Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Mar 1.
Published in final edited form as: Health Place. 2019 Feb 10;56:147–154. doi: 10.1016/j.healthplace.2019.01.015

Neighborhood built environment associations with adolescents’ location-specific sedentary and screen time

Carolina M Bejarano a, Jordan A Carlson b,*, Christopher C Cushing c, Jacqueline Kerr d, Brian E Saelens e, Lawrence D Frank f, Karen Glanz g, Kelli L Cain d, Terry L Conway d, James F Sallis d
PMCID: PMC6477917  NIHMSID: NIHMS1521186  PMID: 30743089

Introduction

Sedentary time, screen time in particular, appears to have potential as a significant intervention target for obesity prevention and improved cardiometabolic health in youth. Although evidence is mixed regarding the health consequences of objectively measured sedentary time in youth (Carson et al., 2014; Cliff et al., 2016; Saunders et al., 2013; Tremblay et al., 2011, van Ekris et al., 2016), there is a large literature linking TV and screen time with greater risk of youth obesity (Tremblay et al., 2011). Furthermore, there are more consistent associations between greater objective sedentary time and numerous poor health outcomes in adults (Owen et al., 2010; Proper et al., 2011), and adolescence is a key developmental period where health-related habits are established and can be carried into adulthood (Telama, 2009). A next step in this research is to better understand the determinants of sedentary time to inform effective interventions.

Ecological models posit that health behaviors are influenced by individual-, interpersonal-, and environmental-level factors (Sallis et al., 2015). Neighborhood attributes, specifically land use mix and residential density, have shown promise as environmental-level factors that are related to physical activity in adolescents (Ding et al, 2011), but less is known about how neighborhood environments relate to youth’s sedentary time. Although there appears to be a clear link between neighborhood environments and walking in adults (Owen et al., 2007) and youth (Carlson et al., 2015), a small increase in walking time likely has little impact on offsetting total sedentary time because a very large proportion of the day is spent being sedentary. The home setting is where a large proportion of sedentary time occurs and is supported (e.g., by couches, TVs, computers; Maitland et al., 2013). A more substantial way neighborhood environments may impact sedentary time could be by creating appeal for youth to spend time outside their home where they may be more likely to engage in non-sedentary activity. However, this hypothesis has been seldom investigated (Chastin et al., 2016).

Few studies have examined neighborhood environments in relation to objectively measured sedentary time in youth, and findings have generally been null (McDonald et al., 2012; Timperio et al., 2017; Veitch et al., 2011) or mixed (Hinckson et al., 2017). One study found that better objective sidewalk quality was related to less total sedentary time in youth (Jago et al., 2006). Hinckson et al. (2017) found that favorable perceived, but not objective, neighborhood environment attributes were associated with less total sedentary time in adolescents. Timperio and colleagues studied typologies (i.e., seven objectively measured neighborhood attributes that yielded four distinct clusters) of children’s neighborhood environments and found no significant differences in objectively measured sedentary time between typologies (2017). However, having a neighborhood profile indicative of little mixed use development, few playgrounds, low street connectivity, and low traffic was related to greater television viewing in 5 to 6 year olds, while having greater mixed land use and many playgrounds was associated with less TV viewing three years later (Timperio et al., 2017).

Due to the lack of comprehensive and consistent findings, further research is necessary to clarify the potential impact of the neighborhood environment on sedentary time in youth. There has been more research examining these relationships in adults, and a recent review paper found mixed associations between the neighborhood environment and sedentary time (Koohsari et al., 2015). There were consistent findings of lower sedentary time for residents of urban versus rural and suburban areas, but mixed findings regarding the associations of walkability, crime, streetscapes, and aesthetics with sedentary time.

It is notable that previous studies, both in adults and youth, have been limited by examining total sedentary time without regard to where the sedentary time occurs (e.g., within the home, in the home neighborhood, at school, at work, or elsewhere). For example, studies examining residential neighborhood environments have not isolated the sedentary time spent in one’s home neighborhood. Authors of studies involving youth and adults have suggested that future research focusing on context-specific sedentary time would provide more clarity regarding how environment is related to sedentary behavior (Hinckson et al., 2017; Compernolle et al., 2017).

Present Study

The objective of the present study was to investigate relations of neighborhood environment attributes to objectively measured total sedentary time, home sedentary time, home neighborhood sedentary time, and time spent at home, and self-reported TV time and total screen time in adolescents. We hypothesized that adolescents living in a more favorable neighborhood environment for physical activity (Ding et al., 2011) would spend less time at home, in total sedentary time, watching TV, and engaging in total screen time as compared to those living in a less activity-supportive neighborhood environment. We also hypothesized that those living in a more favorable neighborhood environment would spend less time being sedentary when in their neighborhood and at home.

The present study addressed limitations of past studies by including both perceived and objective measurements of the neighborhood environment, as well as investigating both self-reported and objective measures of sedentary time. Following recommendations for research in this area (Hinckson et al., 2017), the present study provided more precision in the dependent variables by using the combination of global positioning systems (GPS) and accelerometer data to examine the places where sedentary time occurred.

Methods

Participants and Procedures

Participants were 524 adolescents who participated in the Teen Environment and Neighborhood (TEAN) observational study of neighborhood environments and physical activity. This study was conducted in the Baltimore, Maryland-Washington, DC and Seattle-King County, Washington metropolitan areas in the United States. Adolescents and one parent/guardian per adolescent were selected from census block groups representing high or low walkability and high or low income. Four block group types were created based on median splits of GIS-based measures of walkability (Frank et al., 2010) and census-based household income (2 X 2 matrix defined by high/low walkability X high/low household income) as described previously (Sallis et al., 2018). Participants were recruited through contacting households with adolescents ages 12-16, via mail and telephone, identified by a list purchased from a marketing company. Participants were screened and considered ineligible for the study if they had a physical disability that affected their physical activity, any conditions that affected their dietary habits, and any developmental disability that would limit their ability to participate. Parents and adolescents provided informed consent and assent, respectively. All procedures were approved by the Institutional Review Board of the sponsoring university. Participants were mailed an accelerometer and Global Positioning System (GPS) tracker with instructions to wear the devices during all waking hours for seven days, except during water activities. Participants had the option of completing study surveys online or on paper sent through the postal mail. Of the 928 participants from the full sample, 130 were not given a GPS device and 152 did not wear the provided accelerometer and GPS tracker together for ≥1 valid school day and ≥1 valid non-school day; these participants were excluded from present analyses. Participants who attended homeschool, did not provide their school address, or had geocoding errors were also excluded (n = 122). Therefore, the analytic sample for the present study was 524 participants.

