Preventive Medicine 76 (2015) 31–36

Contents lists available at ScienceDirect

Preventive Medicine journal homepage: www.elsevier.com/locate/ypmed

Neighborhood, family and individual characteristics related to adolescent park-based physical activity Susan H. Babey a,b,⁎, Diane Tan b, Joelle Wolstein a,b, Allison L. Diamant a,c a b c

UCLA Center for Health Policy Research, 10960 Wilshire Blvd, Suite 1550, Los Angeles, CA 90024, USA Department of Health Policy and Management, UCLA Fielding School of Public Health, 650 Charles Young Dr. S., 31-269 CHS Box 951772, Los Angeles, CA 90095-1772, USA Division of General Internal Medicine and Health Services Research, David Geffen School of Medicine at UCLA, 911 Broxton Plaza, Los Angeles, CA 90024, USA

a r t i c l e

i n f o

Available online 11 April 2015 Keywords: Physical activity Parks Adolescent Adolescent health Youth Park-based physical activity

a b s t r a c t Background. Local parks are an important community resource that may influence levels of physical activity among youth. However, few population-based studies have investigated park-based physical activity among youth. Purpose. This study examines sociodemographic, family, and neighborhood characteristics associated with park-based physical activity among adolescents. Methods. Data were from the 2007 California Health Interview Survey (CHIS), a population-based survey of California households, and were analyzed in 2012–2013 and 2015. This study examined individual (age, gender, race/ethnicity, and nativity), family (household income, parental education, and nativity), and neighborhood characteristics (urbanicity, perceived park availability, perceived park safety, neighborhood income, and racial composition) associated with engaging in park-based physical activity among adolescents. Results. In California, 71% of adolescents reported being physically active the last time they visited a park. In adjusted multinomial logistic regression analyses, older adolescents and females were less likely to be physically active in parks. Adolescents with a park within walking distance of home and those with a safe park nearby were more likely to be physically active during a park visit. Conclusions. Park availability and perceptions of park safety are important predictors of engaging in parkbased physical activity among adolescents. These findings provide information that can help inform interventions intended to increase physical activity among youth. Strategies that increase availability and safety of parks and other recreation spaces may be particularly effective. © 2015 Elsevier Inc. All rights reserved.

Background Obesity is a major public health concern, especially among adolescents in the US. Obesity during early childhood and adolescence has been associated with increased risk of obesity in adulthood (Berenson and Srnivasan, 2005; Kvaavik et al., 2003; Singh et al., 2008). Additionally, obesity is a risk factor for many chronic diseases, including heart disease and type 2 diabetes (Eckel and Krauss, 1998; Wannamethee and Shaper, 1999). Insufficient physical activity contributes to obesity among adolescents, and regular physical activity has positive effects on musculoskeletal development, cardiovascular health and adiposity, and weight gain (Delva et al., 2007; Strong et al., 2005; US Department of Health and Human Services, 1996). National guidelines recommend that children and adolescents engage in at least 60 minutes of physical activity each day (U.S. Department of Health and Human Services, 2008). ⁎ Corresponding author at: UCLA Center for Health Policy Research, 10960 Wilshire Blvd, Suite 1550 Los Angeles, CA 90024, USA. Fax: +1 310 794 2686. E-mail address: [email protected] (S.H. Babey).

http://dx.doi.org/10.1016/j.ypmed.2015.04.001 0091-7435/© 2015 Elsevier Inc. All rights reserved.

Parks and other places to engage in physical activity provide an important resource that may influence levels of physical activity among adolescents (Babey et al., 2008; Cohen et al., 2006). Several studies have reported on barriers to and correlates of park use. In addition, a growing body of research has examined the association of access to parks and park characteristics with overall physical activity. However, relatively few studies have examined park-based physical activity among adolescents. Common barriers to park use include not having a park in the community, safety concerns, and poor maintenance of park facilities (Carlson et al., 2010; Mowen et al., 2005), while proximity to local parks has been associated with greater park use (Cohen et al., 2007). In addition, studies examining the association of access to parks with overall physical activity suggest that both proximity to parks and having more parks or greater park area are associated with greater levels of physical activity among youth (Brownson et al., 2001; Cohen et al., 2006; Frank et al., 2007; Kaczynski and Henderson, 2007; Roemmich et al., 2006). Research on the contribution of parks to physical activity also suggests that the availability and quality of park facilities, amenities, and programmed activities are positively associated with physical

