Co-benefits of designing communities for active living: an exploration of literature

To reverse the global epidemic of physical inactivity that is responsible for more than 5 million deaths per year, many groups recommend creating “activity-friendly environments.” Such environments may have other benefits, beyond facilitating physical activity, but these potential co-benefits have not been well described. The purpose of the present paper is to explore a wide range of literature and conduct an initial summary of evidence on co-benefits of activity-friendly environments. An extensive but non-systematic review of scientific and “gray” literature was conducted. Five physical activity settings were defined: parks/open space/trails, urban design, transportation, schools, and workplaces/buildings. Several evidence-based activity-friendly features were identified for each setting. Six potential outcomes/co-benefits were searched: physical health, mental health, social benefits, safety/injury prevention, environmental sustainability, and economics. A total of 418 higher-quality findings were summarized. The overall summary indicated 22 of 30 setting by outcome combinations showed “strong” evidence of co-benefits. Each setting had strong evidence of at least three co-benefits, with only one occurrence of a net negative effect. All settings showed the potential to contribute to environmental sustainability and economic benefits. Specific environmental features with the strongest evidence of multiple co-benefits were park proximity, mixed land use, trees/greenery, accessibility and street connectivity, building design, and workplace physical activity policies/programs. The exploration revealed substantial evidence that designing community environments that make physical activity attractive and convenient is likely to produce additional important benefits. The extent of the evidence justifies systematic reviews and additional research to fill gaps. Electronic supplementary material The online version of this article (doi:10.1186/s12966-015-0188-2) contains supplementary material, which is available to authorized users.


Introduction
As demonstrated in Designed to Move (www.designedtomove.org), physical inactivity is a major challenge to health and well-being globally. There are 5 million deaths worldwide attributable to physical inactivity (Lee, et al, 2012), most people are not sufficiently active, physical activity is declining in many countries, and this is a global problem with the biggest burden in low and middle income countries (Ng & Popking, 2012). Increasing physical activity is a goal of the United Nations through its non-communicable disease initiative (UN, 2011).
Physical activity has been engineered out of people's lives, but efforts are underway worldwide to re-integrate physical activity into daily lives through environmental changes. The UN, World Health Organizations, national physical activity plans, US Centers for Disease Control and Prevention, US Institute of Medicine, and other scientific groups worldwide have identified creating built environments and implementing policies that support active living as essential for increasing physical activity and improving health.
There are special opportunities to create more activity-friendly environments through actions taken by city governments. Actions at the city level can affect urban design, transportation systems, parks and recreation facilities, schools, and workplaces/buildings. These actions are often led by mayors. Thus, we look to mayors for leadership in creating activity-friendly environments, and mayors around the world are showing leadership and achieving results.
An important barrier to environmental changes that support physical activity is that decision-makers do not consider physical activity an important enough rationale to justify policy changes and investments. Even if a decision-maker is well aware of the important effects of physical activity on health and health care costs, health may not be on the agenda of the decision-maker. Decisions taken in the transportation, city planning, parks and recreation, education, architecture, and business fields often determine whether environments facilitate, or create barriers to, physical activity. Health is often not part of the mission of these groups. Though mayors, city councilors, and similar officials work everyday to balance competing interests, most of them do not understand that environments that support physical activity produce additional benefits for their communities. Thus, creating activity-friendly environments can be a solution that solves multiple problems. For example, changing zoning codes to favor mixed use developments is likely to enhance property values and reduce carbon emissions. Having parks in neighborhoods has been linked with physical health and mental health benefits.
However, there is no resource that describes the full range of co-benefits of activity-friendly environments. We undertook an exploration of an extremely diverse literature ranging from mental health and injury prevention to real estate prices and climate change. We searched in the scientific literature and beyond. We asked for guidance from scholars and advocates working in diverse fields. We looked hard for international evidence and dug into "green" benefits that related to pollution and carbon emissions. The hunt uncovered hundreds of relevant studies and reports that can be used to create a profile of the multiple benefits of each environmental feature. We believe this report can be a useful decision tool. Because mayors and city councilors must address a very broad set of issues, knowing how a single decision can affect several outcomes may change that decision. Built environment decisions often involve large financial commitments, and they create the form and character of cities that last for decades, even centuries. Decisions about built environments should consider the full range of consequences, including physical activity, and this report is meant to be a tool for decision-makers.
The main audience for this report is mayors, city officials, practitioners from multiple sectors, and advocates. The goal is to summarize the available science in a non-technical way so non-scientists can use the evidence in their decisionmaking.

Making the Case Methodology
Active Living Research's (ALR) primary task was to identify physical activity-related built environment features, then conduct a literature review to examine multiple potential benefits across a wide range of sectors, including chronic diseases, injury, mental health, social benefits, economic benefits, automobile congestion, air pollution, and carbon emissions. Thus, the focus was on potential benefits of each built environment feature, in addition to physical activity. The findings of the literature review were summarized in tables, figures, and quotes to aid the reader in understanding the level of evidence for multiple benefits of built environment features that can support physical activity

