Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Sep 1.
Published in final edited form as: J Transp Health. 2018 Jun 8;10:304–314. doi: 10.1016/j.jth.2018.05.010

Latent profile analysis of young adolescents’ physical activity across locations on schooldays

Kelsey B Borner a, Tarrah B Mitchell b, Jordan A Carlson c,*, Jacqueline Kerr d, Brian E Saelens e, Jasper Schipperijn f, Lawrence D Frank g, Terry L Conway d, Karen Glanz h, Jim E Chapman i, Kelli L Cain d, James F Sallis d
PMCID: PMC6289522  NIHMSID: NIHMS975307  PMID: 30555787

Abstract

Purpose

To investigate whether adolescents cluster into profiles based on where they accumulate moderate-to-vigorous physical activity (MVPA), if overall MVPA differs across profiles, and if walking to school and participant and neighborhood characteristics explain profile membership.

Methods

Adolescents (N=528; mean age=14.12±1.44; 50% girls) wore accelerometers and Global Positioning Systems (GPS) trackers for 3.9±1.5 days to assess MVPA minutes in five locations: at home, at school, in home neighborhood, in school neighborhood, and other. Walking to school and participant characteristics were assessed by questionnaire, and neighborhood environment by Geographic Information Systems (GIS). Latent profile analysis (LPA) was used to identify profiles/groups of participants based on accumulation of physical activity across the five locations. Mixed-effects regression tested differences in overall MVPA, walking to school, and other characteristics across profiles.

Results

Four initial profiles emerged: one Insufficiently Active profile and three “Active” profiles (Active Around School, Active Home Neighborhood, and Active Other Locations). The Insufficiently Active profile emerging from the first LPA (90% of participants) was further separated into four profiles in a second LPA: Insufficiently Active, and three additional “Active” profiles (Moderately-Active Around School, Moderately-Active Home Neighborhood, and Active At Home). Those in the six Active profiles had more overall MVPA (41.1–92.7 minutes/day) than those in the two Insufficiently Active profiles (34.5–38.3 minutes/day). Variables that differed across profiles included walking to school, sports/athletic ability, and neighborhood walkability.

Conclusions

Although most participants did not meet the MVPA guideline, the six Active profiles showed the places in which many adolescents were able to achieve the 60-minute/day guideline. The home and school neighborhood (partly through walking to school), “other” locations, and to a lesser extent the home, appeared to be key sources for physical activity that distinguished active from insufficiently active adolescents. Finding the right match between the individual and physical activity source/location may be a promising strategy for increasing active travel and MVPA in adolescents.

Keywords: accelerometry, built environment, global positioning systems

1. INTRODUCTION

Because of the numerous physical and psychosocial health benefits of physical activity (Booth et al., 2012; Strong et al., 2005; Vaisto et al., 2014), the U.S. Department of Health and Human Services recommends that all children and adolescents achieve 60 minutes of moderate-vigorous physical activity (MVPA) per day (eDpartment of Health and Human Services, 2008). Adolescence marks a time of significant concern for inactivity, as research with objective measures suggests that only about 8% of adolescents meet the MVPA guidelines (Troiano et al., 2008). Additional research is needed to better understand MVPA patterns during this critical period of adolescence as a basis for improving physical activity promotion strategies and preventing the declining trajectory of MVPA into adulthood (Gordon-Larsen et al., 2004; Ortega et al., 2013; Telama, 2009).

Adolescents engage in MVPA through active transportation and recreation in multiple locations, including homes, schools, neighborhoods, and recreation areas (Grow et al., 2008). Active travel, in particular, has been related to more overall physical activity in numerous studies (Bassett et al., 2013; Cooper et al., 2005; Faulkner et al., 2009; Larouche et al., 2014). Previous studies that examined MVPA across locations in adolescence have done so by describing and comparing the amount of time spent in individual locations and by exploring how demographic and other explanatory variables are associated with MVPA in individual locations (Carlson et al., 2017; Maddison et al., 2010; Oreskovic et al., 2012; Rainham et al., 2012). These previous studies have relied on variable-centered analyses. However, variable-centered analyses do not allow for understanding of co-occurrence of physical activity across different locations within individuals. A complementary approach is to use person-centered analyses to identify subsets of adolescents based on their patterns of MVPA across locations. This approach allows for the identification of naturally occurring profiles/groups of adolescents based on where they accumulate MVPA and allows for the exploration of differences between these profiles. In particular, knowing whether patterns of MVPA accumulation across locations are related to overall MVPA, walking to school, and other participant and environment characteristics would inform settings-based and cross-setting (e.g., transportation) interventions.

The present study was among the first to use person-centered analyses (i.e., latent profile analysis) to identify naturally occurring patterns of young adolescents’ MVPA across locations using GPS and accelerometry. Study objectives were to investigate (1) whether youth clustered into profiles based on where they accumulated MVPA, and (2) whether overall MVPA, walking to school, and participant and neighborhood characteristics differed across profiles. Only schooldays were investigated because they account for most of the week and adolescents’ overall MVPA (Comte et al., 2013), likely involve distinct location-patterns of physical activity as compared with weekend days, and provide more heterogeneity in physical activity locations than provided by weekend days, particularly in regard to school physical activity, neighborhood activity, and active travel to school.

2. METHODS

2.1. Participants and Procedures

The current data were drawn from the Teen Environment and Neighborhood (TEAN) study of built environments and physical activity. This study was conducted in the Baltimore, MD/Washington, DC and Seattle/King County, WA regions of the USA from 2009 to 2011. A total of 928 healthy adolescents (ages 12–16 years), and one of their parents, participated in the larger TEAN study. TEAN participants were selected from 447 census block groups and were evenly distributed across four block group types of (high or low) walkability and (high or low) income, as assessed by previously reported methods (Frank et al., 2010; Sallis et al., 2018). Recruitment was conducted via mail and telephone. Ineligibility criteria included any condition affecting their physical activity, dietary habits, or ability to participate.

A subset of participants was instructed to wear a Global Positioning System (GPS) tracker and an accelerometer for seven days during all waking hours, except during times when the devices could get wet. From the initial 928 participants, a total of 400 were determined ineligible for present analyses and were excluded due to not being given a GPS device (N = 130), not wearing an accelerometer and a GPS simultaneously for at least one valid school day (N = 148), and geocoding errors (N = 122). Thus, the final sample for the current study included 528 adolescents.