Measures

Demographic information and anthropometrics

Adolescents completed a survey to report their age, gender, and race/ethnicity. Parents completed a survey to report on their highest level of education, marital status, and number of vehicles per licensed driver in the household. The parent of each participant measured and reported their child’s height and weight based on detailed instructions provided by the research team. BMI percentiles were calculated using the Centers for Disease Control and Prevention growth charts (CDC 2000).

Perceived neighborhood environment

Parents completed a subset of the Neighborhood Environment Walkability Scale for Youth (NEWS-Y), which assessed various aspects of the perceived neighborhood environment around their home. The subscales that were examined in the present study were land use mix-access (2 items; e.g., parking is difficult near shopping), street connectivity (2 items; e.g., many different routes to get from place to place), walking facilities (3 items; e.g., sidewalks on most of streets), neighborhood aesthetics (4 items; nice things to look at), traffic safety (3 items; traffic makes it unpleasant to walk), pedestrian safety (3 items; crosswalks on busy streets), and crime safety (5 items; high crime rate). As compared to the original NEWS-Y, the “pedestrian and automobile traffic safety” scale was split into separate scales to investigate traffic and pedestrian safety separately. Parents could indicate responses on a scale ranging from 1-4, with items averaged within each subscale and higher numbers representing greater walkability or safety (some items were reverse scored). Test-retest ICCs for NEWS-Y subscales have ranged from .61 to .78 in prior work (Rosenberg et al., 2009).

Objective neighborhood environment

Objective measures of the built environment were derived from data from the county tax assessor, regional land use at the parcel level, and street networks. These data were integrated into GIS to derive built environment features within .5 km street network buffers around each participant's home. Based on recommendations from past research, this study examined neighborhoods using a smaller buffer size than sometimes used in physical activity studies, but that may be more appropriate for detecting the hypothesized effects, particularly in youth (McDonald et al., 2012). Using methods described in a previous study (Frank et al., 2010; Saelens et al., 2012), the variables of residential density (housing units per residential parcel), street connectivity (intersections per square km), retail floor area ratio (building square ft per parcel square ft, with higher values reflecting more pedestrian-oriented design), mixed land use (includes residential, retail, food and entertainment, and office land use types; 0 = single use and 1 = even distribution across the 5 uses), cul-de-sac density (number of cul-de-sacs per square km), and number of parks were calculated.

Screen time

Adolescents reported on time spent watching television, videos, or DVDs (one item) during a typical school day but not during school hours, which made up the TV time dependent variable. Response options included None, 15 minutes, 30 minutes, 1 hour, 2 hours, 3 hours, and 4 or more hours per day. Reported time spent watching television was combined with reported time engaged in sedentary computer or video games and reported time using the internet or electronic media for leisure, to create the total screen time dependent variable (sum of three items). Past studies using similar measures for time spent watching television have found good test-retest reliability and construct validity based on associations with home sedentary environment and psychosocial variables (Norman et al., 2005).

Objectively measured sedentary time and physical activity

Participants wore the accelerometer (ActiGraph models 7164, GT1M & GT3X, Pensacola, FL) on their right hip. Data from the accelerometers were downloaded and converted to minutes engaged in sedentary time using a cut point of ≤ 100 counts per minute (Evenson et al., 2008; Treuth et al., 2004) applied to 30-second epochs. The Evenson cut point applied to 30-second epochs was used for calculating moderate-intensity physical activity (MVPA), which was used as a covariate (Evenson et al., 2008; Physical Activity Guidelines for Americans, 2008). Non-wear periods were considered >60 consecutive epochs (30 minutes) with count = 0, and such periods were excluded from analyses. A minimum of 8 hours of wear time per day was necessary for inclusion, which is considered an acceptable threshold for wear time in youth (Mattocks et al., 2008). Days with >960 minutes (16 hours) of daily accelerometer wear time were excluded because the participant could have worn the device while sleeping, which could have profound impacts on their estimated sedentary time.

GPS-assessed locations

Latitude and longitude data were collected at 30-second epochs from the GPS tracker. Participants who met the accelerometer wear criteria and also had >8 hours of daily GPS signal were included. The GPS and accelerometer data were merged using the Personal Activity and Location Measurement System Version 4 (PALMS; Center for Wireless and Population Health). The devices were time synchronized during initialization and linked in PALMS with a time stamp. Details of the spatial analysis methods have been previously presented (Carlson et al., 2016). In brief, each participant’s home and school address was geocoded and integrated into ArcGIS (ESRI, Inc., Redlands, CA). Exposure to the home neighborhood was defined in two different ways. A 50-m-radius circular buffer around the point resulting from geocoding the home address was used to define the ‘at-home’ location. A 1-km street network buffer around the home address, excluding the ‘at-home’ buffer and when applicable excluding the participant’s school (15 meter buffer around the school parcel), was used to define the ‘home neighborhood’ location. Total time (while the devices were worn) and sedentary time were calculated specific to the at-home and home neighborhood locations, as well as overall (across all locations). These calculations involved filters for spending at least 60 minutes per day at home for total sedentary time and home sedentary time, spending at least 60 minutes per day at or in the home neighborhood for home neighborhood sedentary time, and spending at least 15 minutes per day at home for the time spent at home variable. These procedures were meant to exclude days when participants had minimal exposure to their home neighborhood environment by spending little time at home, which would also result in abnormally large or small proportions of time in the at-home and home neighborhood locations being sedentary.