32

S.H. Babey et al. / Preventive Medicine 76 (2015) 31–36

activity (Cohen et al., 2007, 2009, 2013; Kaczynski and Henderson, 2007; Kaczynski et al., 2008; Ries et al., 2009; Shores and West, 2008). Despite the growing number of studies examining the contribution of parks to physical activity, relatively few studies have examined the factors that influence physical activity that occurs in parks. Research in this area suggests that males, younger people, and whites are more active in parks (Floyd et al., 2008a; Kaczynski et al., 2011). Studies conducted in parks or in neighborhoods surrounding parks suggest that park characteristics such as the availability of facilities, staffing, and programming are positively related to park-based physical activity (Cohen et al., 2012; Kaczynski et al., 2010). Other research has suggested that neighborhood factors such as lower neighborhood income and higher concentrations of Hispanic or African American residents are related to greater park-based physical activity (Floyd et al., 2008a, 2008b). Among studies that have examined park-based physical activity, very few have focused on youth. The few studies that have focused on youth have found that boys are more active than girls, and white youth are more active than non-white youth in parks (Floyd et al., 2011; Kaczynski et al., 2013). In addition, previous research has found that park features (i.e., activity area type, and the number of recreation facilities, and formal activities available) were positively associated with physical activity in the park, while neighborhood characteristics (i.e., urban form and racial heterogeneity) were not (Floyd et al., 2011; Spengler et al., 2011). The existing research has not used population-based samples to examine correlates of park-based physical activity among adolescents. These studies consisted mainly of observational studies of park users, often relying on smaller local surveys. For example, one of the studies observed children and adolescents using sampled parks in Durham, NC, and the other observed users in selected parks in Kansas City, MO (Floyd et al., 2011; Kaczynski et al., 2013). In addition, previous studies have not examined park-based physical activity among adolescents using a wide range of individual and family characteristics (e.g., race/ ethnicity, household income, and parental education) as well as neighborhood characteristics (e.g., perceptions of park availability and safety and neighborhood racial composition). This study examines individual sociodemographic, family, and neighborhood characteristics associated with park-based physical activity among adolescents using populationbased data, which includes important factors not available in previous studies of park-based physical activity for this age group. Methods Conceptual framework The current research utilized a socio-ecological model to inform the analyses (Kaczynski et al., 2009; Sallis et al., 2006). This model suggests that health behaviors are influenced by a number of factors at multiple levels. These factors include individual and social as well as community and environmental factors. When applied to park-based physical activity, this model suggests that parkbased physical activity will be influenced by personal characteristics such as demographics, family characteristics, and perceived environment (e.g., safety and availability). This model also predicts that behavior settings including neighborhood characteristics and the recreation environment will influence park-based physical activity. Data source and population This research used data from the 2007 California Health Interview Survey (CHIS), a random-digit dial telephone survey of more than 43,000 households designed to be representative of California's noninstitutionalized population. Data were analyzed in 2012–2013 with additional analyses conducted in 2015. CHIS 2007 is the most recent CHIS data set that included questions about physical activity in parks as well as the covariates used in the current analysis. One randomly selected adult (aged 18 years or older) was interviewed in each household. In households with adolescents ages 12 to 17 years, one adolescent was randomly selected and interviewed directly after obtaining parental permission and consent from the adolescent. Older adolescents were slightly

less likely to complete an interview (California Health Interview Survey, 2009d). However, the CHIS survey weights account for demographic and geographic characteristics to reduce potential nonresponse bias. In addition, previous research suggests that the CHIS sample is representative of the California population (Lee et al., 2009). This research also found no evidence of nonresponse bias in the CHIS sample. A total of 3,638 adolescents completed the survey, representing a completion rate of 44.1%. Interviews were conducted in English, Spanish, Chinese, Vietnamese, and Korean. Detailed information about California Health Interview Survey methodology is available elsewhere (California Health Interview Survey, 2009a, 2009b, 2009c, 2009d, 2009e). Measures The outcome measure of interest was based on adolescent responses to two questions: “In the past 30 days, did you go to a park, playground, or open space?” and “The last time you went to a park were you physically active while you were there?” The second question was only asked if adolescents responded yes to the first. Responses to these questions were used to create a variable with three levels: active during last park visit, not active during last park visit, no recent park visit. These levels are analyzed as distinct categories because it is possible for adolescents to make the decision to go to a park separately from the decision to be active at the park. The following individual sociodemographic characteristics were included in these analyses: age, gender, race/ethnicity, and nativity. Family characteristics included parental education, parental nativity, and annual household income. Neighborhood characteristics included urbanicity, adolescent perception of the safety of his or her nearby park, adolescent perception of park availability near home, and neighborhood income and racial/ethnic composition. Adolescents reported their age, gender, nativity (coded as US-born or foreign-born), and race/ethnicity (white, Latino, Asian, African American, American Indian, Pacific Islander, or mixed race). In addition, adolescents reported whether there was a park, playground, or open space within walking distance of home and how strongly they agreed that the park closest to home was safe during the day (strongly agree, agree, disagree, or strongly disagree). The adult respondent reported household income, household address, parental educational attainment, and maternal and paternal nativity. Household income was examined as a percent of the federal poverty level (0–99%, 100– 199%, 200–299%, and 300%, and above). Maternal and paternal nativity were coded as US-born or foreign-born. Parental educational attainment for the responding adult was coded as high school or less, some college, college graduate. Household address was used to determine the census tract in which a family lived at the time of the survey. Data from the American Community Survey, 2005–2009, were linked to the CHIS data set by census tract to examine neighborhood income and neighborhood racial/ethnic composition. Census tracts in which 50% or more of the households were below 200% of the federal poverty level were considered lower income. Neighborhood racial/ethnic composition was examined as percent of the population in the census tract that was Hispanic or Latino (less than 25% versus 25% or more Hispanic or Latino). Using data obtained from CLARITAS, households were assigned to urbanicity levels (urban, suburban, rural) based on population density of the household's ZIP code and surrounding areas. Analyses Multinomial logistic regression analyses were used to examine the association of individual, family, and neighborhood characteristics with park-based physical activity, where being active during the last park visit and not being active during the last visit were each compared to not having a recent park visit. The model included age, gender, race/ethnicity, household income, urbanicity, respondent's nativity, maternal nativity, paternal nativity, responding parent's educational attainment, perceived park availability near home, perceived safety of nearby park, neighborhood income, and neighborhood racial/ethnic composition. Predictive margins were used to calculate adjusted percentages for variables that were significant predictors of park-based physical activity in the multinomial regression model (Graubard and Korn, 1999). Due to missing values for the neighborhood income variable, regression analyses included 3,636 adolescents. Data were analyzed with SAS and STATA. Analyses were weighted to be representative of the California population and adjusted for the complex survey design of the California Health Interview Survey. The University of California, Los Angeles (UCLA) Office for the Protection of Research Subjects certified this research exempt from review.