2.2. Measures

Demographics

Participating adolescents self-reported their age, gender, and ethnicity. Parent participants self-reported their highest level of education. Ethnicity was represented as whether or not the participant identified as white (non-Hispanic). Parent education was represented as whether or not either parent/guardian had received a college degree. Using the methods for identifying neighborhoods to include in the study, the median annual household income (based on 2000 census) was deciled and then dichotomized by defining the 1st through 4th deciles as “low income” (Baltimore: $15,466 to $40,700; Seattle/King County: $25–431 to $49,012) and the 7th, 8th, and 9th deciles as “high income” (Baltimore: $62,450 to $86,932; Seattle/King County: $65,752 to $81,865). The highest decile of income was omitted to eliminate outliers with very high income and thus maximize generalizability.

Physical activity

Minutes of MVPA were assessed with ActiGraph accelerometers (Models 7164, 71256, GT1M, GT3X). Epoch length was set to 30-seconds, and MVPA was derived from the Evenson cut points for vertical axis acceleration counts (Evenson et al., 2008). Non-wear periods were considered >60 consecutive epochs (30 minutes) with count = 0, and such periods were excluded from analyses. Only school days were included in the analyses, defined as any weekday during which the participant spent ≥200 minutes of time at school based on the GPS. Minutes spent in MVPA were used for the current analyses. Participants were also categorized into meeting or not meeting the recommended 60 minutes/day of MVPA, calculated as whether or not participants’ average MVPA (across days) fell above or below this cutoff.

GPS tracking

Participants were instructed to wear a GlobalSat DG-1 GPS tracker (GlobalSat, New Taipei City, Taiwan), and latitude and longitude data were collected at 30-second epochs (Rainham et al., 2012). GPS and accelerometer data were merged using the Personal Activity and Location Measurement System (Center for Wireless & Population Health Systems, 2012; Center for Wireless and Population Health Systems, La Jolla, CA). Only combined GPS and accelerometer data could be included in the analysis. A minimum of eight hours of daily accelerometer wear time was required for the day to be included (Mattocks et al., 2008), and a minimum of one valid day for the participant to be included. All accelerometer wear time must have coincided with valid GPS data to be included. Participants who never left their home over the one-week period (as determined by the GPS) were considered to have not worn the device.

Locations of physical activity were categorized into the following: “home,” “home neighborhood,” “school,” “school neighborhood,” and “other” locations. Each participant’s home and school addresses were geocoded and integrated into ArcGIS (ESRI, Inc., Redlands, CA). Buffers around each address (home and school) created a radius to define each location of interest (at home: 50-m-radius circular buffer around home address geocoded point; at school: 15-m buffer around school parcel; near home: 1-km street network buffer around home address, excluding the “at-home” and “at-school” buffers; near school: 1-km street network buffer around school address, excluding the “at-home” and “at-school” buffers; and all other locations: any location not included in the previous four locations). The participant-specific location and GPS information were merged into a PostgreSQL database (PostgreSQL Global Development Group, Berkeley, CA) to analyze when and for how long participants were within any of the five locations. Number of minutes per day and number of minutes of MVPA spent in each location were then calculated. For participants whose home and school neighborhood overlapped (N = 110), the overlapping wear time and MVPA time were divided by 2 and split evenly across the two locations.

Neighborhood walkability

A walkability index was derived in GIS for a 1km street-network buffer around each participant’s home neighborhood, using measures of net residential density, road intersection density, mixed land use, and pedestrian design of retail space (i.e., the ratio of retail building square footage divided by retail land square footage; a lower ratio often reflects a higher amount of parking, which decreases likelihood of pedestrian access; Frank et al., 2010). The four variables were standardized to have a mean of 0 and standard deviation of 1 (i.e., z score), based on the neighborhoods selected for recruitment, and summed so each would contribute equal weight to the index. Walkability scores were deciled and categorized by median split procedures to represent lower or higher walkability.

Park access

Data from the county tax assessor on land use was integrated into GIS to determine the presence or absence of at least one park within a 500-meter street network buffer around the participant’s home (yes/no). The smaller 500-meter buffer was used for park access because prior studies have found stronger associations with MVPA at this buffer size (Sallis et al., 2016), and a majority of adolescents had parks within 1km buffers, providing limited variability.

Neighborhood safety

Neighborhood safety was assessed via the parent-report Neighborhood Environment Walkability Scale for Youth (NEWS-Y; Rosenberg et al., 2009). Traffic safety (3 items; e.g., traffic makes it unpleasant to walk), pedestrian safety (3 items; e.g., crosswalks on busy streets), and crime safety (5 items; e.g., high crime rate) items were averaged into individual subscale scores for analyses. Previous investigations have demonstrated adequate test-retest reliability (ICCs = 0.61–0.78; Rosenberg et al., 2009).

Parent physical activity rules

Parents reported on 18 possible rules (yes/no) they enforced regarding their child’s participation in physical activity and sedentary activity (Tandon et al., 2012), and the 11 items expected to restrict adolescent physical activity were summed for present analyses1 (e.g., do not go places alone, do not cross busy streets, do homework before going out, stay in the neighborhood).

Distance to school

GIS data were used to calculate the shortest street-network distance from home to school in meters.

Regular walking to school

An adaptation of a measure from the Centers for Disease Control and Prevention’s Kids-Walk-to-School program (Carlson, Sallis, et al., 2014) was used, which asked adolescents to report both the number of days they travelled to, and from, school by walking in a typical school week. Previous investigations have determined adequate test-retest reliability of this measure (Joe et al., 2012; Timperio et al., 2006). Participants were dichotomized into regular walking (5–10 trips per week) or not (<5 trips per week) to capture a level of regular walking that is achievable for youth in the US, and contributes meaningfully to overall physical activity.

Involvement in team sports/structured activities

Participants self-reported the number of sports teams or after school physical activity classes they participated in over the past year. Individual items assessed participation at school and outside of school. Item responses were transformed from count to any sports participation (i.e., yes/no) for current analyses.

Athletic ability

Participants self-reported their perceived athletic ability (“How do you rate your athletic ability compared to your peers?”) on a scale of 1–5 (much lower (1), somewhat lower (2), same level (3), somewhat higher (4), much higher (5)).

2.3. Data Analyses

Latent profile analysis (LPA) was conducted with five participant-level indicator variables representing MVPA minutes/day in each of five locations (i.e., home, home neighborhood, school, school neighborhood, and other locations). Models were compared to derive the number of profiles that best fit the data, taking into consideration the Lo–Mendell–Rubin test (LMR), McLachlan’s bootstrapped likelihood ratio test (BLRT) test, change in Bayesian Information Criterion (BIC), entropy values, group sample sizes, and interpretability of the profiles. Due to the large proportion of participants in one profile from the latent profile analysis, a second latent profile analysis was conducted on the participants in this large profile. Differences in demographic, MVPA (total and in each location), and other key variables were explored across the emerging profiles by conducting mixed effects linear regression analyses with dummy variables to compare each profile. These models were estimated in Mplus version 6, with block group identifier entered as a cluster variable to account for the nested design and using the MLR estimator which is robust to non-normal distributions. Between-group differences with a p value < 0.01 were considered significant.