Data Analysis

Independent variables were those calculated from the NEWS-Y scales and GIS measures of the neighborhood built environment. All independent variables were computed as sample-specific z-scores to standardize across various scales. Dependent variables were minutes per day of TV time, total screen time, total sedentary time, home sedentary time, home neighborhood sedentary time, and time spent at home. Index scores were calculated for the perceived and objective neighborhood environment, respectively. The indices were calculated by taking an average of the component variable z scores (7 for the perceived index and 5 for the objective index). The objective index excluded cul-de-sac density as its effects appeared to be distinct from the other objectively measured variables. As preliminary data analysis, Pearson correlations were run between time spent at home and each of the sedentary time dependent variables. Next, mixed effects models were run in SPSS Version 24 to account for the nested data structure, with census block group included as a random effect. Each independent variable was entered into a separate model, and each model was adjusted for location-specific wear time; participant characteristics of gender, age, race/ethnicity, and BMI percentile; city, neighborhood income, vehicle access, and parent education and marital status (90 models total for 15 independent variables × 6 dependent variables). Models were also adjusted for MVPA, with the exception of the models investigating time spent at home. Participant level data files were used for the analyses.

Results

Descriptive Results

The sample size for the study was 524 adolescents (Mean age = 14.13, 50.6% female, 70.1% White, non-Hispanic) after excluding participants as described in the procedures. Table 1 provides demographic information and Table 2 provides means and standard deviations of the study variables. Associations were examined between time spent at home with TV time (r = .02), total screen time (r = .05), total sedentary time (r = .09), home sedentary time (r = .98), and home neighborhood sedentary time (r = .61). Of the days of data included, 54% were school days (n = 1,581) and 46% were non-school days (n = 1,858). The mean number of school days per participant was 3.5, SD = 1.5, and the mean number of non-school days per participant was 3.0, SD = 1.6.

Table 1.

Demographic Information

Mean (SD) or n (%)
Site
   Baltimore/MD 253 (48.3%)
   Seattle/King County 271 (51.7%)
Age 14.13 (1.44)
Gender
   Boys 259 (49.4%)
   Girls 265 (50.6%)
Ethnicity
   White, Non-Hispanic 366 (70.1%)
   Non-White or Hispanic 156 (29.9%)
Walkability of neighborhood
   High Walkability 240 (45.8%)
   Low Walkability 284 (54.2%)
Income level of neighborhood
   High Income 264 (50.4%)
   Low Income 260 (49.6%)
Vehicle Access 1.09 (.40)
Parental Education
   Some college or less 183 (35.1%)
   College degree or more 338 (64.9%)
Parent Marital Status
   Married or living with partner 437 (83.7%)
   Not married or living with partner 85 (16.3%)
Youth BMI Percentile 64.17 (26.44)
Youth BMI z-score .47 (.96)
Open in a new tab

Note. n = 524. Vehicle access indicates ratio of drivable motor vehicles to licensed drivers in the household. BMI = Body Mass Index. Ethnicity (n = 2), parental education (n = 3), and parental marital status (n = 2) was not reported for some participants.

Table 2.

Descriptive characteristics for study variables

Mean (SD)
Sedentary and wear time
 Total wear time (minutes/day) 778.32 (69.08)
 Time at home (minutes/day)a 310.73 (157.27)
 Time in home neighborhood (minutes/day) 362.08 (152.26)
 Total sedentary time (minutes/day)b 511.05 (70.29)
 Home sedentary time (minutes/day)b 219.61 (99.81)
 Home neighborhood sedentary time (minutes/day)c 238.99 (106.24)
 TV time (minutes/day) 84.50 (70.81)
 Total screen time (minutes/day) 206.68 (150.93)
Perceived neighborhood environment around home
 Land use mix-access (1-4) 2.45 (.57)
 Street connectivity (1-4) 2.76 (.62)
 Walking facilities (1-4) 2.62 (.89)
 Neighborhood aesthetics (1-4) 3.10 (.63)
 Traffic safety (1-4) 2.51 (.56)
 Pedestrian safety (1-4) 2.82 (.64)
 Crime safety (1-4) 3.04 (.69)
Objective neighborhood environment around home
 Residential density (housing units/ parcel) 6.15 (11.17)
 Street connectivity (intersections/sq km) 91.05 (30.60)
 Retail floor area ratio (building:parcel sq ft) .11 (.22)
 Mixed use (0=single 1=mixed) .08 (.14)
 Cul-de-sac density (cul-de-sacs/sq km) 18.55 (18.64)
 Number parks (parks/sq km) .49 (.77)
Open in a new tab

Note. n = 524

a

Filter required participants to have spent at least 15 minutes at home.

b

Filter required participants to have spent at least 60 minutes at home.

c

Filter required that participants to have spent at least 60 minutes at home or in the home neighborhood

Perceived Neighborhood Environment

See Table 3 for results of the mixed models. Perceived neighborhood environment variables of land use mix-access, and walking facilities were significantly negatively associated with home neighborhood sedentary time. A one standard deviation increase in land use mix-access was associated with 9.83 fewer minutes of home neighborhood sedentary time, while a one standard deviation increase in walking facilities was associated with 16.31 fewer minutes of home neighborhood sedentary time. Significant negative associations also emerged between perceived neighborhood aesthetics and the dependent variables of daily TV time, daily total screen time, and time spent at home. An increase of one standard deviation in neighborhood aesthetics was associated with 9.77 fewer minutes of daily TV time, 17.94 fewer minutes of daily total screen time, and 15.7 fewer minutes per day of time spent at home.

Table 3.