S.H. Babey et al. / Preventive Medicine 76 (2015) 31–36

Results Table 1 displays characteristics of the studied population, California adolescents ages 12–17 years. Seventy-one percent reported they were physically active during their last park visit, 6.4% were not active on their last park visit, and 22.3% had not visited a park in the past 30 days. In addition, 86% of adolescents reported having a park within walking distance of home, 35% strongly agreed, and 53% agreed that the park closest to home was safe during the day. In unadjusted analyses (Table 2), the percent of adolescents engaging in recent park-based physical activity was higher among those who reported having a park within walking distance than among those

Table 1 Characteristics of California adolescents (ages 12–17 years), 2007.

Physically active in park during last park visit Active Not active No visit Age (mean) Gender Female Male Race/ethnicity Latino White Asian African American American Indian Mixed race or other Nativity US-born Foreign-born Income Below 100% FPL 100–199% FPL 200–299% FPL 300% FPL and above Mother's nativity US-born Foreign-born Father's nativity US-born Foreign-born Parental educational attainment High school or less Some college College graduate Urbanity Urban Suburban Rural Park/playground within walking distance from home Yes No Nearby park/playground safe during day Strongly agree Agree Disagree Strongly disagree Neighborhood racial composition Less than 25% Latino/Hispanic 25% or more Latino/Hispanic Neighborhood income c Lower income Higher income a

Unweighted n (N = 3,638)a

Weighted %b

2,623 244 771 3638

71.3 6.4 22.3 14.6

1,823 1,815

51.1 48.9

746 1,890 286 156 43 517

29.3 36.5 9.8 7.5 0.7 16.3

3,277 361

87.5 12.5

451 586 487 2,114

19.7 19.0 13.4 47.9

2,471 1,167

59.3 40.7

2,425 1,213

56.0 44.0

1,300 840 1,498

43.1 23.8 33.0

2,123 808 707

66.1 21.8 12.1

3,038 600

86.6 13.4

1,376 1,949 251 62

35.4 53.5 8.8 2.3

2,188 1450

49.6 50.4

593 3043

22.6 77.4

Some sample sizes may not add to total due to missing values. Estimates are weighted to be representative of the California population and are adjusted for complex survey design effects. c Census tracts in which at least 50% of the households were below 200% of the federal poverty level were considered lower income. b

33

Table 2 Unadjusted percent of adolescents engaging in park-based physical activity, not active during last park visit, or with no park visit, California, 2007.a

Factor Individual characteristics Age (mean) Gender Male Female Race Latino American Indian or Alaskan native Asian African American White Mixed race or other Adolescent's nativity Foreign-born US-born Family characteristics Household income Below 100% FPL 100–199% FPL 200–299% FPL 300% FPL and above Mother's nativity Foreign-born US-born Father's nativity Foreign-born US-born Parental educational attainment HS or less Some college College graduate Neighborhood characteristics Urbanity Urban Suburban Rural Park/playground within walking distance from home Yes No Nearby park/playground safe during day Strongly agree Agree Disagree Strongly disagree Neighborhood racial composition Less than 25% Latino 25% or more Latino Neighborhood incomeb Lower income Higher income

Active last park visit

Not active

No visit

Weighted % (95% CI)

Weighted % (95% CI)

Weighted % (95% CI)

14.5 (14.4–14.5) 15.4 (15.2–15.6) 14.7 (14.6–14.9) 76.5 (73.6–79.4) 65.9 (62.8–69.1)

4.1 (2.8–5.3) 8.8 (6.9–10.6)

19.5 (16.8–22.2) 25.3 (22.5–28.2)

69.6 (65.1–74.1) 74.6 (57.6–91.6)