Literature Review
We began the literature review with a snowball sampling of topic experts. Using our ALR network of diverse experts, we asked 20 leaders with expertise in the five DTM settings for their help in identifying: • Groups/organizations working on these issues worldwide • Must see reports/papers, both peer-reviewed and grey literature • Websites • Cities that have implemented activity-friendly built environment changes • Recommendations for other experts in the field we should speak with Of the 20, we received input from 13 experts (Table 3). From November 2013 through February 2014, ALR staff reviewed the recommended resources, looking for evidence focused on our identified built environment features and cobenefits. Some of the recommended resources focused specifically on physical activity as an outcome and were disqualified from the abstracting. ALR staff used the recommended resources to find additional papers and reports using the citation lists. Both peer-reviewed and "gray" literature and reports were used. To fill gaps, staff also performed

Abstracting Process
During the abstracting process, basic information on the built environment feature, co-benefit, study sample characteristics, study methods, and major research findings were listed in tables specific to each DTM setting (see Excel spreadsheet titled "Data Abstracting Table" at http://activelivingresearch.org/making-case-designing-active-cities). Staff graded the scientific rigor of each piece of evidence (Table 4) and noted the direction of association of the benefit with the environmental feature (Table 5). To simplify interpretation, "+" denotes an environmental feature is associated with a co-benefit in a "favorable" direction, which could be better mental health or fewer carbon emissions. As much as possible, quantitative findings were included so as to offer specificity in terms of impact of the built environment feature on the co-benefit. ALR staff were instructed to also look for findings relevant to international examples, disparities by income or race/ethnicity, and public opinion polling to include in the abstracting process. Public opinion polling was included because such data are relevant to decision-making by government leaders. However, it was not feasible to grade the quality of the studies, as is done in systematic reviews.

+
A favorable association was found between feature and co-benefit (feature associated with "better" level of co-benefit -An unfavorable association was found between feature and co-benefit (feature associated with "worse" level of co-benefit 0 (zero) No association or inconsistent evidence was found between feature and co-benefit The initial goal was to be inclusive in finding relevant sources of information. For scientific literature, reviewers were instructed to find systematic or non-systematic reviews first. If reviews were located, then the individual studies did not need to be searched, except for publications since the latest review. In cases where a review paper did not provide adequate specificity or quantification in the findings, selected primary studies from that review paper were abstracted to illustrate specific findings. If reviews were not located, then individual studies were searched. For gray literature, reports from credible organizations were targeted, from such groups as government agencies, academic centers, and selected advocacy groups. Newspapers, magazines, and blogs were not searched, except to identify citations of or links to more credible reports. Abstracting tables were cross-checked by other staff for accuracy and clarity.
To ensure a global perspective, ALR contracted with two international consultants to replicate the literature review process as well as provide additional expertise. Nick Cavill, PhD and Mike Parker of Cavill Associates, a public health consultant firm based in the United Kingdom, conducted a search of grey European literature, uncovering and abstracting a total of 46 reports. Ding Ding, PhD, MPH of the University of Sydney and Klaus Gebel, PhD of James Cook University, Australia, conducted a literature search focused on the association between the built environment features and climate change related outcomes, and case studies of successful programs, policies, or planning initiatives that were successful in reducing carbon emissions. Ding and Gebel added a total of 31 reports to the overall MTC abstracting.

Synthesizing the Findings
On completion of the abstracting process, ALR staff and the international consultants selected 'key findings' that they judged to be good examples for illustrating findings and providing specific quantitative results. These key findings were identified based on relevance and quality of the study or finding. The key findings are used as evidence and examples in this overall summary report.
Communicating so many findings is a challenge, so several ways of summarizing the results in a visual manner were created. To illustrate areas with strong research evidence as well as research gaps, ALR staff created summary matrices that visually show which built environment features have (and do not have) evidence supporting a co-benefit (see Table  6 for example; for full document, see Excel spreadsheet titled "Summary Matrix" at http://activelivingresearch.org/ making-case-designing-active-cities. Each finding from the Data Abstracting tables is listed in the appropriate cell along with its associated grade of evidence and direction of association. Individual papers that were counted as part of a review paper, and therefore also abstracted, are coded in red on the Summary Matrix to help prevent double counting of findings. The Summary Matrix helps to identify specific policy and built environment changes that cities can pursue within a DTM setting. An overall "Making the Case" summary table identifies which DTM settings have the most overall evidence. The table was calculated by aggregating the "grading the evidence" scores for each built environment feature for each co-benefit within each DTM setting (see Tables 8 -12). Dark green cells have the highest levels of evidence, light green cells have moderate evidence, white cells lack sufficient evidence to make a judgment, and pink cells have evidence of unfavorable associations. Thus, decision-makers can see at a glance which environmental features have the best evidence of multiple benefits.