3. RESULTS

3.1 Demographics

Participants had a mean age of 14.12 (SD = 1.44), 50% were girls, 69% were White non-Hispanic, 46% lived in a high-walkability neighborhoods, 50% in high-income neighborhoods, and 52% resided in the Seattle/King County, WA region. Participants wore the acceleromter and GPS devices for a mean of 3.9 (SD = 1.5) school days.

3.2 Latent Profile Analysis

The BLRT (p < 0.001) and reduction in BIC supported a 4-profile over 3-profile solution. Although the model fit estimates indicated a 5-profile solution was also possible, the improvement in BIC was very small (0.67%). Thus, the 4-profile solution was retained and an additional LPA was conducted on the largest profile, which captured 90% of participants. Similar to the first LPA, the model fit indices supported a 4- vs. 3-profile solution. BLRT (p < 0.001), but not LMR (p = 0.380), supported a 5- vs. 4-profile solution. The 4-profile solution was chosen because it produced reasonable sample sizes per class and interpretable profiles. Entropy for the first LPA (4-profile solution) was 0.99, and 0.94 for the second LPA (4-profile solution), indicating a clear delineation of classes (values closest to 1 are optimal).

3.2.1 First Latent Profile Analysis

The first latent profile analysis produced four profiles of activity locations (Figure 1). Adolescents in all four profiles had similar amounts of MVPA during the school day, ranging from 20.0 to 30.3 minutes/day, and in the home (4.9 – 5.7 minutes/day). Adolescents in the Insufficiently Active profile (N = 475, 90%) generally had low amounts of MVPA in each of the five locations. The remaining three profiles were considered the “Active” profiles and each was distinguished by higher amounts of MVPA in one location (“Active Around School”, N = 17, 3%; “Active Home Neighborhood”, N = 14, 3%; and “Active Other Locations”, N = 22, 4%) as compared with the other profiles. Adolescents in the three Active profiles had substantially higher amounts of overall MPVA (70.2 – 92.7 minutes/day for the three Active profiles vs. 38.3 minutes/day for Insufficiently Active profile) and were substantially more likely to meet the 60-minutes/day MVPA guideline (59% – 86% for Active profiles vs. 12% for Insufficiently Active profile; Table 1) than adolescents in the Insufficiently Active profile.

Figure 1.

Figure 1

Open in a new tab

Four latent profiles of patterns of phsyical activity by location from the first latent profile analysis (N = 528 adolescents)

Note: MVPA = moderate to vigorous physical activity

Table 1.

Differences in physical activity, demographics, and other physical activity-related factors across profiles in the first latent profile analysis (N = 528 adolescents)

Insufficiently Active Active Around School Active Home Neighborhood Active Other Locations
N = 475 (90%) N = 17 (3%) N = 14 (3%) N = 22 (4%)
Mean or % (SE) Significant differencesa Mean or % (SE) Significant differencesa Mean or % (SE) Significant differencesa Mean or % (SE) Significant differencesa
Physical activity by location
 Home MVPA (min/day) 5.5 none 5.7 none 5.5 none 4.9 none
(0.3) (1.0) (3.6) (1.0)
 Home neighborhood MVPA (min/day) 4.1 3 8.0 3 50.3 1, 2, 4 5.1 3
(0.2) (2.1) (3.7) (1.0)
 School MVPA (min/day) 23.0 none 25.9 none 30.3 none 20.0 none
(0.7) (3.4) (3.6) (1.7)
 School neighborhood MVPA (min/day) 1.6 2 20.3 1, 3, 4 2.4 2 3.5 2
(0.1) (1.2) (0.9) (0.9)
 Other location MVPA (min/day) 4.0 4 10.3 4 4.2 4 38.1 1, 2, 3
(0.2) (3.1) (1.7) (2.7)
Overall physical activity
 Meeting 60 min/day guideline (yes) 12% 2, 3, 4 59% 1 86% 1 82% 1
(1.5) (12.0) (8.8) (8.3)
 Overall MVPA (min/day) 38.3 2, 3, 4 70.2 1 92.7 1 71.7 1
(0.9) (6.2) (5.8) (2.9)
Demographics
 Age (years) 14.2 none 14.2 none 13.4 none 13.8 none
(0.1) (0.4) (0.4) (0.3)
 Female (yes) 52% 3 41% none 21% 1 41% none
(2.4) (12.0) (9.3) (10.6)
 White non-Hispanic (yes) 71% none 53% none 57% none 64% none
(2.3) (12.1) (13.0) (10.4)
 Parent with college degree (yes) 64% none 82% none 50% none 59% none
(2.4) (9.3) (13.4) (10.6)
 High income (yes) 50% none 41% none 57% none 55% none
(3.6) (12.0) (12.1) (10.8)
Environment
 High walkability (yes) 44% none 77% none 57% none 55% none
(3.6) (10.3) (14.6) (10.8)
 Park access (yes) 35% none 41% none 50% none 41% none
(2.8) (12.0) (15.2) (10.6)
 Traffic safety [1–4] 2.6 none 2.6 none 2.6 none 2.6 none
(0) (0.2) (0.2) (0.1)
 Pedestrian safety [1–4] 2.8 none 2.8 none 2.9 none 2.8 none
(0) (0.2) (0.1) (0.1)
 Crime safety [1–4] 3.0 none 3.1 none 3.2 none 3.0 none
(0) (0.2) (0.3) (0.2)
 Parent Rules [0–11] 4.1 none 3.8 none 4.2 none 3.8 none
(0.1) (0.6) (0.6) (0.4)
Travel to school
 Distance to school (km) 5.4 none 5.5 none 5.1 none 7.4 none
(0.3) (1.7) (0.8) (1.4)
 Regular walking to school (yes) 18% 2 47% 1, 3 7% 2 23% none
(1.8) (12.1) (6.4) (9.1)
 Sports
 Any school sports (yes) 69% none 77% none 79% none 86% none
(2.1) (10.3) (12.9) (7.4)
 Any sports outside school (yes) 68% 4 88% none 79% none 91% 1
(2.2) (7.8) (13.7) (6.2)
 Athletic ability [1–5] 3.4 4 3.8 none 3.7 none 3.9 1
(0) (0.2) (0.4) (0.2)
Open in a new tab

Note:

a

p < 0.01

Demographics, neighborhood characteristics, school travel, and rates of sports/activity class participation differed among profiles. Participants in Active Home Neighborhood were less likely to be female (21%) compared with participants in Insufficiently Active; (52%). Active Around School was distinguished by higher levels of regular walking to school (47%) when compared with Insufficiently Active (18%) and Active Home Neighborhood (7%). Although only significant at p < 0.05, Active Around School also demonstrated higher levels of neighborhood walkability (77% for Active Around School vs. 44% – 57% for Insufficiently Active, Active Home Neighborhood, and Active Other Locations). When compared with Insufficiently Active, adolescents in Active Other Locations self-reported higher levels of outside-school sport participation (91% for Active Other Locations vs. 68% for Insufficiently Active) and athletic ability (3.9 [out of 5] for Active Other locations vs. 3.4 for Insufficiently Active; Table 1).

