Journal Logo

EPIDEMIOLOGY

Assessing the Contribution of Parks to Physical Activity Using Global Positioning System and Accelerometry

EVENSON, KELLY R.1,2; WEN, FANG1; HILLIER, AMY3; COHEN, DEBORAH A.4

Author Information
Medicine & Science in Sports & Exercise: October 2013 - Volume 45 - Issue 10 - p 1981-1987
doi: 10.1249/MSS.0b013e318293330e
  • Free

Abstract

Given the high levels of physical inactivity in the United States and its contribution to chronic diseases, morbidity, and mortality, efforts to reduce it are warranted (20,32). Countries like Finland and Brazil have been demonstrating success in increasing levels of physical activity at the population level by investing in parks and recreational facilities and programming (23,36). Whether this paradigm could be useful in the United States depends on the current degree with which parks are already used for moderate to vigorous physical activity (MVPA), the intensity of activity recommended in the 2008 National Physical Activity Guidelines (33).

In the United States, it is estimated that residents travel an average of 7 miles to access their closest park, with shorter distances in more urban areas (40). Quantitative and qualitative reviews indicate that access to parks is an important correlate of physical activity (15,16,26). In support of this, increasing access to places for physical activity was incorporated as part of the US National Physical Activity Plan, developed to help implement the 2008 guidelines (34). Public parks offer a free option for physical activity in most communities. For many, access is not a barrier to park use; nevertheless, it is unclear how much time people spend using parks and the contribution that park use makes to their overall physical activity.

A challenge to answering how important parks are for physical activity relates to how park use is measured because several methods exist. First, through surveys or interviews, participants report their own park use (e.g., [37]). However, this self-report is susceptible both to recall bias and social desirability bias, and if the study focuses on younger children, it must rely on parental recall. Second, participation in scheduled programs or classes, provided by parks and recreation departments or outside organizations, has been used in parks to indicate use of services, although uncommonly used for research (5). However, this method misses people who may use the park but have not signed up for any park services. Instead of focusing on use by individuals, a third method relies on observations by park users to indicate use, such as with the System for Observing Play and Recreation in Communities method (27). Limitations of the observational methods are the time and expense, as this method requires multiple observations for different days and seasons of the year to be reliable (7).

More recently, researchers are using global positioning systems (GPS) to assess where physical activity occurs (18,22). This method requires participants to wear a GPS monitor that locates where they are and requires researchers to access electronic maps of parks (in a format called shape files) to overlay with the participant locations from the GPS data to determine whether parks were visited. Researchers focusing on children have added accelerometry, in addition to GPS, to explore how much physical activity happens at parks (e.g., [14,19,29,38]). Expanding on these studies of youth, we used accelerometry with GPS to objectively measure park visits among a diverse sample of adults, a population for whom this work has not been explored. Using GPS and accelerometry, we build on currently available methods to describe patterns of park use and the contribution of parks to physical activity. Secondarily, we document the contribution to overall physical activity of the trip to and from the park. This research contributes to our understanding of how people are actually interacting with their environments. GPS data combined with accelerometry provides empirical evidence of location and level of physical activity, including visits to parks, and thus helps to fill this conceptual and empirical void.

METHODS

Study sample and recruitment

Participant enrollment occurred during the spring, summer, and fall between May 2009 and April 2011 (n = 92 in 2009, n = 148 in 2010, and n = 4 in 2011). Participants came from study centers in five states: Los Angeles, California; Albuquerque, New Mexico; Chapel Hill and Durham, North Carolina; Columbus, Ohio; and Philadelphia, Pennsylvania. At each of the five locations, participants were recruited at or near six (for New Mexico, North Carolina, Ohio, and Pennsylvania) or seven (for California) study parks (3). Compared with all available parks for each city, the study parks had more physical activity facilities, such as basketball courts, picnic areas, and fields. In the parks, participants were recruited in person, after completion of a brief park survey, and through posted flyers. In neighborhoods surrounding the park, participants were recruited after household interviews. More detail on recruitment is described elsewhere (3,12). Inclusion criteria for enrollment were age ≥18 yr, English speaking, ambulatory, and either living within 1 mile from the study park or recruited during a visit at the study park.