Relations of neighborhood environment attributes to TV, screen, total and location-specific sedentary time, and time spent at home in youth

TV time
(minutes/day)a
n = 524		 Total screen time
(minutes/day)a
n = 524		 Total sedentary
time
minutes/day)a
n = 397		 Home sedentary
time
(minutes/day) a
n = 397		 Home neighborhood
sedentary time
(minutes/day)a
n = 509		 Time at home
(minutes/day)b
n = 397
B (95% CI) p B (95% CI) p B (95% CI) p B (95% CI) p B (95% CI) p B (95% CI) p
Perceived neighborhood environment around home
 Land use mix-access (z score) −1.67 (−7.62, 4.28) .582 −7.20 (−19.83, 5.44) .264 2.14 (−2.39 6.68) .354 .03 (−2.38, 2.44) .982 −9.83 (−17.74, −1.92) .015 −5.80 (−20.91, 9.31) .451
 Street connectivity (z score) 1.13 (−4.67, 6.93) .383 6.29 (−6.06, 18.65) .317 1.57 (−2.81, 5.94) .482 −1.11 (−3.45, 1.23) .352 −2.38 (−10.38, 5.62) .559 10.23 (−4.42, 24.88) .171
 Walking facilities (z score) −2.41 (−8.45, 3.63) .433 −8.30 (−21.07, 4.46) .202 2.02 (−2.66, 6.71) .396 1.47 (−1.01 3.96) .245 −16.31 (−24.23, −8.39 .000 15.26 (−.24, 30.76) .0.54
 Neighborhood aesthetics (z score) −9.77 (−15.71, −3.82) .001 −17.94 (−30.65, −5.23) .006 2.56 (−2.14, 7.26) .285 −1.40 (−3.91, 1.11) .274 .17 (−8.07, 8.40) .968 −15.70 (−30.97, −.43) .044
 Traffic safety (z score) −3.58 (−9.35, 2.18) .223 −6.84 (−19.15, 5.46) .275 −1.57 (−6.01, 2.87) .489 .29 (−2.08, 2.66) .810 −6.12 (−13.99, 1.74) .127 −3.38 (−17.99, 11.22) .649
 Pedestrian safety (z score) .99 (−4.90, 6.89) .741 4.95 (−7.59, 17.50) .438 1.28 (−3.25, 5.81) .579 −.01 (−2.42, 2.40) .991 −6.26 (−14.23, 1.71) .123 .03 (−15.07, 15.14) .997
 Crime safety (z score) −4.63 (−10.74, 1.48) .137 −12.82 (−25.81, .17) .053 −1.24 (−5.86 3.39) .600 −.70 (−3.16, 1.75) .573 −1.28 (−9.53, 6.98) .761 −1.27 (−16.73, 14.19) .872
 Perceived neighborhood environment index (z score) −13.60 (−26.67, −.53) .041 −28.15 (−55.96, −.35) .047 4.71 (−5.45, 14.87) .363 −1.03 (−6.42, 4.35) .707 −29.09 (−46.59, −11.60) .001 −.83 (−34.47, 32.82) .962
Objective neighborhood environment around home
 Residential density (z score) .58 (−5.32, 6.48) .063 4.36 (−8.19, 16.89 .495 1.63 (−2.35, 5.61 .420 1.18 (−.94, 3.30) .273 −3.45 (−11.32, 4.42) .390 1.20 (−12.46, 14.86) .863
 Street connectivity (intersection density) (z score) −5.23 (−11.38, .91) .067 4.27 (−8.74, 17.29) .519 6.53 (1.71, 11.25) .008 2.42 (−.16, 5.00) .066 −10.90 (−19.04, −2.76) .009 6.21 (−9.92, 22.34) .449
 Retail floor area ratio (z score) 2.83 (−3.22, 8.89) .358 2.98 (−9.86, 15.83) .648 3.21 (−1.07, 7.48) .141 1.73 (.−.54, 3.99) .136 −5.15 (−13.21, 2.91) .210 −3.73 (−18.32, 10.86) .615
 Mixed use (z score) −1.25 (−7.35, 4.84) .686 −4.46 (−17.36, 8.44) .142 4.61 (−.00, .72) .050 2.84 (.39, 5.28) .023 3.15 (−4.99, 11.29) .447 −1.63 (−17.26, 13.98) .838
 Number parks (z score) −5.51 (−11.45, .42) .069 −.37 (−12.98, 12.24) .955 .53 (−3.98, 5.05) .816 .53 (−1.87, 2.93) .664 5.76 (−2.14, 13.66) .153 −12.60 (−27.35, 2.16) .094
 Cul-de-sac density (z score) −1.99 (−8.27, 4.28) .532 −5.44 (−18.68, 7.81) .420 −5.25 (−9.84, −.66) .025 −3.19 (−5.63, −.76) .010 −6.43 (−14.77, 1.91) .131 6.87 (−8.62 22.37) .383
 Objective neighborhood environment index (z score) −5.07 (−15.52, 5.38) .341 4.10 (−17.99, 26.19) .716 9.11 (1.59, 16.62) .018 4.90 (.91, 8.89) .016 −6.28 (−20.25, 7.69) .378 −6.75 (−32.38, 18.88) .605
Open in a new tab

Note.

a

Adjusted for participant MVPA (total or location-specific), location-specific wear time, gender, age, race/ethnicity, and BMI percentile, city, neighborhood income, vehicle access, and parent education and marital status.

b

Adjusted for location-specific wear time, gender, age, race/ethnicity, and BMI percentile, city, neighborhood income, vehicle access, and parent education and marital status.

Perceived index = average of 7 perceived variables; Objective index = average of objective variables, excluding cul-de-sac density.

Crime safety was marginally associated with total screen time (p = .053), such that parents who perceived greater crime safety had adolescents who spent fewer minutes in daily total screen time. A one standard deviation increase in perceived crime safety was associated with 12.82 fewer minutes of daily total screen time.

