englewood Cliffs, NJ: Prentice Hall, 1986 . 35. Kemper HCG , Ooijendijk WTM , Stiggelbout M. Consensus about the Dutch recommendation for physical activity to promote health [Consensus over de Nederlandse Norm voor Gezond Bewegen] . Tijdschrift voor Gezondheidswetenschappen 2000 ; 78 : 180 - 3 . 36. Kawachi I , Berkman LF . Social cohesion, social capital, and health . In: Berkman LF , Kawachi I (eds). Social Epidemiology . New York: Oxford University Press, 2000 . 37. Van Berkel-Van Schaik AB , Tax B . Towards a Standard Operationalisation of Socioeconomic Status for Epidemiological and Socio-medical Research [Naar een standaardoperationalisatie van sociaal-economische status voor epidemiologisch en sociaal-medisch onderzoek] . Rijswijk: Ministerie van Welzijn, Volksgezondheid en Cultuur , 1990 , 129 . 38. Dempster AP , Laird NM , Rubin DB . Maximum likelihood from incomplete data via the EM algorithm . J Roy Stat Soc B Met 1977 ; 39 : 1 - 38 . 39. Rothman KJ , Greenland S , Lash TL. Modern Epidemiology . Philadelphia. PA: Lippincott Williams & Wilkins, 2008 . 40. Hosmer DW , Lemeshow S. Confidence interval estimation of interaction . Epidemiology 1992 ; 3 : 452 - 6 . 41. Knol MJ , van der Tweel I , Grobbee DE et al. Estimating interaction on an additive scale between continuous determinants in a logistic regression model . Int J Epidemiol 2007 ; 36 : 1111 - 8 . 42. Rabe-Hesketh S , Skrondal A , Pickles A . Maximum likelihood estimation of limited and discrete dependent variable models with nested random effects . J Econom 2005 ; 128 : 301 - 23 . 43. Gauvin L , Riva M , Barnett T et al. Association between neighborhood active living potential and walking . Am J Epidemiol 2008 ; 167 : 944 - 53 . 44. Humpel N , Owen N , Iverson D et al. Perceived environment attributes, residential location, and walking for particular purposes . Am J Prev Med 2004 ; 26 : 119 - 25 . 45. Diez Roux AV . Complex systems thinking and current impasses in health disparities research . Am J Public Health 2011 ; 101 : 1627 - 34 .


This is a preview of a remote PDF: https://academic.oup.com/her/article-pdf/28/2/220/1614863/cyt002.pdf

Beenackers, Mariëlle A., Kamphuis, Carlijn B. M., Mackenbach, Johan P., Burdorf, Alex, van Lenthe, Frank J.. Why some walk and others don't: exploring interactions of perceived safety and social neighborhood factors with psychosocial cognitions, Health Education Research, 2013, 220-233, DOI: 10.1093/her/cyt002