Identification of Effective Programs to Improve Access to and Use of Trails among Youth from Under-Resourced Communities: A Review

The primary purpose of this paper is to identify and review studies evaluating the effectiveness of programs to increase access to trails and trails use (physical activity) among youth from under-resourced communities. Three additional goals include identifying: (1) Correlates of physical activity/trail use and features of transportation systems and/or built environment and land use destinations, that may inform and support the planning and implementation of programs to promote trail use among youth, (2) benefits associated with trail use, and (3) barriers to trail use. Under-resourced communities are defined as those lacking sufficient resources (i.e., under-funded). METHODS: A review of the literature was conducted to identify, abstract, and evaluate studies related to programs to promote trail use among youth and youth from under-resourced communities. In anticipation of very few studies being published about this topic, studies were also reviewed to identify correlates of transportation systems and built environment and land use destinations related to increases in physical activity, and benefits of, and barriers to trail use. PUBMED, MEDLINE, PsycINFO, Sportdiscus, Annual Reviews, American Trails, and Google Scholar databases were searched using terms including trails, built environment, physical activity, exercise, walking, children, adolescents, and youth to identify studies that potentially related to the purposes for conducting this review. Review methods identified, 5278 studies based on our search terms. A review of study titles, abstracts, and select full article screens determined that 5049 studies did not meet the study inclusion criteria, leaving 221 studies included in this review. RESULTS: No studies were located that evaluated programs designed to promote and increase trail use among youth, including youth from under-resourced communities. Eight studies used longitudinal or quasi-experimental designs to evaluate physical activity and neighborhood characteristics prospectively among adolescent girls (n = 1), the effects of the path or trail development on physical activity behaviors of children, youth, and adults (n = 4), marketing or media campaigns (n = 2), and wayfinding and incremental distance signage (n = 1) to promote increased trail use. Correlates of transportation systems (e.g., trail access, road traffic congestion related to safe active travel, lack of sidewalks, closer proximity to trails, access to transportation), destinations (e.g., park availability and access, park improvements, greenspaces), or both routes and destinations (e.g., perceptions of safety, lighting), were identified. These correlates may support the planning and implementation of programs to increase trail use among youth, or may facilitate the connection of trails or routes to destinations in communities. Barriers to trail use included costs, crime, lack of transportation, lack of role models using trails, and institutional discrimination. Conclusions: Scientific evidence in support of addressing the underrepresentation of trail use by youth from under-resourced communities is lacking. However, there is a related body of evidence that may inform how to develop programs that support trail use by youth from under-resourced areas. Dedicated, deliberate, and systematic efforts will be required to address research and knowledge gaps, and to evaluate programs and practice related to trail use among youth from low income, often racially or ethnically diverse under-resourced neighborhoods or communities.


Trails, Built Environment Features, Physical Activity and Health
Community infrastructure is often considered a foundation for health and wellness that may affect planning and land-use decisions that may also be related to greater physical activity and improved health outcomes [24][25][26]. Researchers have identified trails, as part of the integral infrastructure for physical activity [27][28][29], and trails have been found to be associated with regular physical activity participation [30,31]. Researchers have also recommended pedestrian or bicycle routes connect with destinations to promote physical activity for transportation [23,[30][31][32][33][34] and leisure physical activity. Access to recreational trails is widely accepted to influence physical activity participation among varying populations including children and youth [35][36][37][38][39][40][41][42][43][44][45].
Several approaches are being taken to understand the importance of trails to support physical activity and enhance health. Reed and colleagues examined activity behaviors in 25 parks and found trails to be the most frequently used feature in the parks [46]. Sixty-percent of adult males and 81% of adult females observed in all 25 parks were on trails. The South Carolina Trails Survey highlighted by American Trails, however, found few children and youth using a prominent rail-trail conversion [47]. Trail Development and increased access to trails may promote regular physical activity. Librett and colleagues [48] examined the physical activity levels among trail users in the U.S., and found that individuals who reported using trails at least once a week were twice as likely to meet physical activity recommendations as individuals who reported rarely or never using trails. Hughey and colleagues [24] examined the associations between adults' use of a paved trail, their weight status and self-rated health and found that trail users were half as likely to be overweight or obese as trail nonusers. Trail users were also significantly more likely to report high self-rated health than nonusers. Kaczynski and colleagues [49] found that park features (e.g., trails and/or paths in a park, playgrounds), contributed to participation in physical activity, and trails had the strongest relationship with activity participation of all park features in their study. A review article by Mitten et al. identified numerous health benefits associated with hiking behavior. Benefits included increased time spent in social contacts, and enhanced mental, emotional, and physical health. These benefits may be acute such as reduced blood pressure and stress, or improved immune system functioning and restored attention. Chronic benefits derived from hikers over time have included weight loss, reduced depression, and greater wellness [25]. This body of work is consistent with the idea that trails could increase physical activity levels, however, most of these studies are cross-sectional.
Nevertheless, the U.S. Community Preventive Services Task Force (Task Force) released a recommendationon the influence of the built environment on physical activity [50]. The Task Force recommends, based on 90 studies (16 longitudinal and 74 cross-sectional), "built environment strategies that combine one or more interventions to improve pedestrian or bicycle transportation systems with one or more land use and environmental design interventions to increase physical activity." The Task Force further recommended that "Coordinated approaches must combine new or enhanced elements of transportation systems with new or enhanced land use and environmental design features. Intervention approaches must be designed to enhance opportunities for active transportation, leisure-time physical activity, or both." The Task Force recommendations pertain to all ages and include children and adolescents. However, there is scarce evidence about how to promote and optimize the use of routes connected with destinations especially among youth and disparate populations such as different racial/ethnic groups. Considerations about "how to" tailor interventions to promote and increase trail use among disparate groups are especially important, since education and income are both positively associated with trail use [41]. Greater efforts are needed to specify how connectivity can best be addressed in support of increased trail use for under-resourced communities.