The perceived neighborhood index score was associated with less TV time, less total screen time, and less home neighborhood sedentary time. A one standard deviation increase in overall perceptions of activity-supportiveness of the neighborhood environment was associated with watching 13.60 fewer minutes of TV per day, engaging in 28.15 fewer minutes of total screen time per day, and spending 29.09 fewer minutes per day in home neighborhood sedentary time.

Objective Neighborhood Environment

Results pertaining to the objective neighborhood environment variables indicated that street connectivity was positively associated with total sedentary time and negatively associated with home neighborhood sedentary time, such that a one standard deviation increase in street connectivity (i.e., more intersections), was associated with 6.53 more minutes per day of total sedentary time and 10.90 fewer minutes per day of home neighborhood sedentary time. Mixed use was positively associated with home sedentary time, with a one standard deviation increase in the mixed use score being associated with 2.84 more minutes per day of home sedentary time. Mixed use was marginally associated with total sedentary time (p = .050), such that a one standard deviation increase in mixed use was associated with 4.61 more minutes per day of total sedentary time.

Cul-de-sac density was significantly negatively associated with total sedentary time, with a one standard deviation increase in cul-de-sac density relating to 5.25 fewer minutes per day of total sedentary time. Cul-de-sac density was also significantly negatively associated with home sedentary time, indicating that a one standard deviation increase in cul-de-sac density was associated with 3.19 fewer minutes per day of sedentary time at home.

The objective neighborhood environment index was significantly positively associated with total sedentary time and home sedentary time. A one standard deviation increase in the objective neighborhood environment index was associated with 9.11 more minutes per day of total sedentary time and 4.90 more minutes per day of sedentary time at home.

Discussion

A physical activity-supportive perceived neighborhood environment was related to less TV time, total screen time, and sedentary time in the home neighborhood, but not total sedentary time. Findings were more mixed with regards to the objective neighborhood environment, with mixed use and the neighborhood environment index being positively related to total and home sedentary time, cul-de-sac density being negatively related to total and home sedentary time, and street connectivity being negatively related to home neighborhood sedentary time but positively related to total sedentary time. The findings pertaining to perceived neighborhood aesthetics supported the primary hypothesis that favorable neighborhood environments for physical activity would be related to less time spent at home, and also less total sedentary time. However, overall there was limited support for this hypothesis. Findings for perceived aesthetics, perceived crime, the perceived neighborhood environment index, and cul-de-sacs density provided partial support for the hypothesis that neighborhood environments related to greater amounts of physical activity (Ding et al, 2011) would also relate to less sedentary time for youth, but cul-de-sac density was the only objective environment attribute related to total sedentary time in the expected direction.

The Perceived Neighborhood Environment and Sedentary Time

Better perceived aesthetics was related to less TV and total screen time. However, associations with TV and total screen time were strongest for the perceived neighborhood environment index, showing the importance of the combination of environment attributes. When parents have positive perceptions of attractive neighborhood environments, they may be more comfortable with their adolescent spending more time out of the home and away from non-portable electronic entertainment. Though the same associations were not found with the accelerometer measures of sedentary time, TV time and total screen time are important sedentary behaviors as they have been more consistently associated with health markers in youth (Tremblay et al., 2011), and are often targets for interventions.

Examining time spent at home as a dependent variable was one unique contribution of the present study. The significant association between neighborhood aesthetics and time spent at home provided some support for our primary hypothesis. The finding suggested that adolescents may be more likely to leave their home, and perhaps spend time outside, if their neighborhood is more aesthetically pleasing (e.g., trees along the streets, many beautiful natural sights). This pattern is consistent with previous findings that children engaged in significantly less total screen time when parents perceived their neighborhood as having access to good quality parks, playgrounds, and sidewalks (Carson et al., 2010). Thus, improving neighborhood aesthetics may be an important way to support adolescents to spend less time at home and being sedentary in front of a screen. Indeed, more time away from home may lead to less opportunity to engage in sedentary time, especially TV and total screen time in the home, and previous research shows that time at home is mostly inactive for adolescents (Carlson et al., 2016). Although the present study found small-to-no correlation between time spent at home and TV and total screen time, these activities were not measured with specificity to the home. The correlation between time spent at home and total sedentary time was larger and meaningful, but still small.

Although the association between greater perceived crime safety and less total screen time approached statistical significance, perceived safety was not associated with time at home. Other (albeit few) studies of associations between perceived neighborhood safety and total screen time have had null findings (Carson et al., 2010), so more research is needed in this area before drawing conclusions.

Although the perceived environment variables were not related to total or home sedentary time, two perceived environment variables (land use mix-access and walking facilities) were associated with less home neighborhood sedentary time, as was the perceived neighborhood environment index. These findings are likely at least partly explained by the neighborhood environment supporting more walking and less vehicle time, which is discussed more below in regards to objectively measured street connectivity.

The Objective Neighborhood Environment and Sedentary Time

Several associations between the objective neighborhood environment attributes and total and home sedentary time were in the unexpected direction, particularly for mixed land use, street connectivity, and the overall neighborhood environment index, indicating that these attributes may be related to both more sedentary time and more physical activity among adolescents (Ding et al, 2011). While mixed land use near the home provides destinations to walk to, sedentary time may be more likely than physical activity to occur inside these destinations. However, mixed land use was not associated with home neighborhood sedentary time, so the potential impact of mixed land use mix on sedentary time may be more complex. There could be moderating effects of other variables, such as aesthetics and perceived safety, that should be explored. The potential impact of street connectivity on sedentary time also appears complex. The negative (expected) association between street connectivity and home neighborhood sedentary time is likely at least partly explained by transportation mode. As shown previously in this sample, greater street connectivity was related to more walking and less vehicle time (Carlson et al., 2015). However, the drivers of the positive association between street connectivity and total sedentary time are less clear. Although street connectivity was not related to time spent at home, it is possible that connected streets with heavy traffic deter youth from spending time outdoors.