5.3 (3.4–7.2) 6.4 (0–13.9)

25.1 (21.1–29.1) 19.0 (2.9–35.2)

63.7 72.7 73.0 74.4

(57.0–70.3) 11.8 (6.7–16.9) (64.5–80.9) 5.6 (2.4–8.8) (70.2–75.9) 6.5 (4.7–8.3) (69.1–79.6) 5.0 (2.9–7.0)

24.5 21.7 20.5 20.7

(18.8–30.2) (14.6–28.8) (17.9–23.0) (15.5–25.9)

66.4 (59.6–73.2) 72.0 (69.8–74.3)

5.9 (2.9–8.9) 6.4 (5.2–7.6)

27.7 (21.2–34.2) 21.6 (19.5–23.6)

65.7 72.5 71.8 73.0

5.3 6.0 8.9 6.2

29.0 21.5 19.3 20.8

(59.7–71.7) (67.5–77.5) (65.8–77.8) (70.4–75.7)

(2.6–8.0) (3.3–8.7) (5.1–12.7) (4.8–7.6)

(23.3–34.7) (17.0–26.1) (14.0–24.6) (18.4–23.2)

68.7 (65.1–72.4) 73.1 (70.5–75.7)

6.8 (4.9–8.7) 6.0 (4.7–7.4)

24.5 (21.1–27.9) 20.9 (18.5–23.3)

71.1 (67.7–74.6) 71.5 (68.8–74.1)

6.5 (4.6–8.3) 6.3 (4.9–7.6)

22.4 (19.3–25.6) 22.3 (19.8–24.8)

68.3 (64.7–72.0) 72.7 (68.3–77.1) 74.2 (71.2–77.3)

6.0 (4.2–7.7) 6.2 (3.7–8.6) 7.0 (5.2–8.8)

25.7 (22.3–29.2) 21.1 (17.1–25.1) 18.8 (16.1–21.5)

71.2 (68.4–73.9) 70.9 (66.5–75.2) 72.9 (68.0–77.9)

6.4 (5.0–7.9) 5.7 (3.9–7.6) 7.0 (3.6–10.3)

22.4 (19.8–24.9) 23.4 (19.3–27.6) 20.1 (15.9–24.3)

73.3 (71.0–75.6) 58.8 (53.1–64.4)

6.3 (5.1–7.5) 6.5 (3.6–9.5)

20.4 (18.3–22.5) 34.7 (29.3–40.2)

76.6 70.0 64.7 46.3

7.4 5.7 6.1 6.6

16.0 24.3 29.2 47.1

(73.4–79.8) (67.1–73.0) (56.0–73.5) (30.4–62.3)

(5.4–9.5) (4.3–7.0) (1.8–10.5) (0.5–12.7)

(13.3–18.7) (21.6–27.1) (20.8–37.5) (30.8–63.4)

74.9 (72.3–77.4)

6.2 (4.8–7.5)

19.0 (16.6–21.3)

67.8 (64.4–71.2)

6.5 (4.8–8.3)

25.6 (22.5–28.8)

67.0 (61.7–72.3) 72.6 (70.3–74.9)

6.0 (3.4–8.7) 6.5 (5.2–7.7)

27.0 (22.0–32.0) 21.0 (18.9–23.1)

a Results are weighted to be representative of the California population and are adjusted for complex survey design effects b Census tracts in which at least 50% of the households were below 200% of the federal poverty level were considered lower income. FPL, federal poverty level.

34

S.H. Babey et al. / Preventive Medicine 76 (2015) 31–36

who did not (73% vs. 59%, respectively, p b 0.001). Being physically active during their last park visit was also related to perceptions of park safety. Specifically, adolescents who strongly agreed that the nearby park was safe had higher rates of park-based physical activity than those who disagreed or strongly disagreed (77% compared to 65% and 46%, respectively, p's b 0.01). Table 3 displays the multinomial regression results. Adjusting for the individual, family, and neighborhood characteristics displayed in the table, the present analyses show that increasing age was associated with a decreased likelihood of being physically active during their last park visit, and females were less likely than males to have been physically active the last time at a park. Adolescents with foreign-born fathers Table 3 Multinomial logistic regression of factors associated with park-based physical activity among adolescents, California, 2007.a

Factor Individual characteristics Age Gender Female Male Race Latino White Asian African American American Indian/Alaskan native Mixed race or other Adolescent's nativity Foreign-born US-born Family characteristics Household income Below 100% FPL 100–199% FPL 200–299% FPL 300% FPL and above Mother's nativity Foreign-born US-born Father's nativity Foreign-born US-born Parental educational attainment HS or less Some college College graduate Neighborhood characteristics Urbanity Urban Suburban Rural Park/playground within walking distance from home Yes No Nearby park/playground safe during day Strongly agree Agree Disagree Strongly disagree Neighborhood racial composition Less than 25% Latino/Hispanic 25% or more Latino/Hispanic Neighborhood incomec Lower income Higher income a