Creating Color-Coded Matrix Tables to Summarize Results for Each Setting
The color-coded "matrix" tables that follow summarize the evidence of co-benefits for each key feature. There is a separate table of features for each of the five built environment settings. Some of the features are programs or social environment attributes that can complement a favorable built environment or possibly compensate for an unfavorable built environment.
We used a "quasi-quantitative" approach to summarizing the results of the literature reviews. Summary scores were created by summing the weights of evidence from each resource. Thus, the summary scores roughly indicate both the quantity and quality of the evidence. Cells in each table are color coded based on three levels of summary scores. We judge a positive finding from a literature review to be moderate evidence, because if coders found a good-quality review, they did not necessarily continue searching for specific studies. We did not want to set the bar too high, because we expected many of these topics not to be well studied. Thus, one high quality study does not merit a summary of "moderate" evidence, but a good review is considered "moderate" evidence.
We calculated "net" scores by subtracting the weight of negative and zero scores from the weight of plus scores (Table 7). Thus, evidence of negative or null findings were counted against positive findings. The color-coded "matrix" tables can be interpreted as follows: • "4+" net score indicates moderate evidence (light green), because this is equivalent to one non-systematic review. • "10+" net score indicates good evidence (medium green), because this is equivalent to more than two reviews.
• "15+" net score indicates strong evidence (dark green), because this is equivalent to more than three reviews.
• A net score of less than 4 indicated insufficient evidence and was coded with a white background.
• The same rules were used to summarize negative and null findings, with net negative or null findings coded in shades of red. • A white cell with no numbers indicated that we found no credible evidence on this topic.
• In creating the summary scores, resources with quality scores of "1" or "2" were not counted because they were not considered credible enough.

Summary of Results by Setting
There are 3 parts of the presentation of results for each setting.  Table", to give a flavor of the types of studies and in some cases, to quantify the strength of associations or effects. See the "Data Abstracting Table" for full citations of items listed as key highlights (available at http://activelivingresearch.org/making-casedesigning-active-cities).

Open Spaces / Parks / Trails
There were 69 entries in the open space/parks/trails category. Of 36 cells in the matrix table (Table 8), 3 had strong evidence of co-benefits, 3 had good evidence, and 7 had moderate evidence. Specific findings include: • Park presence/proximity had good to strong evidence of all co-benefits, except economic.
• Physical activity programs and promotions had moderate evidence for 4 co-benefits.
• Public gardens had moderate evidence of social and injury prevention benefits.
• Trails had good evidence of economic benefits.
• There are many gaps in research on co-benefits of all parks and trails features, except park proximity and physical activity programs and promotion.
Current evidence supports a conclusion that having a park nearby with substantial programs and promotion produces a wide range of health and environmental benefits beyond physical activity.

Mental Health
a. Presence/Proximity: People living more than 1km away from a green space resulted as having 1.42 higher odds of experiencing stress than those living less than 300m from a green space. An analysis of the association between stress and visits to green spaces showed that respondents who do not report stress have 1.57 (95% CI 1.40-1.76) higher odds of visiting a green space at least a few days a week than those reporting stress. The results also showed that the more often respondents visited green spaces, the less stress they experienced. (Stigsdotter, 2010) b. Presence/Proximity: 8 studies found that separation from nature via modern living is detrimental to human development, health, and wellbeing and that regular contact with nature, such as provided by parks, is required for mental health. (Maller, 2008) c. Presence/Proximity: Scientists in the Netherlands examined the prevalence of anxiety disorders in more than 345,000 residents and found that people who lived in residential areas with the least green spaces had a 44 percent higher rate of physician-diagnosed anxiety disorders than people who lived in the greenest residential areas. The effect was strongest among those most likely to spend their time near home, including children and those with low levels of education and income. In addition, the prevalence of physician diagnosed depression was 33 percent higher in the residential areas with the fewest green spaces, compared to the neighborhoods with the most. (NRPA Report, 2010) d. Presence/Proximity: In one study, researchers found that a 20-minute walk in an urban park benefited concentration performance of children with ADHD. The performance boost was at least as large as the peak performance boost shown for two widely prescribed ADHD medications. Researchers aren't the only ones who have tracked these benefits. In a nationwide survey, parents of kids with ADHD rated after-school activities in outdoor settings as significantly more helpful than those not in these settings.

Social Benefits
a. Presence/Proximity: Significant relation between both the percentage of green in 1 km (p < 0.05) and in 3 km (p < 0.01) radii around people's home and their feelings of loneliness. People with more green space in their living environment feel less lonely. (Maas, 2009) b. Presence/Proximity: Spending time in green outdoor common spaces is systematically related to stronger social integration and a stronger sense of local community. In a study of inner-city older adults in Chicago, Ill., USA, the total amount of variation in sense of local community explained by exposure to green common spaces was 5%. (Kweon, 1998) c. Design Features: As part of Mayor Michael Bloomberg's PlaNYC 2030 and a three-way partnership between the board of education, the department of parks and recreation and The Trust for Public Land, New York City is rebuilding derelict schoolyards for students and opening them to the full community outside school hours. Each playground renovation begins with an in-depth participatory design phase and includes a 3-month student design phase. Following reconstruction of playgrounds, observations revealed a 25% increase in structured games and competition and a 240% increase in unstructured play. Unstructured activity is valuable in helping less athletic children enjoy recreation and develop social skills and imagination. (Harnik, 2011)

Economics
a. Presence/Proximity: The value of an average home increases with proximity to neighborhood parks, special parks, lakes, and rivers, with benefits ranging from a low of 0.0035% of sales price for every one percent decrease in the distance to the nearest neighborhood park, to a high of 0.0342% for every one percent decrease in the distance to the nearest lake. These effects are statistically significant at or near the 1% level. (

Urban Design / Land Use
There were 202 entries in the urban design category. Of 30 cells (Table 9), 8 had strong evidence of co-benefits, 5 had good evidence, and 6 had moderate evidence. In the urban design category, 4 cells had moderate or good evidence of negative effects, and 1 cell had strong evidence of negative effects. Specific findings include: • Mixed use, greenery, street scale design, and connectivity had evidence of 4 to 5 co-benefits.
• All urban design features had strong evidence of green/environmental benefits, except evidence was good for streetscale design. • All urban design features had evidence of economic benefits, and the evidence was strong for mixed use.
• Only greenery had strong evidence of mental health benefits. None had evidence of injury prevention benefits.
• Residential density had the most complex pattern with good evidence of negative health effects, strong evidence of environmental sustainability, and good evidence of economic benefits.
In general, we found very strong evidence of multiple health, environmental, and economic benefits of most of the urban design features. Creating walkable communities (dense, mixed use, connected streets), with substantial greenery, and pedestrian-friendly street designs can help meet multiple goals of city decision-makers. Finding ways to reduce apparent negative health effects of high density remains a challenge to city planners and politicians.