3.2.2 Second Latent Profile Analysis

A second LPA was conducted on the participants from the Insufficiently Active profile of the original LPA. Four additional profiles emerged (Figure 2). In terms of MVPA locations, two profiles were similar to Active Around School and Active Home Neighborhood from the first latent profile analysis, and were therefore distinguished with the “moderate” characterization in each title. Adolescents in the second Insufficiently Active profile (N = 353, 74%) had low amounts of MVPA (<5 minutes in 4 of the 5 locations) across the locations. The remaining three Active profiles were characterized by higher amounts of MPVA in one location (“Moderately-Active Around School”, N = 39, 8%; “Moderately-Active Home Neighborhood,” N = 48, 10%; and “Active At Home”, N = 35, 7%). Adolescents in the three Active profiles had higher amounts of overall MVPA (41.1 – 62.6 minutes/day for Active Profiles vs. 34.5 minutes/day for Insufficiently Active). Only adolescents in Active At Home were more likely to meet the 60-minutes/day MVPA guideline (48.6% for Active At Home vs. 7% – 23% for Insufficiently Active, Moderately-Active Around School, and Moderately-Active Home Neighborhood; Table 2).

Figure 2.

Figure 2

Open in a new tab

Four latent profiles of patterns of phsyical activity by location from the second latent profile analysis (N = 475 adoloscents from Profile 1 in the first latent profile analysis)

Note: MVPA = moderate to vigorous physical activity

Table 2.

Differences in physical activity, demographics, and other physical activity-related factors across profiles in the second latent profile analysis (N = 475 adolescents from Profile 1 in the first latent profile analysis)

Insufficiently Active Moderately-Active Around School Moderately-Active Home Neighborhood Active At Home
N = 353 (75%) N = 39 (8%) N = 48 (10%) N = 35 (7%)
Mean or % (SE) Significant differencesa Mean or % (SE) Significant differencesa Mean or % (SE) Significant differencesa Mean or % (SE) Significant differencesa
Physical activity by location
 Home MVPA (min/day) 4.2 4 5.1 4 3.1 4 22.9 1, 2, 3
(0.2) (0.7) (0.7) (1.2)
 Home neighborhood MVPA (min/day) 2.5 2 ,3 5.2 1, 3 15.6 1 ,2, 4 3.4 3
(0.1) (0.6) (0.6) (0.5)
 School MVPA (min/day) 23.2 3 23.0 none 17.2 1, 4 28.9 3
(0.8) (2.1) (1.7) (2.6)
 School neighborhood MVPA (min/day) 1.0 2 8.3 1, 3, 4 0.9 2 1.8 2
(0.1) (0.3) (0.2) (0.3)
 Other location MVPA (min/day) 3.6 none 6.0 none 4.4 none 5.6 none
(0.2) (0.9) (0.7) (1.1)
Overall physical activity
 Meeting 60 min/day guideline (yes) 7.1% 4 23.1% none 8.3% 4 48.6% 1, 3
(1.4) (6.8) (4.1) (8.0)
 Overall MVPA (min/day) 34.5 2, 3, 4 47.5 1, 4 41.1 1, 4 62.6 1, 2, 3
(0.9) (2.9) (2.0) (3.3)
Demographics
 Age (years) 14.2 none 14.2 none 14.3 none 13.6 none
(0.1) (0.2) (0.2) (0.2)
 Female (yes) 57% 4 44% none 40% none 34% 1
(2.8) (8.0) (7.1) (7.5)
 White non-Hispanic (yes) 71% none 74% none 65% none 77% none
(2.6) (7.2) (7.5) (6.6)
 Parent with college degree (yes) 62% 2 82% 1 58% none 74% none
(2.8) (6.2) (7.7) (7.7)
 High income (yes) 50% none 46% none 58% none 49% none
(3.9) (8.0) (8.2) (9.5)
Environment
 High walkability (yes) 44% none 54% none 42% none 40% none
(3.8) (8.0) (8.7) (9.3)
 Park access (yes) 35% none 33% none 40% none 34% none
(2.9) (7.6) (8.8) (9.1)
 Traffic safety [1–4] 2.6 none 2.6 none 2.6 none 2.7 none
(0.0) (0.1) (0.1) (0.1)
 Pedestrian safety [1–4] 2.8 none 2.8 none 2.7 none 2.9 none
(0.0) (0.1) (0.1) (0.1)
 Crime safety [1–4] 3.0 none 2.9 none 3.1 none 3.1 none
(0.0) (0.1) (0.1) (0.1)
 Parent Rules [0–11] 4.3 none 3.6 none 4.0 none 3.7 none
(0.1) (0.4) (0.3) (0.4)
Travel to school
 Distance to school (km) 5.7 none 4.4 none 4.8 none 4.2 none
(0.3) (0.8) (0.4) (0.7)
 Regular walking to school (yes) 15% 2 44% 1, 3 15% 2 24% none
(0.0) (8.3) (4.7) (7.3)
Sports
 Any school sports (yes) 70% none 72% none 58% none 77% none
(2.4) (7.5) (7.3) (7.9)
 Any sports outside school (yes) 67% none 77% none 60% none 80% none
(2.6) (6.8) (7.2) (7.7)
 Athletic ability [1–5] 3.4 none 3.5 none 3.4 none 3.6 none
(0.1) (0.1) (0.2) (0.2)
Open in a new tab

Note:

a

p < 0.01

Participant and environmental variables were also associated with membership in each of the four profiles from the second LPA. Adolescents in Active At Home were less likely to be female than adolescents in Insufficiently Active (34% vs. 57%), and adolescents in Moderately-Active Around School were more likely than adolescents in Insufficiently Active to have parents with a college degree (82% vs. 62%). Adolescents in Moderately-Active Around School had the highest levels of MVPA in the school neighborhood (8.3 minutes/day) and also showed higher levels of walking to school (44.4% for Moderately-Active Around School vs. 15% for both Insufficiently Active and Moderately-Active Home Neighborhood), as well as a trend toward significance for higher levels of walkability (54% for Moderately-Active Around School vs. 40% – 44% for Profiles Insufficiently Active, Moderately-Active Home Neighborhood, and Active At Home). Adolescents in Active At Home were characterized by being active at home (22.9 minutes/day).