Active in park

Not active in park

Odds 95% CI ratiob

Odds 95% CI ratiob

0.90

0.84–0.96 1.28

1.15–1.44

0.67 1.00

0.53–0.85 1.75 1.00

1.13–2.69

0.92 1.00 0.64 1.13 1.03 1.11

0.61–1.37 0.61 1.00 0.39–1.05 1.47 0.68–1.89 0.74 0.31–3.43 1.21 0.77–1.6 0.74

0.27–1.41

0.83 1.00

0.55–1.25 0.55 1.00

0.27–1.13

0.81 1.24 1.29 1.00

0.55–1.20 1.00 0.85–1.81 1.39 0.85–1.95 2.00 1.00

0.47–2.12 0.65–2.96 1.07–3.71

0.82 1.00

0.57–1.18 1.10 1.00

0.58–2.11

1.67 1.00

1.14–2.45 1.36 1.00

0.75–2.48

0.76 0.89 1.00

0.53–1.08 0.74 0.64–1.23 0.78 1.00

0.43–1.28 0.44–1.39

0.67–3.21 0.29–1.89 0.25–5.82 0.36–1.52

were more likely than adolescents with US-born fathers to have been physically active during their last park visit than to have not visited the park. They were also less likely to have not been active in the park than to have not visited the park. Table 3 also shows that reporting having a park within walking distance of home and perceiving the nearby park to be safe during the day were significantly associated with park-based physical activity. Adolescents who reported having a park within walking distance of home were more likely to have been physically active during their last park visit compared with not visiting the park. They were also more likely to have not been active in the park than to have not visited a park. Adolescents who agreed or strongly agreed that their nearby park was safe during the day were more likely compared to those who strongly disagreed to have engaged in physical activity during their last park visit than to have not visited a park. Table 4 displays the adjusted probability of engaging in park-based physical activity for significant predictors. The largest differences in the adjusted percent of adolescents engaging in park-based physical activity were seen for age, having a park near home, and park safety. Discussion Using a large, population-based sample of adolescents, the current findings suggest that there are several individual, family, and neighborhood characteristics associated with engaging in physical activity in parks. Perceptions of park availability and of park safety were associated with engaging in physical activity in parks, while being female or older were associated with being less likely to engage in park-based physical activity. Adolescents who lived near a park, were female, and were older were more likely to not be active in the park than to not visit park. The current results highlight considerable variation in park-based physical activity as a function of living near a park (75% of those with a nearby park were predicted to engage in park-based physical activity compared to 61% among those with no park, adjusting for covariates) and park safety (78% of adolescents who strongly agree that their park is safe were predicted to be active in the park compared to just 52% among those who strongly disagree, adjusting for covariates). Neither household income nor neighborhood income were associated with parkbased physical activity among adolescents. These findings are consistent with some previous studies, but not with others. In previous research, younger age, being male, and having Table 4 Adjusted probability of engaging in park-based physical activity, adolescents, California, 2007. Factor

1.00 0.83 1.14

1.00 0.62–1.11 0.81 0.81–1.62 1.19

2.12 1.00

1.57–2.84 1.66 1.00

1.01–2.75

3.83 2.47 2.09 1.00

1.93–7.61 2.90 1.27–4.82 1.64 0.97–4.49 1.43 1.00

0.91–9.29 0.53–5.06 0.39–5.29

1.00 0.77

1.00 0.58–1.02 1.12

0.92 1.00

0.65–1.30 1.00 1.00

0.50–1.32 0.64–2.20

0.67–1.88 0.53–1.87

Results are weighted to be representative of the California population and are adjusted for complex survey design effects. b Adjusted for all variables presented in this table. c Census tracts in which at least 50% of the households were below 200% of the federal poverty level were considered lower income. FPL, federal poverty level.

Individual characteristics Age 12 14.5 17 Gender Female Male Family characteristics Father's nativity Foreign-born US-born Neighborhood characteristics Park/playground within walking distance from home Yes No Nearby park/playground safe during day Strongly agree Agree Disagree Strongly disagree

%a

95% CI

80.2 73.8 64.9

76.8–83.5 71.7–75.9 60.7–69.1

68.1 78.0

64.9–71.2 75.2–80.8

78.2 69.5

74.2–82.1 65.6–73.4

75.2 60.9

73.0–77.4 54.9–66.9

77.8 71.8 68.7 52.4

74.5–81.1 68.9–74.8 60.7–76.7 37.0–67.8

a Probabilities calculated as predictive margins controlling for all factors included in the multinomial logistic regression model held at mean values (see Table 3).