Trails or Other Built Environment Features, and Physical Activity among Youth from Under -Resourced Communities
Multi-use recreational trails influence adult physical activity participation [23,30,[35][36][37]48,51,52]. An understudied research topic is the impact of trails and other built environment interventions on the physical activity behaviors of children and adolescents including those living in under-resourced (also referred to as "underserved" in the literature) neighborhoods or communities [53][54][55][56][57][58][59][60][61]. The use of the terms under-resourced children and youth in this paper refers to those who are living in under-resourced communities. These communities are often predominantly made up of residents from diverse racial, ethnic, and cultural backgrounds, who also typically lack the access to health care, economic, financial, and social benefits accruing to people from more affluent backgrounds and communities.
Community and built environments to promote physical activity earned a "C" from the 2018 U.S. Report Card on Physical Activity for Children and Youth ("The Report Card is a resource that summarizes health statistics related to physical activity levels among children and youth in the U.S. More importantly, the Report Card is an advocacy tool that provides a level of accountability and call-to-action for decision-makers regarding how we, as parents, teachers, health professionals, community leaders, and policymakers, can implement new initiatives, programs, and policies in support of healthy environments to improve the physical activity levels and health of our children and youth."). This grade decrease from a B-in the two previous evaluations was primarily related to children and youth having barriers to access to parks and other recreation facilities. The 2018 Report Card considered new elements of the community and built environment such as: safety, walkability of the community, and complete streets policies [62]. These additional elements were the key reason for the lowered grade. Nonetheless, the Report Card highlighted the importance of improving the community and built environments to support child and youth physical activity, with a focus on underserved (under-resourced) communities [62]. Considering less than one-third of U.S. States report having at least 30% of youth residing in high walkable communities, improving these environments is essential [63] including creating or improving trails and trail connectivity.
Survey data indicate that members of under-resourced communities are underrepresented among trail users [64]. The National Visitor Use Monitoring (NVUM) survey [64], an ongoing questionnaire administered to U.S. Forest Service park users over a 10-year timeframe, indicates that 95% of people engaged in recreation on U.S. Forest Service lands are White; 6% are Spanish, Hispanic, or Latino; about 3% Asian; 2% American Indian/Native Alaskan; and only 1% are Black or African American (respondents could choose more than one racial group). This contrasts with the overall race/ethnic composition of the U.S. which was 61% Non-Hispanic White, 15.3 Hispanic, 13.4% Black or African American, 5.9% Asian with the remainder Native American, Alaska Native, Native Hawaiian or two or more races as of 2019. Survey findings also indicate that youth, age 16 to 19 years, recreate on U.S. Forest Service lands less than any other age group (about 4%), except for adults 70 years and older (5%). Although U.S. Forest Service trails survey data may not be representative of trail users in more urban or even rural parts of America, the existing evidence indicates that youth, especially youth from diverse backgrounds and cultures, are underrepresented among trail users [22].
To address this gap in knowledge, the National Collaborative on Childhood Obesity Research (NCCOR)-a public-private partnership among the National Institutes of Health, the Centers for Disease Control and Prevention (CDC), the Robert Wood Johnson Foundation, and the U.S. Department of Agriculture, along with collaboration for this project from staff from the Federal Highway Administration formed a scientific workgroup of member organizations and partners to investigate effective interventions and programs for increasing trail use among children and youth from under-resourced communities. NCCOR also engaged a Principal Investigator (PI) (author Julian A. Reed) to conduct the review, with feedback from the workgroup. Methods used to conduct the review, results, and discussion of findings are highlighted next.

Methods
A review of the peer-reviewed literature was conducted to identify efficacious and effective programs for increasing trail use (urban, nonurban, and/or more remote trail use) among children and youth of all abilities who are from underserved/under-resourced communities. Studies were also reviewed to identify correlates of physical activity/trail use and features of transportation systems and/or built environment and land use destinations. For this review, the PI searched health, medical, public health and sport sciences journal databases in PUBMED, MEDLINE, PsycINFO, Sportdiscus, Annual Reviews, American Trails, and Google Scholar from 2000-2018 for research related to trail use and physical activity by children and youth with a primary focus on intervention studies. These databases identified publications in health, medical, public health and sport sciences fields with a focus on peer-reviewed, primarily, journal articles including abstracts and studies related to our topics of interest.
Eligibility criteria included papers addressing: (a) children and youth (ages six to 17) and physical activity in the built environment(s) with a primary focus on pedestrian or multi-use trails (i.e., trails, trails in parks, rails-to-trails, sidewalks, trails in greenspaces, walking paths, neighborhood trails, city trails), and (b) underserved children, and adolescents and built environment, trails and physical activity. Although we prefer to use the term under-resourced rather than underserved regarding our priority population, we used underserved as a search term, because it is a term that has been used frequently in other studies). The references of all reviewed articles were also examined to identify other studies related to our review (see Figure 1). More than 50% of all abstracts identified were cross-referenced by the primary investigator and at least one other member of the research team. All (n = 324) selected papers for potential relevance were reviewed by two or more research team members and the primary investigator.
A combination of the following search terms was used: "Trails"; "Trails and Physical Activity"; "Trails and Children and Youth"; "Built Environment and Adolescents, Youth"; "Children and Youth, Adolescent Physical Activity and Built Environment"; "Exercise and Built Environment"; "Physical Activity and Open Spaces"; "Open Spaces and Adolescents and Youth"; "Recreational Facilities and Youth Physical Activity"; "Walking Trails"; "Walking Trails and Youth and Adolescents"; "Active Transportation and Youth"; "Child Physical Activity"; "Adolescent Physical Activity"; "Urban Trails and Physical Activity"; "Rural Trails and Physical Activity"; "Rural Trails and Child Physical Activity"; "Greenspace and Physical Activity"; "Built Environment Design and Physical Activity"; "Public Facilities and Physical Activity"; "Intervention Studies and Physical Activity, Children, Youth, Adolescents".  [32,66] or informational strategies to promote and increase trail use [66]. Of these eight studies, one focused solely on adolescent girls [67], and is reviewed in more detail below. Two additional studies [68,69] are also highlighted below as examples of the types of longitudinal research needed to advance the science and knowledge related to youth trail use. The remaining five studies [32,65,66,71] focusing on adults or adults and youth combined are also excellent examples of prospective studies related to trails and trail use.