Highlights of Key Urban Design and Land Use Findings
1. Physical Health a. Urban greenness: Patients in hospital rooms facing a park had a 10% faster recovery and needed 50% less strong pain medication compared to patients whose rooms faced a building wall.(Bolund, 1999) b. Walkability: A 5% increase in walkability was associated with a … 0.23 point reduction in BMI. (Frank, et al, 2006) c. Density / Block size: Contrary to expectations, the hypothesised most walkable neighbourhood (high density, small block stratum) had the greatest mean and median BMI. After adjusting for demographic covariates, physical activity and clustering due to neighbourhood, no conclusive effect of population density by block size on BMI was found (β=−1.024, 95% CI −2.408 to 0.359). 4. Environmental Benefits a. Mixed use / Density: Exhaustive review. "When viewed in total, the evidence on land use and driving shows that compact development will reduce the need to drive between 20 and 40 percent…So, as a rule of thumb, it is realistic to assume a 30 percent cut in VMT with compact development.

Transportation Systems
There were 81 entries in the transportation systems category. Of 48 cells (Table 10), 5 had strong evidence of cobenefits, 2 had good evidence, and 6 had moderate evidence. One cell had good evidence of negative impact. Specific findings include: • Pedestrian and bicycle facilities had the best evidence of multiple co-benefits, followed by lower traffic speed and volume.
• Strong evidence of co-benefits was most evident in the injury prevention and economic domains.
• Traffic calming had strong evidence of injury prevention benefits.
• Public transport had strong evidence of economic benefits and mixed evidence of environmental benefits.
• Many cells had inadequate evidence. Even in well-studied topics, there was little study of health consequences of transportation decisions.
Environmental strategies to promote active transportation, provide public transportation, and protect pedestrians and bicyclists from automobile traffic had good to strong evidence of multiple benefits, particularly in the areas of economics, injury prevention, and environmental protection. Physical and mental health and social benefit consequences of transportation systems are poorly studied.  (Gotschi, 2011) b. Traffic volume: A further externality of private motorized transport is traffic congestion that imposes significant costs on economic efficiency as time lost due to congestion reduces productivity. Congestion costs in Canada are as high as US$4.5 billion nationally of which 80% is accounted for by the country's three largest urban regions: Greater Toronto (43%), Montreal (21%) and Vancouver (17%). In the US, congestion has led urban Americans to travel 5.

Schools
There were 27 entries in the school setting category. Of 18 cells (Table 11), 2 cells had strong evidence of co-benefits, 1 cell had good evidence, and 5 cells had moderate evidence. Specific findings include: • Siting schools near the homes of students had strong evidence of environmental sustainability and moderate evidence of mental health and economic benefits. • Having recreation facilities at schools had strong evidence of mental health and good evidence of physical health benefits. • Shared use agreements had moderate evidence of social benefits, injury prevention, and economic benefits.
The co-benefits of school environment features were poorly studied for most outcomes.   (Huesman, 2007). Freshmen who used the SRC (student recreation complex) persisted (didn't drop out) at a greater rate after 1 semester and after 1 year than their counterparts who did not use the SRC. Users earned slightly higher GPAs and more credit hours at the end of the first year. This finding was of particular interest because the study also revealed that the SRC nonusers entered college with higher high school GPAs and higher ACT/SAT scores than SRC users. (Belch, 2000) c. Recreation facilities: Teachers reported that children were more social, creative, and resilient when unstructured play materials were provided on the playground. Children who were creative, rather than very physically capable, became leaders in activity; the unstructured materials promoted play between children who had not played together previously--including children who had formerly been excluded. (Bundy, 2008) 3. Social Benefits a. Joint use agreements: All 68 faculty/staff respondents agreed that the joint use program had a positive impact on the school and that it was beneficial to students. They reported that the program provided needed opportunities for PA, incurred social benefits (e.g., making new friends), kept students out of trouble, and promoted healthy lifestyles. (Maddock, 2008) b. Joint use agreements: Public schools can serve as community "hubs," and offer a range of medical, social, and other support services. (Filardo, 2010) c. Joint use agreements: On-site services that promote intergenerational use, such as health care, job training, or a senior center, give seniors and childless professionals a vested interest in their neighborhood school. (Chung, 2002) 4. Environmental Benefits a. School siting: Neighborhood schools produce less emissions than schools located on outskirts of town. Compared to a sample from existing schools, neighborhood schools would reduce traffic, produce a 13 % increase in walking and biking, and lead to a reduction of at least 15% in emissions of concern. (Ewing , 2003) b. School siting: Emissions are 3−8 times lower for the neighborhood-only scenario as for the current scenario; distance traveled by automobile is more than cut in half. Busing rate drops by more than half and busing distance declines by more than an order of magnitude. (