This study was approved by the institutional review board at each university or organization affiliated with the five study centers, and participants provided written informed consent. Visits were conducted in a research office or at the park. All participants were asked to wear an accelerometer and GPS unit for 3 wk, with weekly exchanges of units with local study staff. At enrollment, participants were weighed with a Tanita Bc551 scale and measured for height using a Seca Portable Stadiometer. Body mass index (BMI) was calculated as weight in kilograms divided by height in meters squared, and participants were grouped into four categories: underweight (<18.5 kg·m−2), normal weight (18.5–<25.0 kg·m−2), overweight (25.0–<30.0 kg·m−2), and obese (≥30.0 kg·m−2). Participants received a monetary incentive after the data collection period ($200–225 US dollars).

Physical activity measurement

For three 1-wk periods, participants were asked to wear an ActiGraph (model #GT1M; ActiGraph LLC, Pensacola, FL) accelerometer on their right hip secured by a belt to measure accelerations in the vertical plane. Each participant received written instructions and a telephone number to call for questions. Data were collected and stored in 60-s epochs. Nonwear time was defined as at least 90 consecutive minutes of zero counts, with allowance of 1 or 2 min of nonzero counts if no counts were detected in a 30-min window upstream and downstream of the 90-min period (4). Counts for nonwear minutes were set to missing.

Physical activity was calculated as average counts per minute and then converted to minutes per day based on time spent in different physical activity intensities. Several ActiGraph studies of adults provide count thresholds or cut points to distinguish MVPA from other forms of less intense activity. We used the cut points originally applied to the National Health and Nutrition Examination Survey (32), calculated by taking the weighted average of cut points from Freedson et al. (13), Yngve et al. (39), Leenders et al. (21), and Brage et al. (2). Vigorous intensity was defined as ≥5999 counts per minute and moderate intensity as 2020–5998 counts per minute. Bouts of MVPA were defined as consecutive sets of time ≥10 min when accelerometer counts were ≥2020 counts per minute, with allowance for interruptions of 1 or 2 min in the moving 10-min window below the threshold. A “lower moderate”–intensity threshold was calculated based on studies that incorporated more lifestyle activities that may be more appropriate for older adults (24), defined as 760–2019 counts per minute. Light-intensity physical activity was defined between 101 and 759 counts per minute, and sedentary behavior was defined as ≤100 counts per minute (25).

GPS measures

For three 1-wk periods, participants also were asked to wear the Qstarz BT-Q1000X portable GPS units (weight, 65 g; dimensions, 72 × 46 × 20 mm) on their waist during all waking hours. The units were set to record latitude, longitude, and speed every minute, with the Wide Area Augmentation System enabled (a system of satellites and ground stations that provides correction data to increase the accuracy of GPS readings). The map datum used was World Geodetic Survey 1984, and the position format was latitude and longitude in degrees and minutes (HD° MM′). Participants were asked to keep the unit dry and to charge it overnight, every night. Each participant received written instructions and a telephone number to call for questions.

GIS data

Shape files for the parks in the study were obtained from the appropriate parks and recreation and planning departments. This was supplemented with a 2010 national park shape file from Esri (Redlands, CA), with use supported by a recent study (11). Each participant’s home address was geocoded using 2010 TIGER/Line shape files in ArcGIS 10 and checked with electronic maps as needed. The Euclidean distance from home to the nearest park edge was calculated using the ArcGIS “near” function.

Statistical analysis

All statistical analyses were conducted in SAS version 9.3 (Cary, NC) and ArcGIS. Significance testing was set at P < 0.10. The GPS data files were downloaded and cleaned, removing data headers, converting coordinate data into decimal degrees, and transforming the data into wide-character ASCI format to enable further processing with SAS and ArcGIS. The accelerometer files were also downloaded and cleaned to match the minute-by-minute data with the GPS data. Each state had local study staff for data collection, ensuring that accelerometer and GPS data were matched within time zone. Because participants exchanged their units between weeks 1–2 and weeks 2–3, the GPS and accelerometer data from the unit being returned by the participant overlapped some with the data from the unit being picked up. These overlapping points were removed, and the 3 wk of data were merged into one file.