Results
The search strategy initially identified 5278 potential articles from the literature search ( Figure 1). Of these, 221 eligible studies were selected for full abstraction and inclusion in this review. The vast majority of studies located were correlational in nature. However, eight studies used prospective, longitudinal, or quasi-experimental designs [32,[65][66][67][68][69][70][71] evaluating increased availability of trails [65,71] or other neighborhood characteristics [67] with physical activity/trail use, and media campaigns [32,66] or informational strategies to promote and increase trail use [66]. Of these eight studies, one focused solely on adolescent girls [67], and is reviewed in more detail below. Two additional studies [68,69] are also highlighted below as examples of the types of longitudinal research needed to advance the science and knowledge related to youth trail use. The remaining five studies [32,65,66,71] focusing on adults or adults and youth combined are also excellent examples of prospective studies related to trails and trail use.  Figure 3) were cross-sectional analyses of convenience or locally representative samples (n = 138). The remaining studies included analyses of nationally representative samples [17], reviews [26], quasi-experimental or longitudinal [8] and a small number of experimental studies, policy statements, study designs and methods research. Most studies occurred in the United States ( Figure 4). Figure 5 illustrates the measurement approaches used in the studies underlying the review. More than half of the studies use multiple measurement approaches, and more than 40 studies use three or more measurement approaches, highlighting the need for multi-disciplinary teams to investigate trail use and its determinants.

Study Primary Purpose: Effective Programs for Increasing Trail Use among Children and Youth from Under-Resourced Communities
The results indicate that there are no published experimental studies that have evaluated the effectiveness of programs designed to encourage or increase the use of trails by youth from under-resourced communities. However, some longitudinal evidence exists on the influence of neighborhood characteristics and transport with 2-year changes in accelerometer-assessed non-school moderate-to vigorous-intensity physical activity, and sedentary behavior in a sample of adolescent girls from diverse race/ethnic groups (53.5% White; 18% Black; 19.1%; Hispanic; 4.8% Asian, Native Hawaiian or Pacific Islander [67]. Physical activity behavior typically declines with adolescents with age [67,72]. For this study, the researchers hypothesized "that girls who perceived a more conducive neighborhood environment for physical activity, more physical activity opportunities, and better transport options at baseline would have more favorable changes from 6th to 8th grade in non-school physical activity and sedentary behavior". This study, however, did not find that neighborhood features such as good lighting and trails were prospectively related to increases in physical activity at 2-years follow-up. Rather neighborhood features were associated with decreases in non-school Met-weighted Moderate-to Vigorous-intensity Physical Activity (MW-MVPA) at follow-up. No neighborhood measures were associated with sedentary behavior [67]. The researchers indicate that there may be two possible reasons for their findings. First, it may be that girls' who reported common physical activities in 6th grade shifted to doing more sedentary behaviors such as talking on the phone and music lessons in middle school that replaced popular physical activities done during elementary school. Second, perhaps relationships among the environment and physical activity change as children age for reasons other than doing preferred types of physical activities. Third, children who engage in physical activity outdoors may be more likely to notice dangerous locations and unpleasant smells in their neighborhood, that actually impede activity, because outdoor activity, gives them greater exposure to these characteristics than peers who are less active.
In addition to the prospective study above, cross-sectional and observational studies have evaluated associations between environmental infrastructure and neighborhood characteristics and physical activity among youth from diverse backgrounds or groups [73][74][75][76][77][78]. One study [21] surveyed diverse youth (White, 48.5%; Black, 18.7%; Hispanic, 14.1%, Asian/Pacific Islander, 3.1%) to evaluate the associations between their perceptions of environmental factors and transportation with their physical activity and active transport to school behaviors. Perceptions of neighborhood walking trails, ease of walking or biking to transit, destinations within walking distance from home, neighborhood safety, aesthetics (more trees, interesting things to look at, lack of garbage or litter), number of activity facilities, and parental support to walk or bike for transportation were all related to physical activity, and in some cases to active transport [21]. www.mdpi.com/journal/ijerph longitudinal [8] and a small number of experimental studies, policy statements, study designs and methods research. Most studies occurred in the United States ( Figure 4). Figure 5 illustrates the measurement approaches used in the studies underlying the review. More than half of the studies use multiple measurement approaches, and more than 40 studies use three or more measurement approaches, highlighting the need for multi-disciplinary teams to investigate trail use and its determinants.      . Measurement approaches used in the studies considered in this review were carried out. Reviews, policy statements and study design paper were categorized as not applicable (N/A).