Buildings / Workplaces
There were 39 entries in the workplace/building category (Table 12). Of 36 cells, 3 cells had good evidence of co-benefits and 3 cells had strong evidence. Specific findings include: • Building site design features (mainly outdoor) had strong evidence of physical and good evidence of mental health benefits. • Features of the building design had strong evidence physical health and good evidence of environmental sustainability and economic benefits. • Physical activity programs and policies had strong evidence of economic benefits. • 5 cells had evidence of moderate evidence.
For workplace and building features, the best evidence was for physical health and economic benefits.  improved productivity and reduced work absence. Additionally those who did not join the fitness center were nearly twice as likely as participants to report health-related work limitations in time management and physical work, and 1.5 times as likely to report limitations in overall output. Assuming an average wage of US$200 per day, it was calculated that nonparticipants cost $258 more in lost work time per employee than participants. (Burton et al., 2005) e. Building design: The "StairWELL to better Health" CDC intervention is a low-cost intervention (less than $16,000) that increases stair users (involves painting and carpeting, framed artwork, motivational signs, and music to the stairwell). Infrared beams were used to track the number of stair users.
(StairWELL CDC) f. Building design: Lighting, particularly daylighting (lighting through daylight) has consistently shown to be important. School children in well lit with daylight received test scores 15% higher than those in classrooms with little or no daylighting. Additionally, daylighting design is linked to a 15% reduction in absenteeism in office environments, and increases between 2.8-20% due to luminance levels have been found in other studies. Due to individual variation and preferences, providing the means of thermal and ventilation control to individuals is ideal. Personal control over temperature, air supply, sound masking and lighting are associated with a 3% increase in productivity. (CABE, 2004) g. Building design: In a case study in CA, Thayer identified 50% savings in lighting, cooling and ventilation energy as well as a 15% reduction in absenteeism, due to an integrated daylighting design that harmonized layout, orientation, and windows. (CABE, 2004) 2. Mental Health a. Building design: High-rise housing is associated with elevated psychological distress, especially among low-income mothers. The quality of the neighborhood, in terms of social and physical attributes, are associated with greater psychological distress and poorer cognitive development in children in lower quality neighborhoods. (Evans, 2003)

Physical Health (Chronic Diseases)
a. Worksite health program: Comprehensive worksite health promotion programs, that include 1) training and support of coordinators; 2) health risk appraisals; 3) on-site classes, safety meetings, and self-help options; 4) environmental changes, e.g., smoking policy, cafeteria offerings, and blood pressure machines; and 5) recognition, decreased the number of employees with three or more risk factors by 14% and risk levels improved among high-risk individuals in 6 of the 7 risk factors (i.e. BP, cholesterol, etc.) over two years. (Bertera, 1993) b. Standing desk: A sit-stand workstation paradigm reduces musculoskeletal complaints without considerably affecting data entry efficiency. (Huseman et al., 2009) c. Standing desk: Differences as high as 25% have been attributed to differences in employees who report being comfortable vs. uncomfortable. Even improved workstations with ergonomic chairs are associated with a 23% increase in computer data entry and editing. (CABE, 2004) d. Building design: Air quality studies have shown that buildings that improve delivery of outside air are attributed to a decrease in Sick Building Syndrome (SBS). SBS is related to productivity, and it is estimated that a 10% reduction in reported SBS symptoms is associated with a 1.1% in productivity among employees. (CABE, 2004) 4. Environmental Benefits a. Mixed use: Residents of mixed-use neighborhoods use less motorized travel than those in purely residential areas, where the average distance per trip was half of those in single use areas; also residents used alternative transportation more--12.2% of the time compared to 3.9% of the trips in single use areas. This can reduce VMT by at least 4% for an entire region or, in specific dense sections, VMT can be reduced by as much as 17.6%. (Ewing & Kreutzer, 2006)

Review of Grey Literature in Europe
Cavill Associates, based in London, were commissioned to search the grey (not peer reviewed) literature for studies of co-benefits in the European region. The purpose was to ensure incorporation of international literature, and we expected most such literature to be from Europe. They found 31 reports with relevant information, mostly from advocacy organizations, and they found 59 mentions of co-benefits of activity-friendly environments. Their examples of specific and quantitative findings are included here. Their findings are also included in the matrix tables reported above.  N, et al, 2012) c. Walkability: Walkability affects consumer transport costs. Improved walkability allows consumers to save on vehicle expenses ("Affordability," VTPI 2008). For example, one study found that households in automobile-dependent communities devote 50% more to transportation (more than $8,500 annually) than households in communities with more accessible land use and more multi-modal transportation systems (less than $5,500 annually) (Litman, 2011) 2. Environmental Benefits, including general environmental benefits; improved air quality; reduced noise; reduced energy use and Carbon dioxide emissions; Integration of environmental and health concerns; improved biodiversity, air quality, water management; greater use of green space. a. Cycling as transport: Between 1990 and 2007 greenhouse gas emissions from transport increased by 36% in the EU. By 2050the EU has set a target to reduce green house gas emissions by 80-95% (60% will need to come from transport sector). Emissions from bicycles are 10 times less than those from motorized vehicles. If levels of cycling in EU were equivalent to those of Denmark, bike use would help achieve 12-26% of the 2050 target dependant on which transport mode the bike replaces. 4. Mental Health Benefits, including improvements to stress, wellbeing, mental health and cognitive function. a. Green environments: Compared with exercising indoors, exercising in natural environments was associated with greater feelings of revitalization and positive engagement, decreases in tension, confusion, anger, and depression, and increased energy. However, the results suggested that feelings of calmness may be decreased following outdoor exercise. Participants reported greater enjoyment and satisfaction with outdoor activity and declared a greater intent to repeat the activity at a later date. (Thomson et al, 2011) b. Green space: Green spaces associated with greater wellbeing and mental health and cognitive function.