4. DISCUSSION

Present findings showed that young adolescents can be grouped into multiple clusters/profiles based on where they accumulated physical activity on school days across five frequently encountered locations. Each four-profile model fit the data better than a model with fewer profiles. However, the great majority of adolescents were grouped into the Insufficiently Active profile/s, and few were grouped into one of the Active profiles. Overall MVPA was substantially higher in the six Active profiles as compared with the two Insufficiently Active profiles, showing there were multiple ways and locations in which adolescents were able to achieve sufficient amounts of MVPA.

This study was among the first to show that location-based patterns of physical activity relate to overall MVPA and active transportation. Each profile of more-active adolescents was characterized by particularly high amounts of MVPA in one location type. Adolescents in profiles with high amounts of MVPA in the home and school neighborhoods, other locations, and to a lesser extent the home location, were about five-eight times more likely to meet the 60-minute/day MVPA guideline (US Department of Health and Human Services, 2008) than those in the Insufficiently Active profiles. These findings suggest that interventions promoting activity in home neighborhoods, school neighborhoods, and other locations are most promising for increasing overall physical activity, but there should be increased focus on finding the right match between the individual and physical activity source/location. For example, supporting active travel and school neighborhood activity may be most beneficial for those who are able to walk to/from school. Additionally, supporting sports participation in "other locations" may be most beneficial for those who enjoy and have perceived competence for athletics.

The profile with the highest (86%) rates of meeting the 60-minute/day physical activity guideline was the Active Home Neighborhood profile. Neighborhood walkability and sports participation were somewhat, but not significantly, higher in this profile than the Insufficiently Active profile, so these factors may explain some, but not all, of the substantial amount of MVPA (mean = 50.3 minutes/day) these participants obtained in the home neighborhood. It is likely this profile is largely explained by neighborhood play, which has been shown to be an important contributor to overall MVPA in other studies (Grow et al., 2008; Veitch et al., 2006). A similar profile emerged, Moderately-Active Home Neighborhood, but with lower overall MVPA, further supporting the importance of neighborhood-based activity for increasing overall MVPA in adolescents. Although walkability was not significantly more favorable in these profiles, other aspects of the neighborhood environment related to “playability” may be important and thus should be investigated in future research, such as cul-de-sacs, meadows, treed areas, yards at home, low speed roads, recreational facilities, and traffic calming features (Janssen & Rosu, 2015).

Those in the Active Other Locations profile had the second highest (82%) rates of meeting the 60-minute/day physical activity guideline. This profile was characterized by high rates of sports participation both at school and outside of school, and the highest self-reported athletic ability. Most (41%, 38.1 minutes/day) of participants’ overall MVPA in this profile occurred in other locations, potentially sports and recreation facilities. Previous studies using self-report showed that primary “other” locations for physical activity in adolescents included parks, open spaces, and sports fields/courts (Grow et al., 2008). Present findings indicated that for some adolescents, sports participation may lead to large amounts of total physical activity. Unfortunately, the number of adolescents in this profile was small (N = 22; 4%), which speaks to the fact that although a large number of youth participate in sports (Leek et al., 2011), the amount of physical activity accrued during sports is often lower than expected (National Council of Youth Sports, 2008). Further, barriers such as drop out, gender bias, cost, and exclusiveness can inhibit sports particiption, particularly for the adolescent age group (Kanters et al., 2008; Kanters et al., 2013; Sirard et al., 2006). On the other hand, higher levels of sport availability, higher quality of sport management, and greater availability of sport grounds at school (Mandic et al., 2012) were associated with greater sports participation. Other internal characteristics, including self-efficacy for physical activity, have been associated with higher participation in sports, and improving confidence to be active should be a focus of attention, particularly among girls (Barr-Anderson et al., 2007).

The profile with the third highest (59%) rates of meeting the 60-minute/day physical activity guideline was the Active Around School profile. This profile was largely explained by active travel to and from school, which was higher in this profile than the other profiles. This finding is in agreement with previous studies showing that youth who engage in active travel obtain more overall MVPA than their counterparts (Bassett et al., 2013; Cooper et al., 2005; Faulkner et al., 2009; Larouche et al., 2014). Walkability was also higher in this profile, but the differences were not significant at the p < 0.01 level. Participants in this profile obtained a high amount of phyical activty in the school neighborhood, but this group also tended to be more activity in the home neighborhood and other locations than those in the other profiles. The Moderately-Active Around school profile was very similar to the Active Around School profile, except the amounts of physical activity in each location and overall were somewhat lower. Active travel to school has been associated with higher amounts of overall MVPA in multiple studies (Faulkner et al., 2009). Although the number of adolescents in this profile was small (N = 17; 3%), supporting active travel to school by improving neighborhood built environments and adopting Safe Routes to School programs (McDonald et al., 2014; Stewart et al., 2014) may expand opportunities for daily activity through active transport.

The fourth highest (49%) rates of meeting the 60-minute/day physical activity guideline were observed for the Active At Home profile. This profile was not explained by neighborhood characteristics, but it did have slightly higher levels of sports participation and athletic ability than most other profiles. Adolescents in this profile likely engaged in larger amounts of active play inside and directly outside of the home than those in other profiles. Home and neighborhood play are particularly important sources of physical activity for children, but home-based recreational play activities (e.g, playground activities, informal sports, backyard games) typically decline during adolescence (Holt et al, 2008). Thus, encouraging adolescents to engage in more physically demanding activites at home (e.g., active video games, family-based active games, home exercise equipment, driveway basketball) in place of sedentary activities may support continued activity at home.

The purpose of conducting the second latent profile analysis was to determine whether additional profiles would emerge from the large (N = 475) Insufficiently Active profile from the first analysis. Two of the emerging profiles, (Moderately-Active Around School and Moderately-Active Home Neighborhood), were similar to profiles from the first analysis, but with lower amounts of physical activity, and similar participant and environment characteristics explained these profiles. Thus, the validity of these profiles was supported through replication, providing additional rationale for the importance of school neighborhood and home neighborhood-based activity to increasing overall physical activity in youth. Findings suggested that most adolescents did not engage in large amounts of school or home neighborhood phyical activity, but a larger proportion of school and home neighbrhood time was spent physically active as compared with most other locations. Youth who engage in school or home neighborhood activity typically have substantially more overall activity than their counterparts (Carlson et al., 2016; Kneeshaw-Price et al., 2013).