Geoprocessing procedures were used to extract points that fell within the study park and to remove points within 50 m of the participant’s residence, accounting for any inaccuracies in point locations. The removal of points near a participant’s residence affected only five participants. Points that corresponded to a speed of ≥30 km·h−1 were further removed to exclude driving within parks. These cleaned data were then processed using SAS. To be defined as a park visit, consecutive points within the park boundaries were required to span ≥3 min. A time gap of at least 45 min between consecutive park points was deemed two separate park visits. Otherwise, the points were considered as part of the same park visit. Certain special cases were further investigated to determine whether they were a park visit, including visits with an average speed ≥15 km·h−1, visits mostly on streets or roads but were inside a park, multiple visits in 1 d for the same participant, and visits occurring overnight. To compare characteristics of participants by recruitment method, Pearson χ2 tests were used. Wilcoxon nonparametric tests were used to compare physical activity on days when parks were and were not visited.

Secondarily, we explored the additional contribution to overall physical activity of travel to and from the park through active travel, such as through walking or bicycling. We expanded the GPS park visit data to include 1 h before and 1 h after each park visit and removed points within 50 m of the participant’s residence. The contribution of overall physical activity of the trip to and from the park was calculated two ways based on time and two ways based on distance. Using time, we counted physical activity that occurred 30 and 60 min before and after each park visit. Using distance, we counted at least light-intensity physical activity that occurred within a 1- and 5-mile buffer around the park within 1 h before or after the park visit.

RESULTS

Description of sample

In total, 248 participants enrolled in the study and provided both GPS and accelerometer data, with 238 providing 3 wk (and included in these analyses), four providing 2 wk, and six providing 1 wk of data. Among the sample (n = 238), age ranged from 18 to 85 yr (median, 37.0 yr; mean, 40.4 yr), 56.3% were females, 26.2% were non-Hispanic black, and 15.2% were Hispanic (Table 1). The sample included a range of educational levels and a relatively even distribution of weight categories. Overall, the predominant self-reported modes of travel to the park were walking (56.1%) and driving (55.9%), but not bicycling (2.7%). The percent that walked or bicycled to the park varied by site: 27.3% California, 47.8% New Mexico, 12.5% North Carolina, 46.8% Ohio, and 78.4% Pennsylvania.

TABLE 1
TABLE 1:
Descriptive characteristics of participants.

Overall, 80.1% of the participants were recruited within the park and 19.9% of the participants were recruited from households within 1 mile of the study park. Participants recruited from households compared with those recruited from parks were similar on most characteristics (sex, age, education, and BMI), except that those recruited from households were more likely to walk or bicycle to the park (P = 0.01) and were more likely to be non-Hispanic white (P = 0.07; data not shown).

Description of park visits

Participants lived on average 0.4 Euclidean miles from the closest park and 2.6 Euclidean miles from the parks they visited (Table 2). Participants visited parks a median of 2.3 times per week and 6.0 visits for the 3-wk period, using the GPS-based measures. During the 3-wk period, participants visited a median of 2.0 different parks. The median duration per day spent in parks was 42.0 min. In the sample, there were 20 participants (8.4%) that had no park visits during the 3-wk period (7 recruited from households and 13 recruited from the park).

TABLE 2
TABLE 2:
Descriptive characteristics of participants using GPS and GIS derived measures, overall and by site (N = 238).

Physical activity and park visits

Participants wore the accelerometer on average 11.5 h·d−1 (median, 11.8; interquartile range, 9.0–13.6). Overall participants engaged in an average of 26.8 min of MVPA per day (Table 3). Overall, a mean of 4.0% of the monitored day was spent in MVPA, 9.3% in low moderate, 24.4% in light activity, and 62.4% in sedentary behavior.