Study Primary Purpose: Effective Programs for Increasing Trail Use among Children and Youth from Under-Resourced Communities.
The results indicate that there are no published experimental studies that have evaluated the effectiveness of programs designed to encourage or increase the use of trails by youth from underresourced communities. However, some longitudinal evidence exists on the influence of neighborhood characteristics and transport with 2-year changes in accelerometer-assessed nonschool moderate-to vigorous-intensity physical activity, and sedentary behavior in a sample of adolescent girls from diverse race/ethnic groups (53.5% White; 18% Black; 19.1%; Hispanic; 4.8%  Reviews, policy statements and study design paper were categorized as not applicable (N/A).
In another study evaluating the relationship between community design features and access to recreational facilities among a diverse population of adolescents (43% ethnic minority) [34], there was some evidence that community design and access to recreational facilities are associated with physical activity. Nearby recreation facilities and the number of nearby parks correlated positively, and intersection density, inversely, with girls' physical activity. The amount of retail floor area correlated positively with boys' physical activity. No design or recreational access features correlated with BMI percentiles. In another study, researchers examined the relationship between parental nativity with active transport to school, active use of sidewalks, use of local neighborhood parks, and use of neighborhood physical activity facilities, in an under-resourced neighborhood in New Jersey, and found Latino youth with foreign-born parents were generally more physically active than were Latino youth with parents born in the U.S. [73]. however out of the four areas evaluated (transport to school, use of parks, sidewalks, and facilities) findings were significant only for the active transport to school. Latino youth with parents living in the U.S. for less than 10 years were 1.5-times more likely to walk to school than the youth of U.S.-born parents. These findings suggest that parental nativity status may be an important influence on policies or interventions designed to increase physical activity among Latino youth. The authors state that "identifying the factors that seem to be health-enhancing among foreign-born populations," may help inform and tailor interventions that can be effectively used with U.S.-born youth with foreign-born parents. In another study [74], a national Safe Routes to School survey was adapted to evaluate if, and how children and adolescent school students in a traditionally under-resourced, predominantly Latino community in East Los Angeles used an urban greenway for transportation to school. The findings suggest that the greenway pathway near an elementary school and high school made "it easier and safer for students to walk or bike to school compared, to using the often inadequate sidewalk and street infrastructures in this urban environment" [74]. Findings may inform the use of greenways for transportation in other communities. Another study focused on the disparities of under-resourced adolescents and physical activity resources in neighborhoods [75]. Data from the 2011-2012 National Survey of Children's Health were analyzed to identify the demographic characteristics of youth, aged 10-17 years, who live in neighborhoods that are supportive of physical activity. Findings indicate that the proportion of youth living with access to built environment supports for physical activity was lower among non-Hispanic Black or Hispanic youth, those overweight or obese, from homes with lower socioeconomic status or from rural areas [75].
Two additional studies conducted outside the U.S. provide insights into the built environment and physical activity among youth from diverse or under-resourced areas. Cross-sectional relationships were examined between neighborhood-built environment and walkability characteristics, and socio -economic status, with multiple physical activity behaviors, sedentary time, and obesity indicators among adolescents from Valencia, Spain [76]. Moderate-vigorous physical activity was highest, and sedentary minutes lowest, in high-SES/high-walkable neighborhoods. Neighborhood SES was also positively related to participation in sports teams, and physical activity classes and, negatively related to time spent in sedentary behaviors. Adolescents living in lower-SES neighborhoods watched more TV and were more obese. Findings from this study indicate it is important to take into consideration the interaction of neighborhood built and SES environments when planning health promotion interventions for adolescents. In another study [77], the prevalence and correlates of active commuting to school (ACS) were examined in a nationally representative sample of Mexican adolescents, using data of adolescents ages 10-14 years, from the 2012 Mexican National Health and Nutrition Survey [77]. ACS was negatively associated with travel time, age, mother's education level, household motor vehicle ownership, family socioeconomic status, living in urban areas, the North region of the country, and overweight/obesity. Each additional minute of ACS was associated with a 1% decrease in the odds of being overweight or obese.
The above studies related to neighborhood characteristics, the built environment and physical activity among youth from diverse and under-resourced communities are difficult to interpret, because of different methods and outcomes evaluated in the studies. However, one consistent finding is that SES is consistently positively related to physical activity, and inversely related to inactivity/sedentary behavior among youth, in the U.S. [75] and across cultures and other countries [78,79].

Study Secondary Purpose 1: Transportation Systems and Built Environment and Land-Use Correlates and
Features that May Support Programs Designed to Provide Increased Opportunities for Physical Activity (i.e., Trail Use) among Youth Cross-sectional studies and several reviews report that greater access to routes and features of destinations are associated with improvements in daily moderate to vigorous physical activity participation [26,29,57,61,76,[79][80][81][82]. To date, these studies have not included a focus on diverse racial or ethnic groups or under-resourced or low SES neighborhoods or communities. Mitchell and colleagues [61], for example, found that boys and girls from neighborhoods with more access to parks having sports fields had significantly higher moderate-and vigorous-intensity physical activity. When the researchers accounted for individual and neighborhood socio-demographic variables, multi-use path space also favorably predicted moderate-to vigorous-intensity physical activity [61]. Another example, typical of this body of research, is an observational study that relied on direct observation of behaviors in 45 parks in a southeastern community. This study identified large numbers of children participating in physical activity on trails. Trails were located in 16 of the 45 parks (36%), and close to 10% of all boys and girls were observed using trails in those 16 parks [22]. Findings from these studies provide examples of trails supporting youth physical activity, and indicate that efforts to prioritize planning trail connectivity to recreational spaces, and facilities to support youth physical activity are important [83].
Many other cross-sectional studies have identified correlates of trails or other routes (e.g., paths, sidewalks, bicycle routes) and physical activity/trails use (see Table A1, Section A) for select examples). Overall, these studies illustrate that access to trails, paths, trails in parks, access to transportation, closer proximity to trails, road traffic-related to safe active travel, intersection density, and lack of sidewalks) are related to physical activity. These correlates may be important to consider in future interventions connecting trails to programs that promote trail use, and that are tailored for use with youth from diverse or under-resourced neighborhoods or communities.
In addition, many studies have identified correlates of destinations (e.g., parks, open spaces, playgrounds), and physical activity/trails use (see Table A1, Section B) for select examples). Overall, these studies illustrate that access to parks, greenspaces, open-spaces, and recreational facilities, and park enhancements (e.g., playgrounds), are also related to physical activity. These correlates may be important to consider during efforts to plan and develop effective programs to promote and increase trail use, outdoor recreational physical activity, and enjoyment of nature that are tailored for use with youth from diverse or under-resourced neighborhoods or communities.
Studies identifying correlates of both transportation systems and destinations and physical activity/trail use are also listed in Table A1 (Section C). These include favorable perceptions of safety, neighborhood size, neighborhood features (e.g., lighting, signage (e.g., distance from route to destinations), and mapping (wayfinding), as examples.