Although differences in school-based MVPA across profiles were smaller than differences in MVPA for each of the other locations, the difference was still meaningful (up to about 10 minutes/day). It should be noted that school MVPA was not a primary explanatory variable of profile membership. Previous literature showed the importance of school-based physical activity to overall MVPA and health (Pate et al., 2006), and it is clear that student physical activity can vary drastically across schools (Carlson, Mignano, et al., 2014; Carlson et al., 2013) Supporting more physical activity in the school setting may also help more youth achieve guidelines for physical activity. Such initiatives may include Coordinated School Physical Activity Programs (CSPAP), increased time in physical education, opportunities for activity during lunch-time, and integration of physical activity into classroom lessons (Carson et al., 2014; Centers for Disease Control and Prevention, 2013; Hills et al., 2015). However, an implication of the present study is that there is likely an upper limit with regards to how much MVPA schools can provide, and outside opportunities such as active travel, active play, and sports/active recreation appear to be critical for supporting youth to meet physical activity guidelines.

4.1. Strengths and Limitations

Study strengths included use of accelerometers and GPS to obectively assess physical activity in different locations. The purposeful inclusion of both high and low walkable home neighborhoods improved our ability to assess the role of neighborhood factors. Among the limitations, only weekdays were investigated, so findings can not be generalized to weekend days or summers. Weekend activities may change the number of participants meeting guidelines (Comte et al., 2013) and may change the importance of certain locations (e.g., no/limited physical activity at school on weekends; Carlson et al., 2017). Location-patterns of physical activity should be investigated on weekend days, and more specific examination of locations within the “other” category would likely be needed. Based on how the locations were classified, there is little understanding of where adolescents were active when they were in “other” locations. In addition, GPS signals can be unreliable when indoors. Specifically, misclassification may have occurred, such as discriminating in-home from home neighborhood and in-school from school neighborhood, although the PALMS system allowed “smoothing” to reduce such interference. Additional limitations included small sample size in some profiles, which limited power to detect meaningful differences between profiles (e.g., neighborhood walkability). Also, the TEAN study design was limited to adolescents living in urban/suburban neighborhoods with either higher or lower walkability and, therefore, may not represent the all types of neighborhoods in which adolescents live. Finally, the current study was observational, which limits the ability to determine causality.

4.2. Implications for policy and practice

The emerging profiles, each characterized by a different primary location of physical activity, demonstrated there are multiple ways to support increases in overall physical activity in adolescents. It was clear that adolescents characteristically were active in a limited number of locations during the week. Although the highly active profiles included only a small number of participants, even moving from the insufficiently active profile to one of the “moderate activity” profiles could lead to meaningful increases in overall physical activity.

Increasing neighborhood recreational activity and active travel appear to be particularly promising approaches for impacting population health. Multiple strategies for increasing overall physical activity are likely needed, including adoption of complete streets policies (Smart Growth America), which require designing streets to meet the needs of all users, including pedestrians, bicyclists, public transport users, and people with disabilities (e.g., buffer between sidewalks and street, protected bicycle facilities, curb ramps, and traffic calming), and Safe Routes to School programs (McDonald et al., 2014; Stewart et al., 2014), which support youth to walk to/from school. In particular, finding the right match between the individual and physical activity source/location should be considered, such as investing in active travel to school promotion for students from schools in walkable neighborhoods, promoting home and home-neighborhood activity among youth who spend large amounts of time at home, and increasing access to sports and recreation facilities to accommodate adolescents interested in sports. However, home neighborhood-related policies such as complete streets, mixed use development, and park enhancements are challenging because local leaders can be lacking and progress may be slow. Lower-cost strategies such as ‘playstreets’ may be effective in dedicating more neighborhood space to play (D'Haese et al., 2015).

4.3. Conclusions

On school days, adolescents accrued physical activity in multiple locations. The large majority of adolescents did not accrue sufficient physical activity to achieve guidelines. However, adolescents who had a profile of physical activity that involved high levels of activity to and around school, in the home neighborhood, in “other” locations, or to a lesser extent at home, were most likely to meet physical activity guidelines. Active travel to school, home neighborhood activity/play, and sports participation appeared to be key characteristics that explained the active profiles. Health promotion efforts targeting the aformentioned locations and characteristics have strong potential to improve overall physical activity in youth. Tailored strategies may be needed to find the best match between the individual and source/location of physical activity, because there appear to be multiple locations that can lead to sufficient overall physical activity. A next step would be to use findings from the present study to create interventions tailored to location characteristics and adolescent preferences to improve physical activity outcomes.

Supplementary Material

1

HIGHLIGHTS.

  • Patterns of physical activity across locations were related to overall activity.

  • Main locations for physical activity varied across adolescent clusters.

  • Neighborhood activity and active travel were key distinguishers of active adolescents.

  • Identifying the best location-person match could support targeted interventions.

Acknowledgments

Funding: This work was supported by NIH grant HL083454

Footnotes

1

The full 18-item Parent Physical Activity Rules measure is included as on online supplement to this manuscript.