S.H. Babey et al. / Preventive Medicine 76 (2015) 31–36

a mother with higher levels of education were significantly associated with more physical activity (Addy et al., 2004; Cohen et al., 2007; Gordon-Larsen et al., 2000; McKenzie et al., 2006; Prins et al., 2009). Although the current study found similar associations for age and gender, parental educational attainment was not associated with park-based physical activity. However, other studies found no association between parental educational attainment and youth physical activity (Sallis et al., 2002). Although household income was rarely included in previous studies of park-based physical activity, observational studies of park users have found that lower neighborhood income was associated with park-based physical activity (Floyd et al., 2008a, 2008b). Another study found that neighborhood poverty was negatively associated with the number of park users and level of physical activity in parks (Cohen et al., 2012). The current study, however, did not find an association between neighborhood income and adolescent park-based physical activity. The current research is consistent with previous research, suggesting that proximity to parks is positively related to physical activity in parks (Cohen et al., 2007, 2012). Some studies have found no association between proximity to parks and park-based physical activity (Kaczynski et al., 2008, 2009). However, those studies did not include youth. Previous research on the relationship of safety to park-based physical activity has been mixed. Cohen and colleagues found that perceived park safety was not associated with park use or physical activity (Cohen et al., 2007). However, other research suggests that perceptions of park safety are positively related to park use among youth, but the relationship varies with urbanicity and race (Loukaitou-Sideris and Sideris, 2010). Previous research is also mixed with regard to the relationship between park-based physical activity and neighborhood racial composition. Studies of park users in Tampa and Chicago found that greater park-based physical activity was associated with higher concentrations of Hispanic residents in Tampa and with higher concentrations of African American residents in Chicago (Floyd et al., 2008a, 2008b). However, these studies included park users of all ages. Other studies focusing on youth did not find an association between neighborhood racial composition and park-based physical activity (Floyd et al., 2011; Spengler et al., 2011). Discrepancies between the current results and some previous studies may be due to differences in correlates of overall physical activity and park-based physical activity. In addition, previous research differs in the age groups included (i.e., some studies included children of all ages or park users of all ages, others included only adults) and definitions of neighborhood racial composition used in the analyses (i.e., Floyd and colleagues selected parks in neighborhoods with census blocks greater than 50% Latino/Hispanic rather than census tracts with 25% or more Latino/Hispanic) (Floyd et al., 2008a, 2008b). Finally, many of the park-based physical activity studies were only able to include characteristics that could be observed by coders (e.g., gender, race), which meant that they could not control for other unobservable factors such as income, education, and nativity. Increasing the availability of parks is often recommended as an environmental change that can promote active living. The current results indicate that adolescents who reported greater availability and safety of nearby parks were more likely to engage in park-based physical activity, adjusting for a range of individual and family characteristics. However, adolescents who reported living near a park were also more likely to not be active in the park than to not visit the park, suggesting that park availability may be more predictive of visiting a park than of park-based physical activity. Increasing the availability of local parks and improving the perceptions of safety of these local parks may be an important strategy for increasing the use of parks for physical activity and contributing to current efforts aimed at reducing the prevalence of obesity among youth. This study has some limitations. Information on the level or amount of physical activity that respondents engaged in during their last park

35

visit is not known, since respondents were only asked whether or not they were physically active during their last park visit. In the current study, 78% of adolescents recently visited a park. However, nearly all of the adolescents with a recent park visit were physically active during their last park visit (92%). As a result, it is difficult to distinguish between predictors of visiting a park and predictors of park-based physical activity. Additionally, it is not clear whether park-based physical activity increases overall physical activity. If adolescents substitute one type of physical activity for another, then park-based physical activity is less likely to impact health. However, previous research suggests that physical activity occurring in parks and overall physical activity may be related. Having access to parks and number of park visits have both been associated with more overall physical activity as well as more parkbased physical activity (Cohen et al., 2006, 2012; Frank et al., 2007; Kaczynski and Henderson, 2007; Kaczynski et al., 2009; Roemmich et al., 2006). Future research should directly examine the relationship between park-based and overall physical activity. Self-reported measures such as perceptions of availability of parks and safety of parks may not perfectly correspond to objective measures of these variables. However, previous research suggests that perceptions of the neighborhood environment are important correlates of physical activity, and can sometimes be better predictors than objective measures (Davison and Lawson, 2006; Janssen, 2014; McGinn et al., 2007; Mota et al., 2005). Also, data on characteristics of the parks were not available, and previous research suggests that park facilities, staffing, and programming are associated with park-based physical activity. Finally, due to the cross-sectional nature of the data, causality, and the direction of the relationship between variables cannot be inferred. This research employed population-based data to examine a wide range of factors associated with adolescent park-based physical activity. To the authors' knowledge, no previous studies of park-based physical activity among adolescents have included this range of sociodemographic, family, and neighborhood characteristics. The current results provide information that can shape and inform interventions intended to promote physical activity in parks among youth. Strategies that increase availability and safety of parks and other recreation spaces may be particularly effective. Conflict of interest The authors declare that there are no conflicts of interest. Acknowledgments This work was supported by grants 20081418 and 20102168 from The California Endowment. References Addy, C.L., Wilson, D.K., Kirtland, K.A., Ainsworth, B.E., Sharpe, P., Kimsey, D., 2004. Associations of perceived social and physical environmental supports with physical activity and walking behavior. Am. J. Public Health 94, 440–443. Babey, S.H., Hastert, T.A., Yu, H., Brown, E.R., 2008. Physical activity among adolescents. When do parks matter? Am. J. Prev. Med. 34, 345–348. Berenson, G.S., Srnivasan, S.R., 2005. Cardiovascular risk factors in youth with implications for aging: the Bogalusa Heart Study. Neurobiol. Aging 26, 303–307. Brownson, R.C., Baker, E.A., Housemann, R.A., Brennan, L.K., Bacak, S.J., 2001. Environmental and policy determinants of physical activity in the United States. Am. J. Public Health 91, 1995–2003. California Health Interview Survey, 2009a. CHIS 2007 Methodology Series: Report 1—Sample Design. UCLA Center for Health Policy Research, Los Angeles, CA. Carlson, S.A., Brooks, J.D., Brown, D.R., Buchner, D.M., 2010. Racial/ethnic differences in perceived access, environmental barriers to use, and use of community parks. Prev. Chronic Dis. 7, A49. Cohen, D.A., Ashwood, J.S., Scott, M.M., et al., 2006. Public parks and physical activity among adolescent girls. Pediatrics 118, e1381–e1389. Cohen, D.A., McKenzie, T.L., Sehgal, A., Williamson, S., Golinelli, D., Lurie, N., 2007. Contribution of public parks to physical activity. Am. J. Public Health 97, 509–514.