TABLE 3
TABLE 3:
Distribution of physical activity overall and during parks visits among all participants (N = 238).

Participants visited parks on average 6 d during the 3-wk period (median, 5 d; interquartile range, 2–9 d). Overall, 8.2% of all moderate activity and 9.4% of all vigorous activity occurred in parks. For time spent in the park, a mean of 12.0% was spent in MVPA, 15.7% in low moderate, 23.1% in light activity, and 49.3% in sedentary behavior.

Among those with at least one park visit (n = 218), counts per minute, lower moderate, moderate, MVPA, number and time in MVPA bouts per day, and sedentary behavior were all higher on days when a park was visited compared with days when a park was not visited (Table 4). Light activity was higher on days when a park was not visited compared with days when a park was visited.

TABLE 4
TABLE 4:
Accelerometer results overall and stratified by days with or without a park visit (n = 218a).

Active travel to and from the park

When considering the four definitions of active travel, either the 1- or 5-mile buffer, or the 30- and 60- minute period before and after the park visit, an additional average 3.7–6.6 min per park visit of MVPA were added (primarily through moderate activity) and 5.5–10.2 min per park visit of light moderate activity (Table 5).

TABLE 5
TABLE 5:
Distribution of physical activity per park visit (averaged for each person) and considering time before and after the park visit (n = 218a).

DISCUSSION

Patterns of park use and association with physical activity

The results show that although our sample of adults visited parks relatively frequently (average of 8.8 times for the 3-wk period), only 8.2% of all moderate and 9.4% of all vigorous activity occurred within parks. Although park users stayed at parks for an average of 53.3 min, parks were functioning more as a destination for light and sedentary behavior and less as a venue for MVPA. We postulate that one reason for this may be insufficient programming. When collecting observational data at selected parks in each study area, we documented that supervised and organized activities provided by parks and recreation departments or outside organizations, when present, were more often geared toward youth rather than adults (3). A Brazilian study found that public parks offering free supervised physical activity classes had higher usage and a higher prevalence of vigorous activity compared with parks not offering programming across age groups (28). Several studies also indicate that renovating parks can increase usage for both youth and adults (30,35). However, another study found that park use declined, despite park improvements, and cited a decline in programming as a possible reason (5,6). Offering free physical activity programming geared toward MVPA at the parks may be a particularly effective strategy to reach nearby residents and could include both youth and adults. In addition to programming, the types of facilities and the quality of those facilities may also be associated with use.

Although the distance to the closest park was short for most participants (median, 0.3 Euclidean miles), the parks that were visited were usually further away (2.4 Euclidean miles). Moreover, participants visited a median of two different parks for the 3-wk period. An understanding of why participants chose to attend certain parks would be useful to discern. This could be accomplished through qualitative query, such as showing participants the parks they attended through a GIS platform and asking questions about each visit. It could also be accomplished through quantitative study, combining use of parks with detailed information on the parks, such as facility offerings and quality of parks.

Additional contribution of travel to and from the park

Active travel to and from parks provides another opportunity for physical activity. A cohort study of US adults ages 38–50 yr who report the presence and use of neighborhood amenities found that destinations most commonly accessed were public transit stations (72%) and parks (46%) when walking and were parks (19%) and recreation facilities (10%) (1) when bicycling. These types of active trips were associated with more favorable BMI, waist circumference, and cardiorespiratory fitness. Moreover, those who bicycled had a lower risk of cardiovascular disease. Factors associated with walking to a park include higher perceived safety and aesthetics, living close to a variety of destinations, nongrid street pattern, children living at home, owning a dog, and walkable routes (8,17,31). In our study, when accounting for active travel to and from the park, estimated four different ways, the additional contribution to MVPA ranged from an average of 3.7–6.6 min per park visit. This contribution adds meaningfully to the average MVPA spent in the park and underscores the proximity to which many participants lived to the parks they visited (average 2.6 miles to the park visited). It also highlights the potential of parks to contribute even more to MVPA as a destination for adults.