Conflict of interest

None

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. Barr-Anderson DJ, Young DR, Sallis JF, Neumark-Sztainer DR, Gittelsohn J, Webber L, … Jobe JB. Structured physical activity and psychosocial correlates in middle-school girls. Prev Med. 2007;44(5):404–409. doi: 10.1016/j.ypmed.2007.02.012. [DOI] [PubMed] [Google Scholar]
  2. Bassett DR, Fitzhugh EC, Heath GW, Erwin PC, Frederick GM, Wolff DL, … Stout AB. Estimated energy expenditures for school-based policies and active living. American Journal of Preventive Medicine. 2013;44(2):108–113. doi: 10.1016/j.amepre.2012.10.017. [DOI] [PubMed] [Google Scholar]
  3. Booth FW, Roberts CK, Laye MJ. Lack of exercise is a major cause of chronic diseases. Compr Physiol. 2012;2(2):1143–1211. doi: 10.1002/cphy.c110025. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Carlson JA, Mignano AM, Norman GJ, McKenzie TL, Kerr J, Arredondo EM, … Sallis JF. Socioeconomic disparities in elementary school practices and children's physical activity during school. American Journal of Health Promotion. 2014;28(3 Suppl):S47–53. doi: 10.4278/ajhp.130430-QUAN-206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Carlson JA, Mitchell TB, Saelens BE, Staggs VS, Kerr J, Frank LD, … Sallis JF. Within-person associations of young adolescents' physical activity across five primary locations: is there evidence of cross-location compensation? Int J Behav Nutr Phys Act. 2017;14(1):50. doi: 10.1186/s12966-017-0507-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Carlson JA, Sallis JF, Kerr J, Conway TL, Cain K, Frank LD, Saelens BE. Built environment characteristics and parent active transportation are associated with active travel to school in youth age 12–15. British Journal of Sports Medicine. 2014;48(22):1634–1639. doi: 10.1136/bjsports-2013-093101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Carlson JA, Sallis JF, Norman GJ, McKenzie TL, Kerr J, Arredondo EM, … Saelens BE. Elementary school practices and children's objectively measured physical activity during school. Preventive Medicine. 2013;57(5):591–595. doi: 10.1016/j.ypmed.2013.08.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Carlson JA, Schipperijn J, Kerr J, Saelens BE, Natarajan L, Frank LD, … Sallis JF. Locations of Physical Activity as Assessed by GPS in Young Adolescents. Pediatrics. 2016;137(1):1–10. doi: 10.1542/peds.2015-2430. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Carson RL, Castelli DM, Beighle A, Erwin H. School-based physical activity promotion: a conceptual framework for research and practice. Child Obes. 2014;10(2):100–106. doi: 10.1089/chi.2013.0134. [DOI] [PubMed] [Google Scholar]
  10. Center for Wireless & Population Health Systems. Physical Activity Location Measurement System. 2012 Retrieved from http://cwphs.ucsd.edu/palms.
  11. Centers for Disease Control and Prevention. Comprehensive School Physical Activity Programs: A Guide for Schools. Atlanta, GA: US Department of Health and Human Services; 2013. [Google Scholar]
  12. Comte M, Hobin E, Majumdar SR, Plotnikoff RC, Ball GD, McGavock J. Patterns of weekday and weekend physical activity in youth in 2 Canadian provinces. Appl Physiol Nutr Metab. 2013;38(2):115–119. doi: 10.1139/apnm-2012-0100. [DOI] [PubMed] [Google Scholar]
  13. Cooper AR, Andersen LB, Wedderkopp N, Page AS, Froberg K. Physical activity levels of children who walk, cycle, or are driven to school. American Journal of Preventive Medicine. 2005;29(3):179–184. doi: 10.1016/j.amepre.2005.05.009. [DOI] [PubMed] [Google Scholar]
  14. D'Haese S, Van Dyck D, De Bourdeaudhuij I, Deforche B, Cardon G. Organizing "Play Streets" during school vacations can increase physical activity and decrease sedentary time in children. Int J Behav Nutr Phys Act. 2015;12:14. doi: 10.1186/s12966-015-0171-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Evenson KR, Catellier DJ, Gill K, Ondrak KS, McMurray RG. Calibration of two objective measures of physical activity for children. Journal of Sports Sciences. 2008;26(14):1557–1565. doi: 10.1080/02640410802334196. [DOI] [PubMed] [Google Scholar]
  16. Faulkner GE, Buliung RN, Flora PK, Fusco C. Active school transport, physical activity levels and body weight of children and youth: a systematic review. Preventive Medicine. 2009;48(1):3–8. doi: 10.1016/j.ypmed.2008.10.017. [DOI] [PubMed] [Google Scholar]
  17. Frank LD, Sallis JF, Saelens BE, Leary L, Cain K, Conway TL, Hess PM. The development of a walkability index: application to the Neighborhood Quality of Life Study. British Journal of Sports Medicine. 2010;44(13):924–933. doi: 10.1136/bjsm.2009.058701. [DOI] [PubMed] [Google Scholar]
  18. Gordon-Larsen P, Nelson MC, Popkin BM. Longitudinal physical activity and sedentary behavior trends: adolescence to adulthood. American Journal of Preventive Medicine. 2004;27(4):277–283. doi: 10.1016/j.amepre.2004.07.006. [DOI] [PubMed] [Google Scholar]
  19. Grow HM, Saelens BE, Kerr J, Durant NH, Norman GJ, Sallis JF. Where are youth active? Roles of proximity, active transport, and built environment. Medicine and Science in Sports and Exercise. 2008;40(12):2071–2079. doi: 10.1249/MSS.0b013e3181817baa. [DOI] [PubMed] [Google Scholar]
  20. Hills AP, Dengel DR, Lubans DR. Supporting public health priorities: recommendations for physical education and physical activity promotion in schools. Prog Cardiovasc Dis. 2015;57(4):368–374. doi: 10.1016/j.pcad.2014.09.010. [DOI] [PubMed] [Google Scholar]
  21. Holt NL, Spence JC, Sehn ZL, Cutumisu N. Neighborhood and developmental differences in children's perceptions of opportunities for play and physical activity. Health & Place. 2008;14(1):2–14. doi: 10.1016/j.healthplace.2007.03.002. [DOI] [PubMed] [Google Scholar]
  22. Janssen I, Rosu A. Undeveloped green space and free-time physical activity in 11 to 13-year-old children. Int J Behav Nutr Phys Act. 2015;12:26. doi: 10.1186/s12966-015-0187-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Joe L, Carlson J, Sallis JF. Active Where? Individual Item Reliability Statistics Adolescent Survey. 2012 Retrieved from http://www.drjamessallis.sdsu.edu/Documents/AW_item_reliability_Adolescent.pdf.
  24. Kanters MA, Bocarro J, Casper J. Supported or pressured? An examination of agreement among parent's and children on parent's role in youth sports. Journal of sport behavior. 2008;31(1):64. [Google Scholar]
  25. Kanters MA, Bocarro JN, Edwards MB, Casper JM, Floyd MF. School sport participation under two school sport policies: comparisons by race/ethnicity, gender, and socioeconomic status. Annals of Behavioral Medicine. 2013;45(Suppl 1):S113–121. doi: 10.1007/s12160-012-9413-2. [DOI] [PubMed] [Google Scholar]
  26. Kneeshaw-Price S, Saelens BE, Sallis JF, Glanz K, Frank LD, Kerr J, … Cain KL. Children's objective physical activity by location: why the neighborhood matters. Pediatric Exercise Science. 2013;25(3):468–486. doi: 10.1123/pes.25.3.468. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Larouche R, Saunders TJ, Faulkner G, Colley R, Tremblay M. Associations between active school transport and physical activity, body composition, and cardiovascular fitness: a systematic review of 68 studies. J Phys Act Health. 2014;11(1):206–227. doi: 10.1123/jpah.2011-0345. [DOI] [PubMed] [Google Scholar]