Study Secondary Purpose 2: Benefits Related to Trails Use
With the exception of one study [56], results did not lead to the identification of physical, mental/emotional, or social benefits of trail use by youth or adult trail users. McCracken, et al. [56] found that more access to greenspace was associated with better health-related quality of life, and self-esteem among children. Although the use of trails has been found to increase physical activity behaviors, our knowledge about other benefits of trails use is very limited, as highlighted in the subsection above titled, Trails or Other Built Environment Features, Physical Activity and Health.

Study Secondary Purpose 3: Barriers to Trail Use among Youth from Under-Resourced Communities
Results indicate that many youths from under-resourced communities lack exposure, experiences, and opportunities to access trails, and obtain benefits of trails use, due to a variety of social determinants of health (e.g., costs, lack of transportation, lack of role models using trails, and institutional discrimination) [50,83,84], crime [85,86] and automobile traffic [78,79,87]. Additional barriers to youth physical activity included schoolwork, weather conditions [88], and strangers [89].
The findings indicating that lack of seeing role models on trails is a barrier to trail use for some population groups, may be interrelated with findings from other studies. These include observations of park users in 45 parks that found a greater frequency and percentage of white youth compared to minorities [22], and a "major inhibitor" to park use that has been reported to include racial heterogeneity of the neighborhood [90]. Among Caucasian youth in Canada aged 8-10, perceptions of a high proportion of minorities in the surrounding area decreased feelings of safety related to physical activity. While for parents, high traffic and lack of perceived community involvement contributed to decreased feelings of safety [85].
Although much research has been published examining physical activity behaviors related to routes (e.g., trails, pathways), and destinations (e.g., parks, greenspaces), the majority of the studies have relied on correlational and cross-sectional designs that do not identify what works or how to intervene to be more inclusive of population groups who are underrepresented trail users. More robust study designs are needed to gain knowledge about programs that promote and increase trail use among youth from under-resourced communities, as are discussed next.
Few studies exist that specifically focus on the influence of trails and trail features on children's and adolescents' physical activity behaviors, and a lack of experimental research limits our understanding of how trails (routes) may be combined with program-based efforts (destinations) to increase trail use. Because changes in the built environment are rarely amenable to randomized controlled trials, observational studies of built environments, policies and programs that are not under investigator control, often termed "natural experiments", may be needed to evaluate such changes. Evaluation of natural experiments with carefully selected control groups and adequate follow-up can provide strong evidence for causality and are recommended [68]. Fitzhugh et. al. [68] report the results of a natural experiment concerning trails that incorporated many of the recommended standards for assessing causal inference. Studies like this can provide strong guidance for planners and policymakers (see Table A3). This study examined the effectiveness of a newly retrofitted urban trail that was implemented in one neighborhood among many similar neighborhoods with poor connectivity in the community [68]. This trail was designed for a neighborhood lacking connectivity between residential pedestrian infrastructure and nonresidential destinations to increase physical activity among children, adolescents, and adults. The effect of the change in infrastructure was evaluated by comparing the behavior of study participants in the neighborhood that underwent the trail infrastructure improvement with two control neighborhoods that did not undergo any improvement. Some of the recommended design standards used include the following: all neighborhoods had similar characteristics at baseline; intervention and control neighborhoods were well matched; data on all communities were assessed before the intervention; and physical activity was assessed with 2-hour direct observation of trail use. This trail infrastructure improvement significantly increased physical activity in the intervention neighborhood, while physical activity declined significantly in control neighborhoods. The pre-and post-intervention improvements for physical activity in the experimental neighborhood compared with the control neighborhoods were statistically significant for total physical activity, walking, and cycling [68]. Other trail and built environment improvements identified in the literature may be considered when developing future interventions to increase physical activity among youth. For example, reducing road traffic is associated with increased physical activity in neighborhoods [87], and park improvements such as adding a new path/trail, have also led to increases in physical activity [49].
Another example of the type of research that could advance knowledge related to trail use by youth from under-resourced communities are studies that examine trail use in conjunction with Safe Routes to Schools programs and policies. More intervention studies have addressed active transportation to school than leisure or recreational walking by children and youth (see Table A2 for a select sample of studies) [50,51]. However, further research and interventions are needed to address trail use and active transportation together with youth-oriented programs, such as Boys and Girls Clubs, YMCA's/YWCA's, Parks and Recreation program offerings, Community Youth Centers, and after school programs. The body of research pertaining to active travel to schools can serve as a model for the type of research that is needed to advance knowledge about how to effectively connect trails with programs for youth that, in turn, promote and facilitate increased use of the trails themselves.
An example, of how a "program" related to active travel to school was incorporated into a comprehensive community intervention illustrates the type of research that can lead to effective trail use programs in under-resourced communities. This study, in Jackson Michigan, was supported by a Robert Wood Johnson Foundation's Active Living by Design grant. Jackson, a community of 35,000 residents (with 15.2% of families below the poverty level) developed a community intervention using the Active Living by Design's 5P Model (Preparation, Promotion, Programs, Physical Projects, Policy). The intervention included partnership building and a multilevel community intervention that focused, in part, on creating safe transportation routes to encourage walking and biking to school and to worksites. Baseline measures were obtained in 2003 prior to intervention implementation and again post-intervention in 2005/2006. Findings indicate that there was a change in attitudes toward active transportation (i.e., 8% increase in children who thought walking to school was "safer" post-intervention), intentions to try active transportation (i.e., 43% of Smart Commute Day participants "would" smart commute more often post-event), and, importantly, increased physical activity (i.e., the percentage of students walking to school more than doubled at three of four intervention schools). In addition, a community level observational study was conducted at 10  In August 2018, the Community for Preventive Services Task Force recommended interventions to increase Active Travel to School, such as the Safe Routes to School (SRTS) program, based on sufficient evidence of effectiveness that the interventions increase walking among students and reduce risks for traffic-related injury [83]. The Task Force also found that the economic benefits exceed the cost for active travel to school interventions. Active travel to school interventions were made possible, in part, due to funding opportunities that may also be needed to support other programmatic activities to increase routes or trails use among youth. In the U.S., the Safe Routes to School program launched in 2005 resulted in more active transportation to school with fewer injuries [90]. Safe routes to school projects are currently eligible under the Transportation Alternatives Set-Aside and the Surface Transportation Block Grant Program for funding. SRTS funds may be used for infrastructure-related projects, which may include sidewalk improvements, off-street bicycle and pedestrian facilities (for example, trails), traffic-calming measures, bicycle lanes, and bike racks [83,84], and for non-infrastructure related projects, which may include student and parent education, public awareness campaigns, and traffic enforcement [82]. The Federal Recreational Trail Program provides funds to the States to develop and maintain recreational trails and trail-related facilities that can support additional trail infrastructure in under-resourced areas [83]. The Federal Recreational Trail Program may be one source of support for bridging trails with programs promoting their use by youth in under-resourced communities [82]. Such infrastructure-related projects could also be useful targets for further research addressing how to increase trail use and physical activity in under-resourced populations [91].