We chose to use Euclidean (straight line) distances to calculate the active travel estimates but recognize that network distances (shortest street trip linking the origin and destination) better represent vehicular paths. Euclidean distances better account for alternative routes that might be taken if walking or bicycling to the park, but they are theoretically always shorter than network distances. Euclidean and network distances are strongly correlated; however, the correlation declines from more urban to suburban and rural areas (40).

Limitations and strengths

There are several limitations worth noting with regard to the objective GPS and accelerometry measures used in this study. GPS units have difficulty recording locations in dense urban environments, especially with large and closely connected buildings, or indoors. This could affect the classification of park visits that were inside recreation centers or other indoor park buildings, or in parks in urban areas surrounded by buildings. Also, the GPS battery could not last an entire week, so participants were asked to charge the unit each night. There is a small chance of missed park visits due to the need to charge the battery. Also, to minimize equipment failure, we met with participants weekly and exchanged both the GPS and the accelerometer units. This overlap in exchange time provided analytic challenges. In the future, the equipment exchange time and date should be recorded to ensure a simpler approach to merging the 3 wk of data. As monitor battery life improves, these concerns will be diminished.

The accelerometers we used in this study have evidence for both validity and reliability (9,13) but only provided information in the vertical plane, so some types of physical activities may have been misrepresented (i.e., bicycling and strength training). Also, participants were told not to wear the monitor during water activities, such as swimming. Thus, all of these activities are underrepresented with the accelerometer data. Newer accelerometers can overcome these limitations. There is likely not one cut point to define MVPA from a single waist-mounted accelerometer (24), particularly for older adults (10), which is why we also explored a lower moderate activity cut point (24).

Although we selected participants with diversity related to geographic location, sex, race/ethnicity, and age, a limitation is that they were volunteers. Moreover, we made an effort to recruit participants both within and around the parks; however, by design, our sample lived on average 0.4 miles from a park and so had higher access to parks than the general population, which in 2008 had national estimate of 7 miles (40). Our sample did have similar MVPA compared with a national sample recorded with accelerometry (32). It would be useful to next apply this technology to a representative sample of adults and to explore potential differences by urbanicity. It also would be useful to contrast parks that were visited to other parks in the area to identify characteristics of parks that are used more often overall and specifically for MVPA.

The strengths of this study include the diverse sample from different geographic regions and the use of objective measurement. Although most studies monitor behavior for 1 wk or less, this sample wore the monitors for 3 wk, minimizing the potential for reactivity. This article also contributes to the development of algorithms to use objective measures to determine park use that provided duration, frequency, and intensity of physical activity in parks. These methods could be replicated by other studies and provide an additional measurement methodology to the existing choices of self-reported park use, park user counts, and park observation. This methodology is also useful as a criterion measure to compare against self-reported park use questionnaires or diaries (12).

CONCLUSIONS

The use of objective technologies allowed us to explore park use for an extended period. This study indicated that participants were more active on days they went to a park, yet a small percentage of total daily MVPA occurred within a park. Parks were a destination that offered additional support for MVPA through active travel. New GIS-based tools and methods to identify active travel patterns should be developed in the future. Another step in this line of inquiry is to understand characteristics of parks as places that can contribute more often as a place and destination for physical activity.

This study was supported by the National Institutes of Health, National Heart Lung and Blood Institute (grant no. R01HL092569). The authors thank the Robert Wood Johnson Foundation, Active Living Research accelerometer lending library. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The authors thank the other SOPARC investigators, study coordinators, and staff for their help with this study; Sara Satinsky for her review of an earlier draft of the article; and the anonymous reviewers.

The authors declare no conflicts of interest.

The results of the present study do not constitute endorsement by the American College of Sports Medicine.