  28. Leek D, Carlson JA, Cain KL, Henrichon S, Rosenberg D, Patrick K, Sallis JF. Physical activity during youth sports practices. Archives of Pediatrics and Adolescent Medicine. 2011;165(4):294–299. doi: 10.1001/archpediatrics.2010.252. [DOI] [PubMed] [Google Scholar]
  29. Maddison R, Jiang Y, Vander Hoorn S, Exeter D, Mhurchu CN, Dorey E. Describing patterns of physical activity in adolescents using global positioning systems and accelerometry. Pediatric Exercise Science. 2010;22(3):392–407. doi: 10.1123/pes.22.3.392. [DOI] [PubMed] [Google Scholar]
  30. Mandic S, Bengoechea EG, Stevens E, de la Barra SL, Skidmore P. Getting kids active by participating in sport and doing it more often: focusing on what matters. Int J Behav Nutr Phys Act. 2012;9:86. doi: 10.1186/1479-5868-9-86. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Mattocks C, Ness A, Leary S, Tilling K, Blair SN, Shield J, … Riddoch C. Use of accelerometers in a large field-based study of children: protocols, design issues, and effects on precision. J Phys Act Health. 2008;5(Suppl 1):S98–111. doi: 10.1123/jpah.5.s1.s98. [DOI] [PubMed] [Google Scholar]
  32. McDonald NC, Steiner RL, Lee C, Rhoulac Smith T, Zhu X, Yang Y. Impact of the safe routes to school program on walking and bicycling. Journal of the American Planning Association. 2014;80(2):153–167. [Google Scholar]
  33. National Council of Youth Sports. Report on Trends and Participation in Organized Youth Sports. 2008 Retrieved from Stuart, FL: http://www.ncys.org/pdfs/2008/2008-ncys-market-research-report.pdf.
  34. Oreskovic NM, Blossom J, Field AE, Chiang SR, Winickoff JP, Kleinman RE. Combining global positioning system and accelerometer data to determine the locations of physical activity in children. Geospat Health. 2012;6(2):263–272. doi: 10.4081/gh.2012.144. [DOI] [PubMed] [Google Scholar]
  35. Ortega FB, Konstabel K, Pasquali E, Ruiz JR, Hurtig-Wennlof A, Maestu J, … Sjostrom M. Objectively measured physical activity and sedentary time during childhood, adolescence and young adulthood: a cohort study. PloS One. 2013;8(4):e60871. doi: 10.1371/journal.pone.0060871. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Pate RR, Davis MG, Robinson TN, Stone EJ, McKenzie TL, Young JC. Promoting physical activity in children and youth: a leadership role for schools: a scientific statement from the American Heart Association Council on Nutrition, Physical Activity, and Metabolism (Physical Activity Committee) in collaboration with the Councils on Cardiovascular Disease in the Young and Cardiovascular Nursing. Circulation. 2006;114(11):1214–1224. doi: 10.1161/circulationaha.106.177052. [DOI] [PubMed] [Google Scholar]
  37. Rainham DG, Bates CJ, Blanchard CM, Dummer TJ, Kirk SF, Shearer CL. Spatial classification of youth physical activity patterns. American Journal of Preventive Medicine. 2012;42(5):e87–96. doi: 10.1016/j.amepre.2012.02.011. [DOI] [PubMed] [Google Scholar]
  38. Rosenberg D, Ding D, Sallis JF, Kerr J, Norman GJ, Durant N, … Saelens BE. Neighborhood Environment Walkability Scale for Youth (NEWS-Y): reliability and relationship with physical activity. Preventive Medicine. 2009;49(2–3):213–218. doi: 10.1016/j.ypmed.2009.07.011. [DOI] [PubMed] [Google Scholar]
  39. Sallis JF, Cerin E, Conway TL, Adams MA, Frank LD, Pratt M, … Owen N. Physical activity in relation to urban environments in 14 cities worldwide: a cross-sectional study. Lancet. 2016;387(10034):2207–2217. doi: 10.1016/s0140-6736(15)01284-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Sallis JF, Conway TL, Cain KL, Carlson JA, Frank LD, Kerr J, … Saelens BE. Neighborhood built environment and socioeconomic status in relation to physical activity, sedentary behavior, and weight status of adolescents. Preventive Medicine. 2018;110:47–54. doi: 10.1016/j.ypmed.2018.02.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Sirard JR, Pfeiffer KA, Pate RR. Motivational factors associated with sports program participation in middle school students. Journal of Adolescent Health. 2006;38(6):696–703. doi: 10.1016/j.jadohealth.2005.07.013. [DOI] [PubMed] [Google Scholar]
  42. Smart Growth America. ( ). National Complete Streets Coalition. 2017 Retrieved from https://smartgrowthamerica.org/program/national-complete-streets-coalition/
  43. Stewart O, Moudon AV, Claybrooke C. Multistate evaluation of safe routes to school programs. American Journal of Health Promotion. 2014;28(3 Suppl):S89–96. doi: 10.4278/ajhp.130430-QUAN-210. [DOI] [PubMed] [Google Scholar]
  44. Strong WB, Malina RM, Blimkie CJ, Daniels SR, Dishman RK, Gutin B, … Trudeau F. Evidence based physical activity for school-age youth. Journal of Pediatrics. 2005;146(6):732–737. doi: 10.1016/j.jpeds.2005.01.055. [DOI] [PubMed] [Google Scholar]
  45. Tandon PS, Zhou C, Sallis JF, Cain KL, Frank LD, Saelens BE. Home environment relationships with children's physical activity, sedentary time, and screen time by socioeconomic status. Int J Behav Nutr Phys Act. 2012;9:88. doi: 10.1186/1479-5868-9-88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Telama R. Tracking of physical activity from childhood to adulthood: a review. Obes Facts. 2009;2(3):187–195. doi: 10.1159/000222244. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Timperio A, Ball K, Salmon J, Roberts R, Giles-Corti B, Simmons D, … Crawford D. Personal, family, social, and environmental correlates of active commuting to school. American Journal of Preventive Medicine. 2006;30 doi: 10.1016/j.amepre.2005.08.047. [DOI] [PubMed] [Google Scholar]
  48. Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Medicine and Science in Sports and Exercise. 2008;40(1):181–188. doi: 10.1249/mss.0b013e31815a51b3. [DOI] [PubMed] [Google Scholar]
  49. US Department of Health and Human Services. Physical activity guidelines for Americans. Washington, DC: 2008. Retrieved from http://www.health.gov/paguidelines/pdf/paguide.pdf. [Google Scholar]
  50. Vaisto J, Eloranta AM, Viitasalo A, Tompuri T, Lintu N, Karjalainen P, … Lakka TA. Physical activity and sedentary behaviour in relation to cardiometabolic risk in children: cross-sectional findings from the Physical Activity and Nutrition in Children (PANIC) Study. Int J Behav Nutr Phys Act. 2014;11:55. doi: 10.1186/1479-5868-11-55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Veitch J, Bagley S, Ball K, Salmon J. Where do children usually play? A qualitative study of parents’ perceptions of influences on children's active free-play. Health & Place. 2006;12(4):383–393. doi: 10.1016/j.healthplace.2005.02.009. http://dx.doi.org/10.1016/j.healthplace.2005.02.009 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1
NIHMS975307-supplement-1.docx (12.5KB, docx)

RESOURCES