36

S.H. Babey et al. / Preventive Medicine 76 (2015) 31–36

Cohen, D.A., Golinelli, D., Williamson, S., Sehgal, A., Marsh, T., McKenzie, T.L., 2009. Effects of park improvements on park use and physical activity: policy and programming implications. Am. J. Prev. Med. 37, 475–480. Cohen, D.A., Han, B., Derose, K.P., et al., 2012. Neighborhood poverty, park use, and parkbased physical activity in a Southern California city. Soc. Sci. Med. 75, 2317–2325. Cohen, D.A., Lapham, S., Evenson, K.R., et al., 2013. Use of neighbourhood parks: does socio-economic status matter? A four-city study. Public Health 127, 325–332. Davison, K.K., Lawson, C.T., 2006. Do attributes in the physical environment influence children's physical activity? A review of the literature. Int. J. Behav. Nutr. Phys. Act. 3, 19. Delva, J., Johnston, L.D., O'Malley, P.M., 2007. The epidemiology of overweight and related lifestyle behaviors: racial/ethnic and socioeconomic status differences among American youth. Am. J. Prev. Med. 33, S178–S186. Eckel, R.H., Krauss, R.M., 1998. American Heart Association call to action: obesity as a major risk factor for coronary heart disease. AHA Nutrition Committee. Circulation 97, 2099–2100. Floyd, M.F., Spengler, J.O., Maddock, J.E., Gobster, P.H., Suau, L., 2008a. Environmental and social correlates of physical activity in neighborhood parks: an observational study in Tampa and Chicago. Leis. Sci.: An Interdiscip. J. 30, 360–375. Floyd, M.F., Spengler, J.O., Maddock, J.E., Gobster, P.H., Suau, L.J., 2008b. Park-based physical activity in diverse communities of two U.S. cities: an observational study. Am. J. Prev. Med. 34, 299–305. Floyd, M.F., Bocarro, J.N., Smith, W.R., et al., 2011. Park-based physical activity among children and adolescents. Am. J. Prev. Med. 41, 258–265. Frank, L., Kerr, J., Chapman, J., Sallis, J., 2007. Urban form relationships with walk trip frequency and distance among youth. Am. J. Health Promot. 21, 305–311. Gordon-Larsen, P., McMurray, R.G., Popkin, B.M., 2000. Determinants of adolescent physical activity and inactivity patterns. Pediatrics 105, E83. Graubard, B.I., Korn, E.L., 1999. Predictive margins with survey data. Biometrics 55, 652–659. California Health Interview Survey, 2009b. CHIS 2007 Methodology Series: Report 2—Data Collection Methods. UCLA Center for Health Policy Research, Los Angeles, CA. California Health Interview Survey, 2009c. CHIS 2007 Methodology Series: Report 3—Data Processing Procedures. UCLA Center for Health Policy Research, Los Angeles, CA. California Health Interview Survey, 2009d. CHIS 2007 Methodology Series: Report 4—Response Rates. UCLA Center for Health Policy Research, Los Angeles, CA. California Health Interview Survey, 2009e. CHIS 2007 Methodology Series: Report 5—Weighting and Variance Estimation. UCLA Center for Health Policy Research, Los Angeles, CA. Janssen, I., 2014. Crime and perceptions of safety in the home neighborhood are independently associated with physical activity among 11–15 year olds. Prev. Med. 66, 113–117. Kaczynski, A.T., Henderson, K.A., 2007. Environmental correlates of physical activity: a review of evidence about parks and recreation. Leis. Sci. 29, 315–354. Kaczynski, A.T., Potwarka, L.R., Saelens, B.E., 2008. Association of park size, distance, and features with physical activity in neighborhood parks. Am. J. Public Health 98, 1451–1456. Kaczynski, A.T., Potwarka, L.R., Smale, B.J.A., Havitz, M.E., 2009. Association of Parkland Proximity with Neighborhood and Park-based Physical Activity: variations by gender and age. Leis. Sci.: An Interdiscip. J. 31, 174–191. Kaczynski, A.T., Johnson, A.J., Saelens, B.E., 2010. Neighborhood land use diversity and physical activity in adjacent parks. Health Place 16, 413–415. Kaczynski, A.T., Wilhelm Stanis, S.A., Hastmann, T.J., Besenyi, G.M., 2011. Variations in observed park physical activity intensity level by gender, race, and age: individual and joint effects. J. Phys. Act. Health 8 (Suppl. 2), S151–S160.