REFERENCES

1. Boone-Heinonen J, Jacobs DR Jr, Sidney S, Sternfeld B, Lewis CE, Gordon-Larsen P. A walk (or cycle) to the park: active transit to neighborhood amenities, the CARDIA study. Am J Prev Med. 2009; 37 (4): 285–92.
2. Brage S, Wedderkopp N, Franks P, Andersen L, Froberg K. Reexamination of validity and reliability of the CSA monitor in walking and running. Med Sci Sports Exerc. 2003; 35 (8): 1447–54.
3. Cohen DA, Lapham S, Evenson KR, et al. Use of neighbourhood parks: does socio-economic status matter? A four-city study. Public Health. 2013; 127 (4): 325–32.
4. Choi L, Liu Z, Matthews CE, Buchowski MS. Validation of accelerometer wear and nonwear time classification algorithm. Med Sci Sports Exerc. 2011; 43 (2): 357–64.
5. Cohen D, Setodji C, Evenson K, et al. How much observation is enough? Refining the administration of SOPARC. J Phys Act Health. 2011; 8 (6): 1117–23.
6. Cohen DA, Golinelli D, Williamson S, Sehgal A, Marsh T, McKenzie TL. Effects of park improvements on park use and physical activity: policy and programming implications. Am J Prev Med. 2009; 37 (6): 475–80.
7. Cohen DA, Sehgal A, Williamson S, Marsh T, Golinelli D, McKenzie TL. New recreational facilities for the young and the old in Los Angeles: policy and programming implications. J Public Health Policy. 2009; 30 (1 Suppl): S248–63.
8. Dills J, Rutt C, Mumford K. Objectively measuring route-to-park walkability in Atlanta, Georgia. Env Behavior. 2013; 44 (6): 841–60.
9. Esliger D, Tremblay M. Technical reliability assessment of three accelerometer models in a mechanical setup. Med Sci Sports Exerc. 2006; 38 (12): 2173–81.
10. Evenson K, Buchner D, Morland K. Objective measurement of physical activity and sedentary behavior among United States adults 60 years and older. Prev Chronic Dis. 2012;9(1):1–10. Available from: http://www.cdc.gov/pcd/issues/2012/11_0109.htm.
11. Evenson KR, Wen F. Using Geographic Information Systems to compare municipal, county, and commercial parks data. Prev Chronic Disease. 2013; 10 (June 6, 2013): 1–8. Available at: http://www.cdc.gov/pcd/issues/2013/12_0265.htm.
12. Evenson K, Wen F, Golinelli D, Rodriguez D, Cohen D. Measurement properties of a park use questionnaire. Env Behavior. 2013; 45 (4): 522–43.
13. Freedson P, Melanson E, Sirard J. Calibration of the Computer Science and Applications, Inc. accelerometer. Med Sci Sports Exerc. 1998; 30 (5): 777–81.
14. Jones AP, Coombes EG, Griffin SJ, van Sluijs EM. Environmental supportiveness for physical activity in English schoolchildren: a study using Global Positioning Systems. Intl J Behav Nutr Phys Activity. 2009; 6: 42.
15. Kaczynski A, Henderson K. Environmental correlates of physical activity: a review of evidence about parks and recreation. Leisure Sci. 2007; 29: 315–54.
16. Kaczynski AT, Henderson KA. Parks and recreation settings and active living: a review of associations with physical activity function and intensity. J Phys Act Health. 2008; 5 (4): 619–32.
17. Koohsari M, Karakiewicz J, Kaczynski A. Public open space and walking: the role of proximity, perceptual qualities of the surrounding built environment, and street configuration. Env Behavior. 2013; 45 (6): 706–36.
18. Krenn PJ, Titze S, Oja P, Jones A, Ogilvie D. Use of global positioning systems to study physical activity and the environment: a systematic review. Am J Prev Med. 2011; 41 (5): 508–15.
19. Lachowycz K, Jones AP, Page AS, Wheeler BW, Cooper AR. What can global positioning systems tell us about the contribution of different types of urban greenspace to children’s physical activity? Health Place. 2012; 18 (3): 586–94.