Discussion
The primary purpose of this review was to identify effective programs for increasing trail use among youth from under-resourced communities, in order to address disparities related to the underrepresentation of trail users from different racial and ethnic groups and/or low-income neighborhoods and communities more broadly. Unfortunately, this review did not find studies evaluating programs that promoted the use of trails among youth from under-resourced areas. In fact, the review highlights a lack in our knowledge on how to promote and increase the use of trails by youth regardless of their socioeconomic status or where they reside. Connecting safe (pedestrian, bicycle, or public) transportation systems to destinations or places where youth reside in programs where they may congregate to use trails to pursue outdoor recreational opportunities and enjoy nature has not been systematically evaluated. It is possible that such programs exist and have just not been studied using experimental designs and methods. This highlights a critical research gap and lack of knowledge about how to intervene to connect youth, and/or programs for youth, to trails to increase their physical activity as a health-enhancing behavior.
To address this research gap, it is recommended that natural experimental research designs be considered to provide insights into possible causal factors that may link youth programs to increased use of trails. Recent workshops and reports of evidence-based reviews, focused on enhancing methods for the evaluation of natural experiments, have provided a clear set of recommendations (e.g., intervention and control neighborhoods well-matched, baseline characteristics similar at intervention outset) on how to improve current research designs and the data resources needed for such research. Furthermore, a September 2018 Federal Partners meeting organized by the National Institutes of Health, Office of Disease Prevention addressed approaches to enhance collaborative efforts in the area of the evaluation of the effects of obesity policy and programs across federal health, transportation, housing, environmental and park agencies-highlighting the need for cross-sectoral work to achieve progress in this arena [92]. These workshops, reports, and meetings hold promise for advancing our knowledge related to trails use among different segments of the populations and addressing disparities among trail users.
Active travel to schools, such as Safe Routes to Schools, has been found to be effective in increasing active commuting to schools among youth. The body of research that represents the evidence for the effectiveness of active commuting to schools or Safe Routes to Schools is also highlighted in this review as an example of what may advance science and knowledge about attracting more youth participants, including those from under-resourced neighborhoods and communities, to use trails for recreational physical activity.
Since trail features themselves may influence trail accessibility and use among youth, a secondary purpose of this study was to identify trail features that may be considered during efforts to connect safe trails to programs that promote their use. Studies have found that it would be beneficial for such programs to be in close proximity to trails [93][94][95][96].
It may also be necessary to provide safe active commuting (walking, bicycling or public) transportation system options to connect with trails that are not in close proximity to youth programs. Trails are one form of the built environment that may influence recreational and transportation physical activity among youth and adults, and this review did identify a variety of built environment features that may impact trail use, such as access to and proximity of trails.
Features of destinations that are attractive among youth may also inform the planning of programs to promote trail use among this population. For example, parks, greenspaces, and open spaces may be enhanced by programs to accommodate youth recreational trail users and their connectedness to opportunities to recreate and enjoy nature. Providing lighting, improving perceptions of, and actual safety [38], and increased accessibility are examples of features important to both trails use and destinations where youth programs are located.
Other secondary purposes of the review were to identify the benefits of, and barriers to trail use that may inform the development of programs for our population of interest. Insights into the barriers associated with using trails, or sidewalks, include neighborhoods with high crimes, automobile traffic congestion making walking unsafe [75], lack of exposure and opportunities to access trails, costs, lack of transportation, lack of role models using trails, and institutional discrimination [51]. It will be beneficial to plan in advance how to address barriers that deter youth from under-resourced areas from using trails for physical activity. These efforts will most likely benefit greatly by programs being centered in destinations where people from different races/ethnicities, backgrounds and cultures intermingle and socialize, since people who do not see role models that look like them attending physical activity programs or pursuing recreational opportunities will be unlikely to initiate involvement and/or sustain participation in recreational physical activities.
Programs designed to support trail use among youth, including youth from under-resourced communities, may also benefit by capitalizing on promoting the social, mental, and physical health benefits that can be derived from trails use, outdoor recreational pursuits, and enjoying nature. Doing so may optimize program participation and success. One study was located that evaluated health benefits in our review reporting a psychological outcome, indicating that youth having greater access to greenspaces also had higher self-esteem and health-related quality of life. There is a need to evaluate the benefits of trail use among youth that may be associated with increases in physical activity as a part of outdoor recreation or connectedness with nature.