Previous evidence has shown that cul-de-sacs can be a setting for children’s leisure-time physical activity (Ding et al, 2011), although findings in adolescents have been somewhat mixed regarding associations between cul-de-sacs and physical activity (Carver et al., 2008). The present finding that cul-de-sacs were associated with less total and home sedentary time fits with the notion that cul-de-sacs provide a safe place for activity (both light and moderate-to-vigorous), potentially due to less traffic and greater perceived safety (Veitch et al., 2006). Thus, cul-de-sac density was the one objective environment attribute investigated in this paper that appears to support both increased total physical activity and reduced total sedentary time.

Integration of Findings

In the present study, the discrepant findings using perceived versus objective environment measures are notable. Perceived environment features positively correlated with physical activity in youth tended to be negatively correlated with sedentary time, as expected. The effect sizes were meaningful, with 14 to 29 fewer minutes of daily sedentary time per one standard deviation increase in the perceived neighborhood environment index. In contrast, with objective environment variables there were sometimes associations with sedentary time in the opposite direction of what was expected based on previous associations with physical activity. However, the effect sizes appeared to be much smaller for the objective environmental variables than perceived environment variables, with 5 to 9 more minutes per day of sedentary time per one standard deviation increase in the objective neighborhood environment index. The few studies that have examined associations between objectively measured neighborhood environments and sedentary time in youth have generally had null findings when sedentary time was measured objectively (Hinckson et al., 2017; McDonald et al., 2012; Timperio et al., 2017; Veitch et al., 2011), but some associations in expected directions have been observed for TV and total screen time (Veitch et al., 2011; Timperio et al., 2017). However, few past studies examined both perceived and objectively measured neighborhood environments in relation to sedentary time in youth (Hinckson et al., 2017; Jago et al., 2006; Veitch et al., 2011), so it is difficult to determine whether differential findings with perceived and objective environment measures generalize across studies. Some studies showing associations between neighborhood environments and sedentary time or TV/total screen time did not appear to adjust for MVPA (Hinckson et al., 2017; Jago et al., 2006; McDonald et al., 2012, Timperio et al., 2017; Veitch et al., 2011), so prior findings may simply show that an increase in MVPA offsets sedentary time, but not necessarily that neighborhood environments can decrease sedentary time over and above their influence on MVPA (i.e., increase light activity). An exception appears to be in regards to the impact of the neighborhood environment on transportation-related sedentary time, with previous studies finding negative associations between objectively measured neighborhood walkability and adolescents’ time spent in a vehicle (Carlson et al., 2015). The discrepancy in findings between perceived and objective neighborhood environments, and between objective sedentary time and TV/total screen time, suggests a complex and unclear role of the neighborhood environment on sedentary time in youth. Taking into account current findings about sedentary time and past findings about MVPA, it is possible that some aspects of the neighborhood environment, such as mixed land use, may support both more sedentary time and more MVPA (and perhaps less light intensity activity that is between sedentary and MVPA), so more work is needed to gain more insight into these associations.

Strengths, Limitations, and Future Directions

Strengths of the present study included the large sample of adolescents from two geographic regions in the US, and equal selection from census block groups representing high or low walkability and high or low income. This strategy maximized variation in walkability and minimized income as a potential confounder of the results. This study examined both perceived and objective environment variables and a wider range of sedentary time measures than prior studies. The use of GPS to define location-specific wear time, and the use of accelerometers to provide objective measurement of sedentary time were also a particular strengths of the study that added information beyond previous work in this area.

Findings of the present study must be considered in light of some limitations. The cross-sectional design of the study does not allow for causal inferences. Perceived neighborhood environments were assessed through parent report, and it is possible that adolescents’ perceptions of their neighborhood environment may reveal different associations. Given potential error in the GPS and the home buffer used some sedentary time classified as occurring in the neighborhood may have actually occurred at home. Future studies should use newer GPS devices that measure signal strength to support identification of indoor vs. outdoor time (Lam et al., 2013). The use of hip-worn accelerometer cut points for the purposes of this study is potentially limited because they do not specifically differentiate sitting from standing behavior (Mitchell et al., 2017). However, these methods have good validity for accurately assessing an individual’s volume of sedentary time in minutes (Evenson et al., 2008; Trost et al., 2011; Carlson et al., 2019), as was done in the present study. Lastly, the decision to exclude the school location from the neighborhood buffer was made in efforts to isolate the potential effects of the neighborhood environment on sedentary behavior from those that are dictated by being at school. However, there remains a possible limitation that participants spent time at school on the weekends, for example using a playground or field, that would not be captured.

In future research it may be important to focus on consistency of measurement across studies. For example, we found it useful to compare our findings of the perceived neighborhood environment to those of Hinckson et al. (2017), as they also used the NEWS-Y. Consistency of measurement methods may allow for more rapid development of consensus regarding results across multiple studies. Future research may also be improved by measuring domain- and/or location-specific sedentary behaviors, such as total screen time, riding in a car, or sitting at a restaurant, that could inform more targeted behavior- and setting-specific interventions to decrease sedentary time. Sedentary time could also be delineated by whether it occurs in a destination (and which type of destination) or during transportation, because the environments within specific destinations are likely important over and above the neighborhood environment. Future studies could examine whether there are differences in associations between the neighborhood environment and sedentary time across demographic subgroups (e.g., SES, gender, race/ethnicity), and whether these associations differ in relation to the home environment and/or other personal attributes (e.g., enjoyment of sedentary time).