20. Lee IM, Shiroma EJ, Lobelo F, Puska P, Blair SN, Katzmarzyk PT. Effect of physical inactivity on major non-communicable diseases worldwide: an analysis of burden of disease and life expectancy. Lancet. 2012; 380 (9838): 219–29.
21. Leenders N, Nelson T, Sherman W. Ability of different physical activity monitors to detect movement during treadmill walking. Intl J Sport Med. 2003; 24: 43–50.
22. Maddison R, Ni Mhurchu C. Global positioning system: a new opportunity in physical activity measurement. Intl J Behav Nutr Phys Activity. 2009; 6: 73.
23. Malta DC, Barbosa da Silva J. Policies to promote physical activity in Brazil. Lancet. 2012; 380 (9838): 195–6.
24. Matthews C. Calibration of accelerometer output for adults. Med Sci Sports Exerc. 2005; 37 (11 Suppl): S512–22.
25. Matthews C, Chen K, Freedson P, et al. Amount of time spent in sedentary behaviors in the United States, 2003–2004. Am J Epidemiol. 2008; 167 (7): 875–81.
26. McCormack G, Rock M, Toohey A, Hignell D. Characteristics of urban parks associated with park use and physical activity: a review of qualitative research. Health Place. 2010; 16: 712–26.
27. McKenzie T, Cohen D, Sehgal A, Williamson S, Golinelli D. System for Observing Play and Recreation in Communities (SOPARC): reliability and feasibility measures. J Phys Act Health. 2006; 3 (1 Suppl): S208–22.
28. Parra DC, McKenzie TL, Ribeiro IC, et al. Assessing physical activity in public parks in Brazil using systematic observation. Am J Public Health. 2010; 100 (8): 1420–6.
29. Rodriguez DA, Cho GH, Evenson KR, et al. Out and about: association of the built environment with physical activity behaviors of adolescent females. Health Place. 2012; 18 (1): 55–62.
30. Tester J, Baker R. Making the playfields even: evaluating the impact of an environmental intervention on park use and physical activity. Prev Med. 2009; 48: 316–20.
31. Tilt JH. Walking trips to parks: exploring demographic, environmental factors, and preferences for adults with children in the household. Prev Med. 2010; 50 (1 Suppl): S69–73.
32. Troiano R, Berrigan D, Dodd K, Masse L, Tilert T, McDowell M. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc. 2008; 40 (1): 181–8.
33. US Department of Health and Human Services. 2008 Physical Activity Guidelines for Americans. ODPHP Publication No. U0036 [cited 2013 Feb 14]. Available from: http://www.health.gov/paguidelines/.
34. US National Physical Activity Plan Coordinating Committee. 2010. National Physical Activity Plan [cited 2013 Feb 14]. Available from http://www.physicalactivityplan.org/NationalPhysicalActivityPlan.pdf.
35. 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): 383–93.
36. Vuori I, Lankenau B, Pratt M. Physical activity policy and program development: the experience in Finland. Public Health Rep. 2004; 119 (3): 331–45.
37. Walker JT, Mowen AJ, Hendricks WW, Kruger J, Morrow JR Jr, Bricker K. Physical activity in the park setting (PA-PS) questionnaire: reliability in a California statewide sample. J Phys Act Health. 2009; 6 (1 Suppl): S97–104.
38. Wheeler BW, Cooper AR, Page AS, Jago R. Greenspace and children’s physical activity: a GPS/GIS analysis of the PEACH project. Prev Med. 2010; 51 (2): 148–52.
39. Yngve A, Nilsson A, Sjostrom M, Ekelund U. Effect of monitor placement and of activity setting on the MTI accelerometer output. Med Sci Sports Exerc. 2003; 35 (2): 320–6.
40. Zhang X, Lu H, Holt JB. Modeling spatial accessibility to parks: a national study. Intl J Health Geographics. 2011; 10: 31.
Keywords:

ACCELEROMETER; ACTIVE TRAVEL; GEOGRAPHIC INFORMATION SYSTEMS (GIS); GLOBAL POSITIONING SYSTEMS (GPS)

© 2013 American College of Sports Medicine