Conclusions
The primary aim of the paper was to identify evidence-based trail studies that promote and/or increase trail use among youth from under-resourced communities. Few intervention studies using trails to increase physical activity among under-resourced youth were identified in this review. Clearly, more studies need to be conducted using access to trails as interventions to promote trail-use among youth. In most of the adult trail studies using trails as the intervention tool provide limited insight to this specific segment of the population. How to increase the use of trails among youth for purposes other than active commuting to school, such as using trails to enjoy nature and outdoor recreational pursuits, is unknown and needs further attention. Planning, implementing and evaluating the use of trails to increase physical activity among youth, and programs specifically designed to facilitate the use of trails for outdoor recreation, could benefit from greater attention from researchers and practitioners in the future. Efforts to directly address disparities related to trails use among youth from low income, often racially/ethnically diverse, under-resourced neighborhoods and communities are especially needed.
The recent reports from the Community Preventive Services Task Force [4] about the influence of the built environment on physical activity and on active travel to schools provide evidence of just how rapidly knowledge about physical activity and the built environment is evolving. These reports highlight the increased rigor of research designs now in use to evaluate the effects of the built environment on physical activity behaviors. However, systematic reviews related to physical activity and the environment do not provide knowledge about "how to" implement effective programs that may increase the use of the built environment especially among select population groups of interest, such as youth from low income, diverse, and/or under-resourced communities.
The Task Force physical activity-built environment recommendation noted the need for ensuring that a transportation system (e.g., pedestrian trails, bicycle routes, or public transit) connect to the built environment and land use destinations, such as a facility housing a program (e.g., Boys Club and Girls Club of America, YMCA/YWCA, or school club). However, this review of the scientific literature did not identify an effective trail use program for youth from under-resourced communities, that is housed in a destination or setting such as a school, YMCA/YWCA, or Boys and Girls Club.

Practical Applications
To make advances in this area, future research and practice efforts are needed to establish programs designed to help more youth enjoy nature and outdoor recreational opportunities. Future research efforts should focus on developing interventions to promote trail use rather than cross-sectional studies limiting causal inferences. This research and practical work should incorporate an evaluation of the intervention and programs' impact on increasing trail use and assessing other outcomes of interest to expand the knowledge base in this under-studied area that can also then be replicated.
Practice-based programs may also provide data on feasibility, even if they have not been evaluated using a well-conducted experimental study design, or published in a peer-reviewed journal. Practice-based programs accompanied by evaluation data may also inform and influence better designed future experimental research. A companion brief based on a review of programs and practices related to trails use among youth from under-resourced communities or neighborhoods, is being developed and when completed will be available on the NCCOR website (https://www.nccor.org). We also thank Van Do, Amanda Sharfman, and Miranda Brna from FHI 360 for their substantial organizational and logistical support to the Trails Use Work Group, that helped lead to the development of this paper. Finally, we would like to acknowledge Caroline Stanton and Zoë Rogers for their assistance with the cross-validation identification process for this review. The findings and conclusions in this paper are those of the authors and should not be construed to represent any official USDA, NIH, CDC or U.S. Government position, determination or policy.

Conflicts of Interest:
The authors declare no conflict of interest.  use and features of transportation systems (routes) (e.g., access to trails, walking paths, bicycle paths, sidewalks) that may inform trail use programmatic efforts to increase physical activity among youth.

Pathways [26] 1999-2011 Review
Pathways for bicycles and walking were more likely to promote regular physical activity in children.

Physical activity
Trail [30] 1989-2004 Review Having "highly walkable" neighborhoods were more likely to encourage people to walk for transportation. For people who reported using the trail in this study, over half of them increased their average amount of walking. Younger adults, men, White, and those with some graduate school education were more likely to perform physical activity on the trail. Higher PA was also associated with usage in cooler months, travelling to the trail in a motorized vehicle, and going with others.
Section A. Correlates of physical activity/trail use and features of transportation systems (routes) (e.g., access to trails, walking paths, bicycle paths, sidewalks) that may inform trail use programmatic efforts to increase physical activity among youth.

Routes Citation Year(s) of Data Data Collection Methods/Measures Results
Trail-proximity [41] 1980-2008 Review The further a trail was from a participant's home, the less frequently it was used. Education and income were both positively associated with trail use. When observing which factors influenced the physical activity levels of Brazilian children, it was found that paved streets, bicycling lanes, and proper street lighting were important features that increased the frequency of activity.
Physical activity

Parks Playgrounds
Playing fields [28] 2004-2008 SOPARC 45 parks in a Southeastern community were evaluated to identify the activity settings used by boys and girls. There were fewer girls observed in the parks than boys, but the most commonly used structure by both genders was playgrounds. Playing fields were popular among both, but more frequently visited by boys. This setting was associated with greater vigorous physical activity. Observations of seven of eight activity settings in the 45 parks indicated a greater frequency and percentage of white youth observed in comparison to minorities. Having a park playground within 1 km of the home significantly decreased a child's odds of being at risk or overweight. These children were five times more likely to be a healthy weight than children who did not have a playground in the nearby park.

MVPA
Healthy weight / Overweight Greenways [74] 2017 Adapted Safe Routes to School survey Greenways make it easier for students to participate in active transportation via walking or cycling due to the improved infrastructure conditions. Active transportation Table A1. Cont.
Section B. Correlates of physical activity/trail use and features of the built environment and land use destinations (e.g., parks, park enhancements, greenspaces, open spaces) that may inform trail use programmatic efforts to increase physical activity among youth.

Home School
Green spaces, including parks, wooded areas, and vacant land [96] 2012 Accelerometers GPS data loggers Questionnaires This study evaluated the MVPA of rural, urban, and suburban students. Rural students were more likely to get most of their MVPA at school, but were three times less likely to partake in this activity than urban students. For both urban and suburban students, active commuting made up a large percentage of their average MVPA. MVPA among youth occurs at locations other than home, school, or through active transportation. Green spaces, including parks, wooded areas, and vacant land are also locations where youth attained MVPA.
Parks [97] 1994-1995 Data from Wave I of National Longitudinal Study of Adolescent Health; GIS The availability of parks was associated with benefits such as increased participation in active sports. Females were more likely to participate in wheel-based activities, such as cycling.