Conclusions

The present study found that having more favorable perceived neighborhood aesthetics was associated with less TV and total screen time, partly by supporting adolescents to spend less time at home. Cul-de-sacs were also related to less total sedentary time, suggesting their potential role in supporting both light and moderate-to-vigorous physical activity among youth. However, mixed land use and the objective neighborhood environment index, which are commonly associated with greater physical activity were associated with more sedentary time in the present study, so they may support both increased physical activity and increased sedentary time. Other environments, such as the home and school environment, and availability of screen-based opportunities (Wiecha et al., 2001), may be more important targets for reducing sedentary time in youth. The apparent complexity of this research area makes it a good fit for studies that examine true multi-level ecological models (Sallis et al., 2015) by including variables at individual, social, and multiple environmental levels and investigating interactions among levels.

Highlights.

  • Neighborhood environment associations with sedentary time differed between perceived and objective measures

  • Perceived neighborhood aesthetics and an overall perceived index were related to less screen time

  • Objective mixed land use and an overall objective index were related to more total sedentary time

  • Cul-de-sac density was related to less home sedentary time and total sedentary time

Acknowledgments

Conflicts Of Interest

This study was funded by NIH grant HL083454. The authors have no conflicts of interest to disclose.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Carson V, Spence JC, Cutumisu N and Cargill L, 2010. Association between neighborhood socioeconomic status and screen time among pre-school children: a cross-sectional study. BMC Public Health, 10(1), p.367. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Carson V, Stone M and Faulkner G, 2014. Patterns of sedentary behavior and weight status among children. Pediatric exercise science, 26(1), pp.95–102. [DOI] [PubMed] [Google Scholar]
  3. Carlson JA, Saelens BE, Kerr J, Schipperijn J, Conway TL, Frank LD, Chapman JE, Glanz K, Cain KL, Sallis JF, 2015. Association between neighborhood walkability and GPS-measured walking, bicycling and vehicle time in adolescents. Health Place 32, 1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Carlson JA, Schipperijn J, Kerr J, Saelens BE, Natarajan L, Frank LD, Sallis JF, 2016. Locations of physical activity as assessed by GPS in young adolescents. Pediatrics 137 (1), e20152430. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Carver A, Timperio AF and Crawford DA, 2008. Neighborhood road environments and physical activity among youth: the CLAN study. Journal of Urban Health, 85(4), p.532. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Centers for Disease Control and Prevention. 2000. CDC Growth Charts for the United States: Methods and Development. Hyattsville, MD: Department of Health and Human Services; 2002. [Google Scholar]
  7. Chastin SF, De Craemer M, Lien N, Bernaards C, Buck C, Oppert JM, Nazare JA, Lakerveld J, O’Donoghue G, Holdsworth M and Owen N, 2016. The SOS-framework (Systems of Sedentary behaviours): an international transdisciplinary consensus framework for the study of determinants, research priorities and policy on sedentary behaviour across the life course: a DEDIPAC-study. International Journal of Behavioral Nutrition and Physical Activity. 13: 83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cliff DP, Hesketh KD, Vella SA, Hinkley T, Tsiros MD, Ridgers ND, Carver A, Veitch J, Parrish AM, Hardy LL and Plotnikoff RC, 2016. Objectively measured sedentary behaviour and health and development in children and adolescents: systematic review and meta-analysis. Obesity Reviews, 17(4), pp.330–344. [DOI] [PubMed] [Google Scholar]
  9. Compernolle S, De Cocker K, Mackenbach JD, Van Nassau F, Lakerveld J, Cardon G, & De Bourdeaudhuij I, 2017. Objectively measured physical environmental neighbourhood factors are not associated with accelerometer-determined total sedentary time in adults. International Journal of Behavioral Nutrition and Physical Activity, 14(1), 94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ding D, Sallis JF, Kerr J, Lee S, Rosenberg D, 2011. Neighborhood environment and physical activity among youth: a review. Am. J. Prev. Med 41 (4), 442–455. [DOI] [PubMed] [Google Scholar]
  11. Evenson KR, Catellier DJ, Gill K, Ondrak KS and McMurray RG, 2008. Calibration of two objective measures of physical activity for children. Journal of sports sciences, 26(14), pp.1557–1565. [DOI] [PubMed] [Google Scholar]
  12. Frank LD, Sallis JF, Saelens BE, Cain K, Conway TL, Hess PM, 2010. The development of a walkability index: application to the Neighborhood Quality of Life Study. Brit. J. Sport Med 44, 924–933. [DOI] [PubMed] [Google Scholar]
  13. Hinckson E, Cerin E, Mavoa S, Smith M, Badland H, Stewart T, Duncan S and Schofield G, 2017. Associations of the perceived and objective neighborhood environment with physical activity and sedentary time in New Zealand adolescents. International journal of behavioral nutrition and physical activity, 14(1), p.145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Jago R, Baranowski T and Baranowski JC, 2006. Observed, GIS, and self-reported environmental features and adolescent physical activity. American Journal of Health Promotion, 20(6), pp.422–428. [DOI] [PubMed] [Google Scholar]
  15. Koohsari MJ, Sugiyama T, Sahlqvist S, Mavoa S, Hadgraft N and Owen N, 2015. Neighborhood environmental attributes and adults' sedentary behaviors: review and research agenda. Preventive medicine, 77, pp.141–149. [DOI] [PubMed] [Google Scholar]
  16. Lam MS, Godbole S, Chen J, Oliver M, Badland H, Marshall SJ, Kelly P, Foster C, Doherty A and Kerr J, 2013, November Measuring time spent outdoors using a wearable camera and GPS. In Proceedings of the 4th International SenseCam & Pervasive Imaging Conference (pp. 1–7). ACM. [Google Scholar]