Co-benefits Identified in European Studies
"Strong body of evidence which suggests that physical activity in green spaces has stronger mental health benefits than physical activity in non-green spaces, and that "more passive forms of usage" can also have a beneficial impact on mental well-being and cognitive function." (BoP consulting, 2013) c. Green Space: There is strong evidence that access to green spaces improves mental health, although the evidence that it increases physical activity is less strong (1). Use of green spaces is associated with: a decrease in health complaints, improved blood pressure and cholesterol levels, reduced stress, improved general health perceptions and a greater ability to face problems (2). Sixty-three per cent of people in social grade AB (managerial and professionals) said they visited green spaces on a weekly basis, compared with 47 per cent in grade DE (manual workers, shop workers, apprentices, casual labourers, state pensioners and the unemployed).

Special Report on Active Living and Climate Change
Some of the most consequential co-benefits of activity-friendly environments are likely to be related to climate change. For example, active transportation can substitute for automobile trips and reduce carbon emissions. More parks and street trees can improve carbon sequestration. The relevant literature is scattered across many disciplines, so we commissioned Ding Ding and Klaus Gebel, based in Australia, to conduct a review of scientific and grey literature on the climate change co-benefits. In their extensive report, they include summaries of findings and identify promising case studies. Their findings have been incorporated into the matrix tables above.

Co-benefits Identified in Climate Change Literature
1. Urban Planning a. Density: Examined determinants of urban travel greenhouse gas emissions based on a travel survey in the Quebec City Region of Canada. They found that compared to residents in the city center (the highest residential density), those leaving in old suburbs, new suburbs, and the periphery (the lowest residential density and farthest from the city) produced 19%, 27%, and 70% more emissions from traveling. Specifically, they also found that if one moves to a neighborhood that is 10% more dense, his/her emission would reduce by 1.2%. (Barla, et al, 2011) b. Walkable neighborhoods: Estimated CO 2 equivalent emissions based on the 2006 Household Activity Survey in the Puget Sound Region of Washington State. The key findings were that households located in denser and well-mixed neighborhoods produce a smaller amount of emissions. In addition, intersection density has an inverse association with emissions. Residential density, entropy (land use mix), and intersection density were inversely associated with CO 2 emissions in both urban and suburban areas. A 100% increase in all three built environment factors could lead to 31.2-34.4% reductions in transportation emissions. (Hong and Goodchild, 2014) 2. Transportation a. Transport and land use policies: The South-western city of Freiburg is considered the environmental capital of Germany. Over the last few decades, a number of transport and land-use policies have encouraged more walking, cycling and public transport use. For instance, an extensive network of bike paths and lanes was created, the city center was converted into a pedestrian only zone, new light rail lines were built and existing ones extended, a land use plan was developed that centers new developments around public transport stops, Germany's first transferable flat-rate monthly public transport ticket was introduced, and thousands of bike parking spots were created in the city. Through these and other measures over the last three decades in Freiburg the number of bicycle trips tripled, public transport ridership doubled, and the share of trips by automobile declined from 38% to 32%.
Since 1990, motorization rates have leveled-off and per-capita CO2 emissions from transport have fallen. Many of the strategies mentioned here would be transferable to car-oriented cities around the world. (Buehler & Pucher, 2011) b. Public transportation (BRT): Estimated household CO 2 emissions in 23 diverse neighborhoods in Jinan, China, based on household survey data collected in 2009-2010. They found that residents living near BRT (bus rapid transit) had 13% lower CO 2 emissions compared to those who did not live near BRT. The difference in CO 2 emissions was likely due to residents' preference to use BRT when opportunities present. In Jinan, BRT offers comfortable and convenient transit conditions, and it is overall energy efficient. (Guo, et al, 2013) 3. Parks and Green Space a. Greenery: It has been well understood that plants can reduce CO 2 concentration in the atmosphere through carbon sequestration, a process of "removing carbon from the atmosphere and depositing it in a reservoir". Trees play a particularly important role in capturing and storing atmospheric CO 2 in vegetation, soils, and biomass products. Resent estimates suggest that trees and forests in U.S. urban areas store 643 million tons of carbon, and the total annual gross carbon sequestration was estimated to be 25.6 million tons. (Nowak, et al, 2013) b. Greenery: In 1995, in order to promote multiple environmental and health benefits, Taiwan's government started to establish air quality enhancement zones (AQEZs). By 2010, a total of 504 AQEZs had been established, including greening landfill restoration sites, bare-land sites, waste disposal sites, green open spaces, metropolitan bike paths, and metropolitan parks. The main approach of the greening process was tree planting. In 2010, the investigators sampled and surveyed 9761 trees from 48 AQEZs and found that the total forest carbon storage in the 48 sites was 296.73 tons and the annual sequestration was 157.64 tons of CO 2 per year. Given that most trees were still young, carbon sequestration by the trees will increase over time. AQEZs provide significant and direct benefits to urban areas. (Wang & Lin, 2012)