PA: Sports participation; Cycling
Parks Park enhancements-Fitness zones [98] 2000 SOPARC Fitness Zones were added to some parks and not added in others (control). Usage of the park increased with the addition of the new equipment, and there are higher estimates of energy expenditure.

Park usage; energy expenditure
Park enhancements-new equipment added Amid decrease in programming activities [99] 2003[99] , 2004[99] , 2006[99] , 2008 This study evaluated the effect of adding park improvements. Contrary to expectations, it was found that physical activity decreased in both parks with new equipment and those without. This decline was attributed to a decrease in programming during that period.
Physical activity PA and BMI z-scores Routes: Neighborhood bicycle or walking trails Ease of walking or biking to transit Destinations: Neighborhood safety Neighborhood aesthetics (more trees, interesting things to look at, and lack of garbage or litter) Number of activity facilities Destinations of interest within walking distance of homes Parental permission to walk in the neighborhood, take public transportation, or walk/bike to transit [21] Youth Survey Examined the test-retest reliability of a survey designed for youth to assess their perceptions of physical environmental factors and transportation. Evaluated the associations of these perceptions with both physical activity and active transport to school Neighborhood walking trails; destinations within walking distance from home, safety, aesthetics, facilities, and being able to walk or bike for transportation were related to physical activity, and in some cases to Active Transport (ATS) Neighborhoods with parks that contain sports fields and multi-use paths were found to significantly increase the average daily MVPA of children outside of school. This can be partially attributed to the fact that paths allow for active transportation.    This study highlights the results of a Robert Wood Johnson Foundation's Active Living by Design grant that supported partnership building and a multilevel community intervention in 2003 in Jackson, MI (Jackson has more families below the poverty level (15.2%) than the average in Michigan and the United States; race/ethnicity demographics not reported). The safe physical activity project, which focused, in part, on creating safe transportation routes to encourage walking and biking to school and to worksites, was shown at follow-up (2005/2006) to change attitudes toward active transportation (8% increase in children who thought walking to school was "safer" post-intervention), intentions to try active transportation (43% of Smart Commute Day participants "would" smart commute more often post-event), and increased physical activity (the percentage of students walking to school more than doubled at three of four intervention schools). In addition, a community level observational study was conducted at 10   Door-to-door surveys; Self-report Direct Observation; SOPLAY A 6-block walking path and a school playground were installed in a low-income, largely African-American intervention neighborhood, in New Orleans, LA. Physical activity levels were measured in the intervention neighborhood and in two matched comparison neighborhoods by self-report, using door-to-door surveys, and by direct observations of neighborhood residents outside before and after the interventions. Self-reported physical activity increased over time in most neighborhoods. The proportion of residents observed who were active increased significantly in the section of the intervention neighborhood with the path compared with comparison neighborhoods. There was no significant pre-playground intervention post-intervention findings noted for physical activity measures. Path use-walking and biking; Physical activity A natural experiment, pre-test post-test quasi-experimental non-control design Conversion of a rail trail to a multi-use trail [65] 2000

MVPA
Telephone survey of adults ≥18 years. Survey questions asked about: Leisure activity Walking and bicycling Moderate and Vigorous Physical Activity Transportation Activity Trail Use A railway of >23 miles was under development for conversion to a multi-use trail. A segment of the trail was evaluated by randomly selecting and telephone interviewing adults living within 2 miles of the planned trail before trail construction began and approximately 2 months after completion of construction. For the total population, living nearby the part of the trail segment evaluated, 41.2% of the residents were black and 47.3% were white. At follow-up, 11.0% of respondents had not heard of the trail, and 23.1% had heard of the trail and had used it at least once. Leisure activity, leisure activity near home, moderate activity, vigorous activity, and walking for transportation did not significantly change for those who used the trail compared to those not using the trail. This prospective study of the building of a multi-use trail did not demonstrate an increase in physical activity among adults. Trail use and other physical activity The Sydney, AU Road and Traffic Authority (RTA) completed the construction of a 16.5-km-long Rail Trail cycleway in Western Sydney in December 2000. A short-term local promotional campaign to increase awareness and use of the Rail Trail was conducted. The media campaign included local press advertisements about the newly constructed Rail Trail, and maps of the trail in English and six other languages. There was a significant increase in unprompted Trail awareness, but post-campaign awareness was low (34%). Pre and post-campaign telephone surveys were conducted with adults 18-55 years of age living in an "inner" area, within 1.5 km of the Trail, and an "outer" area, bike-owners only, 1.5-5 km from the Trail. Trail usage was higher among bike-owners than pedestrians and was moderated by proximity to the Trail. The campaign reached and influenced cyclists in the inner area. Inner cyclists increased mean cycling time by 0.19 h (SD _ 1.5) while outer cyclists decreased cycling, time (_0.24 h, SD _ 1.6).

Routes
Way-finding, and incremental distance signage along trails [66] 2011-2012 Infrared monitors placed on the trails for three periods of 7 days, during a 1-year timeframe.
Comparisons were made between pre-, mid-, and post-intervention usage rates on six trails where signage was added, to usage rates on four control trails. The signage was added to trails after a previous marketing campaign was used to effectively promote and increase trail usage on 8 of 10 trails studied. Wayfinding and incremental distance signage were associated with about a 33% increase in trail use (mean users per hour increased 31% for control trials and 35% for the trails with signage), but the total increase did not vary between the intervention and control groups. The marketing campaign intervention that occurred prior to the wayfinding and distance signage intervention that followed, may have influenced findings.