  17. Maitland C, Stratton G, Foster S, Braham R and Rosenberg M, 2013. A place for play? The influence of the home physical environment on children’s physical activity and sedentary behaviour. International Journal of Behavioral Nutrition and Physical Activity, 10(1), p.99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Mattocks C, Ness A, Leary S, Tilling K, Blair SN, Shield J, Deere K, Saunders J, Kirkby J, Smith GD and Wells J, 2008. Use of accelerometers in a large field-based study of children: protocols, design issues, and effects on precision. Journal of Physical Activity and Health, 5(s1), pp.S98–S111. [DOI] [PubMed] [Google Scholar]
  19. McDonald K, Hearst M, Farbakhsh K, Patnode C, Forsyth A, Sirard J and Lytle L, 2012. Adolescent physical activity and the built environment: a latent class analysis approach. Health & place, 18(2), pp. 191–198. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Mitchell TB, Borner KB, Finch J, Kerr J, Carlson JA, 2017. Using activity monitors to measure sit-to-stand transitions in overweight/obese youth. Med. Sci. Sports Exer 49 (8), 1592–1598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Norman GJ, Schmid BA, Sallis JF, Calfas KJ and Patrick K, 2005. Psychosocial and environmental correlates of adolescent sedentary behaviors. Pediatrics, 116(4), pp.908–916. [DOI] [PubMed] [Google Scholar]
  22. Owen N, Cerin E, Leslie E, Coffee N, Frank LD, Bauman AE, Hugo G, Saelens BE and Sallis JF, 2007. Neighborhood walkability and the walking behavior of Australian adults. American journal of preventive medicine, 33(5), pp.387–395. [DOI] [PubMed] [Google Scholar]
  23. Owen N, Healy GN, Matthews CE, and Dunstan DW, 2010. Too much sitting: the population-health science of sedentary behavior. Exercise and sport sciences reviews., 38(3), pp. 105–113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Physical Activity Guidelines for Americans. 2008; http://www.health.gov/paguidelines/. Accessed January 2, 2012.
  25. Proper KI, Singh AS, Van Mechelen W, & Chinapaw MJ (2011). Sedentary behaviors and health outcomes among adults: a systematic review of prospective studies. American journal of preventive medicine, 40(2), 174–182. [DOI] [PubMed] [Google Scholar]
  26. Rosenberg D, Ding D, Sallis JF, Kerr J, Norman GJ, Durant N, Harris SK and Saelens BE, 2009. Neighborhood Environment Walkability Scale for Youth (NEWSY): reliability and relationship with physical activity. Preventive medicine, 49(2-3), pp.213–218. [DOI] [PubMed] [Google Scholar]
  27. Saelens BE, Sallis JF, Frank LD, et al. , 2012. Neighborhood environmental and psychosocial correlates of adults' physical activity. Med. Sci. Sport Exer 44 (4), 637–646. [DOI] [PubMed] [Google Scholar]
  28. Sallis JF, Owen N, Fisher E, 2015. Ecological models of health behavior. In: Glanz K, Rimer BK, Viswanath K (Eds.), Health Behavior: Theory, Research, and Practice, 5th Edition. Jossey-Bass, San Francisco, CA, pp. 43–64. [Google Scholar]
  29. Sallis JF, Conway TL, Cain KL, Carlson JA, Frank LD, Kerr J, Glanz K, Chapman JE, Saelens BE, 2018. Neighborhood built environment and socioeconomic status in relation to physical activity, sedentary behavior, and weight status of adolescents. Prev. Med 110, 47–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Saunders TJ, Tremblay MS, Mathieu MÈ, Henderson M, O’Loughlin J, Tremblay A, Chaput JP and Quality Cohort Research Group, 2013. Associations of sedentary behavior, sedentary bouts and breaks in sedentary time with cardiometabolic risk in children with a family history of obesity. PloS one, 8(11), p.e79143. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Telama R (2009). Tracking of physical activity from childhood to adulthood: a review. Obesity facts, 2(3), 187–195. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Timperio A, Crawford D, Ball K and Salmon J, 2017. Typologies of neighbourhood environments and children's physical activity, sedentary time and television viewing. Health & place, 43, pp. 121–127. [DOI] [PubMed] [Google Scholar]
  33. Tremblay MS, LeBlanc AG, Kho ME, Saunders TJ, Larouche R, Colley RC, Goldfield G and Gorber SC, 2011. Systematic review of sedentary behaviour and health indicators in school-aged children and youth. International Journal of Behavioral Nutrition and Physical Activity, 8(1), p.98. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Treuth MS, Schmitz K, Catellier DJ, McMurray RG, Murray DM, Almeida MJ, Going S, Norman JE and Pate R, 2004. Defining accelerometer thresholds for activity intensities in adolescent girls. Medicine and science in sports and exercise, 36(7), p.1259. [PMC free article] [PubMed] [Google Scholar]
  35. Trost SG, Loprinzi PD, Moore R, & Pfeiffer KA (2011). Comparison of accelerometer cut points for predicting activity intensity in youth. Medicine and science in sports and exercise, 43(7), 1360–1368. [DOI] [PubMed] [Google Scholar]
  36. Van Ekris E, Altenburg TM, Singh AS, Proper KI, Heymans MW and Chinapaw MJM, 2016. An evidence-update on the prospective relationship between childhood sedentary behaviour and biomedical health indicators: a systematic review and meta-analysis. Obesity reviews, 17(9), pp.833–849. [DOI] [PubMed] [Google Scholar]
  37. Veitch J, Bagley S, Ball K and Salmon J, 2006. Where do children usually play? A qualitative study of parents’ perceptions of influences on children's active free-play. Health & place, 12(4), pp.383–393. [DOI] [PubMed] [Google Scholar]
  38. Veitch J, Timperio A, Crawford D, Abbott G, Giles-Corti B and Salmon J, 2011. Is the neighbourhood environment associated with sedentary behaviour outside of school hours among children? Annals of behavioral medicine, 41(3), pp.333–341. [DOI] [PubMed] [Google Scholar]
  39. Wiecha JL, Sobol AM, Peterson KE and Gortmaker SL, 2001. Household television access: associations with screen time, reading, and homework among youth. Ambulatory, 1(5), pp.244–251. [DOI] [PubMed] [Google Scholar]

RESOURCES