Conclusions
When the results from all features are combined (Table 13), there is impressive evidence of co-benefits in all physical activity settings. When settings are designed to be activity-friendly, there is good reason to expect not only more physical activity, but also other health, environmental, and economic benefits. This table represents 418 findings.
Of the 30 cells in the matrix, 22 have strong evidence and two had good evidence of co-benefits. Five cells had inadequate evidence, and only 1 cell had evidence of a net negative effect. Specific findings include: • Each setting had strong evidence of at least 3 of the 6 co-benefits, and parks and trails had strong evidence of all 6 co-benefits. Thus, for each setting there are multiple features that can be designed to both facilitate physical activity and produce co-benefits. • Of particular relevance to mayors and other government leaders responsible for balancing budgets, activityfriendly environments had strong evidence of economic benefits. A broad range of economic benefits was documented, such as increased home value, greater retail activity, reduced health care costs, and improved productivity. • Activity-friendly design in all settings had strong evidence of environmental co-benefits based on reduced pollution and carbon emissions. • There were many gaps in evidence of co-benefits in the schools and workplace settings.
• Another major gap was evidence of the health consequences of environments that support active travel.
• There was little evidence of negative consequences of activity-friendly environments. However, in the urban design setting there was some evidence of negative physical health and injury outcomes, mainly related to high residential density. However, the overall pattern of results indicated overwhelmingly positive effects for numerous important outcomes from activity-friendly environment designs. We used the same criteria for level of evidence for this summary table as we did in the setting-specific tables. Though it was "easier" to generate positive (dark green) findings in the summary table, we believe the approach is justified. Within each setting there are several ways (features) to create activity-friendliness. If there is good evidence that one of those features is related to the co-benefit, then that strategy can be used to improve that setting. If several features with weaker evidence can be applied to the design of a setting, it is reasonable to conclude this constitutes strong evidence for the overall design of the setting. The fact that the 3/4 of the cells in the summary table had strong evidence simply indicates there is good evidence of multiple co-benefits in all settings, and several features can be used to achieve the cobenefits.

International Evidence on Co-Benefits of Physical Activity-Friendly Environments
Because coders noted the country of origin for specific studies and reports, it is possible to summarize the evidence by country (Table 14). This is important because physical inactivity is a global problem, and it is useful to know the extent to which the findings about co-benefits are similar across countries. Though we were not able to conduct separate analyses by country, we can summarize the extent to which the evidence is international. The table tallies the number of findings by country and for each grade of evidence.
Most of the international evidence came from review papers (evidence grade of 4.5 and 4), for which it was not possible to identify all the specific countries. The mix of non-US and US studies varies by review paper, and it was not possible to determine the international contribution. Most of the country-specific findings were in specific studies (evidence grade of 3.5).
Though worldwide coverage was less than desired, and international findings were dominated by the English-speaking world, half of the findings included international studies. Our inability to search in languages other than English certainly reduced our ability to identify international literature.

Public Support for Physical Activity-Friendly Environments and Related Policies
Public opinion is an important input into the decision-making of public officials in a democracy, in part because officials are elected to represent the will of their constituents. Public opinion polls are a common political tool, and there is evidence from surveys in the US that many people want more and better physical activity-friendly environments. In this section we summarize a few polls and studies of people's preferences.
Two large national surveys in the US (about 6000 and 12,000 adults) assessed support for walkable communities in their areas and assessed trends from 2003 to 2005. In 2003 there was substantial publicity about studies showing people living in suburbs were less active and more likely to be obese and have hypertension than those living in urban, walkable areas. In 2003, 44% expressed support for building more walkable communities where they lived. This support increased significantly to 59% in 2005. Support was strong among all groups except rural residents and increased among all groups. Public support for developing walkable communities is strong, widespread, and growing. Walkable communities had appeal because they were seen as child-and elderly-friendly places (Handy, et al, 2008).
A recent random sample survey of 1200 adults in the US was conducted by the Urban Land Institute, an organization of real estate developers. They found more than half of Americans prefer neighborhoods that are close to shops, have a mix of incomes, and have public transportation. Neighborhoods that are close to a mix of shops, restaurants, and offices are especially appealing to African Americans (75%), members of generation Y (62%), single people (60%), renters (60%), and college graduates (60%). Over 60% of Americans planning to move in the next five years would prefer to settle in mixed-use communities. Many want communities that have access to public transportation (52%) (Urban Land Institute, 2013).
Though the surveys above demonstrate strong support for walkable mixed use communities with access to public transit, it appears that people are not able to find homes in such communities. A study of over 3000 people in Atlanta examined preferences for walkable communities and assessed the types of neighborhoods where people actually live. Many people were "mismatched" and did not live in their preferred neighborhood type. The biggest mismatch was for people who wanted to live in walkable neighborhoods but did not. This suggests walkable neighborhoods are undersupplied (Frank, et al, 2007). An important reason for the under-supply is that zoning laws in most cities favor or require the separate-use suburban-style development that reduces the ability of the real estate industry to respond to the market's demand for walkable communities.
Because policies, including funding priorities, determine what kinds of built environments are constructed, it is useful to assess opinions about relevant policies. About 1800 randomly sampled adults in the US rated their support of various policies that could increase physical activity. Between 89% and 92% of both women and men stated a preference for more local government funds to be devoted to walking/jogging trails, recreation centers, and bike paths. There was also strong support (86% to 88%) for zoning laws to require walking and biking paths (Brownson, et al, 2001).
Bicycling is the least used mode of transportation in the US, and it is well documented that the largest barrier to bicycle use is concern about safety from cars. A study of about 1800 adults assessed current bicycle use and projected changes if bicycling was made safer from traffic. If bicycling was safer, bicycle riding at least once per week might increase from 8% to 40% of adults. Forty-four percent of people who did not own a bicycle said they would start riding at least once per week. The projected increase in bicycling was greatest among racial and ethnic minority participants, populations who tend to be at highest risk of chronic diseases (Sallis, et al, 2013). This study suggests that many people would support changes in transportation policies that invest more in protecting bicyclists from traffic.
In summary, there is good evidence that most US adults want more walkable neighborhoods, like to live in mixed use neighborhoods, and want safer places to ride bicycles. There is strong support for more government funding for walking and bicycling facilities and recreation centers. The evidence we found says that US adults are in favor of the kinds of activity-friendly environments that help people be active and produce important co-benefits.