Kaczynski, A.T., Stanis, S.A., Besenyi, G.M., Child, S., 2013. Differences in youth and adult physical activity in park settings by sex and race/ethnicity. Prev. Chronic Dis. 10, E42. Kvaavik, E., Tell, G.S., Klepp, K.I., 2003. Predictors and tracking of body mass index from adolescence into adulthood: follow-up of 18 to 20 years in the Oslo Youth Study. Arch. Pediatr. Adolesc. Med. 157, 1212–1218. Lee, S., Brown, E.R., Grant, D., Belin, T.R., Brick, J.M., 2009. Exploring nonresponse bias in a health survey using neighborhood characteristics. Am. J. Public Health 99, 1811–1817. Loukaitou-Sideris, A., Sideris, A., 2010. What brings children to the park? Analysis and measurement of the variables affecting children's use of parks. J. Am. Plann. Assoc. 76, 89–107. McGinn, A.P., Evenson, K.R., Herring, A.H., Huston, S.L., Rodriguez, D.A., 2007. Exploring associations between physical activity and perceived and objective measures of the built environment. J. Urban Health 84, 162–184. McKenzie, T.L., Cohen, D.A., Sehgal, A., Williamson, S., Golinelli, D., 2006. System for Observing Play and Recreation in Communities (SOPARC): reliability and feasibility measures. J. Phys. Act. Health 3 (Suppl. 1), S208–S222. Mota, J., Almeida, M., Santos, P., Ribeiro, J.C., 2005. Perceived neighborhood environments and physical activity in adolescents. Prev. Med. 41, 834–836. Mowen, A.J., Payne, L.L., Scott, D., 2005. Change and stability in park visitation constraints revisited. Leis. Sci. 27, 191–204. Prins, R.G., Oenema, A., van der Horst, K., Brug, J., 2009. Objective and perceived availability of physical activity opportunities: differences in associations with physical activity behavior among urban adolescents. Int. J. Behav. Nutr. Phys. Act. 6, 70. Ries, A.V., Voorhees, C.C., Roche, K.M., Gittelsohn, J., Yan, A.F., Astone, N.M., 2009. A quantitative examination of park characteristics related to park use and physical activity among urban youth. J. Adolesc. Health 45, S64–S70. Roemmich, J.N., Epstein, L.H., Raja, S., Yin, L., Robinson, J., Winiewicz, D., 2006. Association of access to parks and recreational facilities with the physical activity of young children. Prev. Med. 43, 437–441. Sallis, J.F., Taylor, W.C., Dowda, M., Freedson, P.S., Pate, R.R., 2002. Correlates of vigorous physical activity for children in grades 1 through 12: comparing parent-reported and objectively measured physical activity. Pediatr. Exerc. Sci. 14, 30–44. Sallis, J.F., Cervero, R.B., Ascher, W., Henderson, K.A., Kraft, M.K., Kerr, J., 2006. An ecological approach to creating active living communities. Annu. Rev. Public Health 27, 297–322. Shores, K.A., West, S.T., 2008. The relationship between built park environments and physical activity in four park locations. J. Public Health Manag. Pract. 14, e9–e16. Singh, A.S., Mulder, C., Twisk, J.W., van Mechelen, W., Chinapaw, M.J., 2008. Tracking of childhood overweight into adulthood: a systematic review of the literature. Obes. Rev. 9, 474–488. Spengler, J.O., Floyd, M.F., Maddock, J.E., Gobster, P.H., Suau, L.J., Norman, G.J., 2011. Correlates of park-based physical activity among children in diverse communities: results from an observational study in two cities. Am. J. Health Promot. 25, e1–e9. Strong, W.B., Malina, R.M., Blimkie, C.J.R., et al., 2005. Evidence based physical activity for school-age youth. J. Pediatr. 146, 732–737. U.S. Department of Health and Human Services, 2008. 2008 Physical Activity Guidelines for Americans. U.S. Department of Health and Human Services, Washington, DC. US Department of Health and Human Services, 1996. Physical Activity and Health: A Report of the Surgeon General. US Dept of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Atlanta, GA. Wannamethee, S.G., Shaper, A.G., 1999. Weight change and duration of overweight and obesity in the incidence of type 2 diabetes. Diabetes Care 22, 1266–1272.