Disparities in Access to Physical-Activity Friendly Environments
One of the biggest problems with health in the US is that there are disturbingly large disparities in life expectancy and health outcomes when comparing high-and low-income people and when comparing those from majority and minority racial-ethnic groups. There are many reasons for these persistent disparities, but increasing focus is being placed on the role of environments. It is known that exposure to toxic chemicals tends to be greater in low-income areas, but evidence is also accumulating that low-income neighborhoods in the US provide lower quantity and quality of opportunities for physical activity.
Active Living Research compiled the research on this topic. There were three main conclusions of this review (Taylor & Lou, 2011).
• Residents of lower-income neighborhoods generally lacked well-maintained sidewalks, attractive scenery, and trees, all features that make for safer and more enjoyable physical activity. • Residents of lower-income neighborhoods were less likely to have parks in their neighborhood, and those without nearby recreation facilities tended to be less active. • Residents of lower-income neighborhoods usually had more crime in their area as well as more physical and social disorder such as graffiti, litter, drug use, and public drinking.
As an example of a specific study, about 2200 adults from two regions of the US reported on environmental characteristics of their neighborhoods. Their neighborhoods were selected to vary widely on both income and walkability. Participant perceptions of specific neighborhood characteristics were compared across lower-and higherincome groups. Residents of lower-income neighborhoods reported (Sallis, Slymen, et al, 2011): • Fewer walking and cycling features, such as well maintained sidewalks, pedestrian paths, and bike trails; • Fewer aesthetically pleasing features, such as landscaping, trees that provide shade, interesting sights, and attractive buildings; • Fewer pedestrian and traffic safety features, such as safe crosswalks and slow traffic speeds; • Less safety from crime, including comfort of walking at night, and neighborhood crime rates; • Fewer nearby recreation areas such as, basketball courts, soccer fields, tennis courts; • Fewer gyms and fitness facilities within walking distance.
Unfortunately there is growing evidence that low-income people in the US live in neighborhoods that have fewer public and private recreation facilities, less-safe environments for walking and bicycling, less-safe intersections, and less attractive scenery that might encourage them to be active. Lower-quality physical activity environments for low-income people and communities of color is a social injustice that likely contributes to higher rates of chronic diseases and higher health care costs. Targeting investments in physical activity environments in low-income neighborhoods could bring a particularly good return on investment.
Countries vary greatly on income and health inequalities, so findings about disparities in the US will not be applicable to many other countries.

Strengths and Limitations of the Review
The primary strength of the literature review was the breadth of topics explored. For each setting we identified several features that were related to physical activity, and for each feature we evaluated for 6 types of co-benefits. We searched the international scientific literature and the grey literature using numerous search terms to account for differences in terminology across fields. We solicited recommendations from experts in all the settings. We coded the level of evidence of each source and created summary scores of the evidence by weighting each finding by the rating of the source. To avoid basing our findings on lower quality evidence, such as poorly-substantiated claims in advocacy documents, we did not include lower levels of evidence in our summaries. However, all relevant documents found in the search are coded in the evidence tables. To be even more conservative, summary scores represented in the final matrix tables are based on positive findings minus null and negative findings.
The main limitations of the review were a consequence of the breadth. Because of the large number of topics searched for, it was not possible to conduct a systematic review or a quantitative meta-analysis. A requirement of systematic reviews is assessment of the quality of each study, but this was not feasible. We relied on existing reviews whenever possible. When reviews were available, we did not include the individual studies, but we did attempt to find more recent studies. A different coder conducted each search, and though they received some training and supervision, they were not experts in literature reviews or syntheses. There are undoubtedly differences across topics in thoroughness of search, classification of levels of evidence, and identification of findings to highlight.
Though we searched the international literature, we were only able to do so in English. This limited our ability to identify and include the full range of international studies. Nevertheless one-quarter of the findings were from specific non-US countries, and another quarter of findings were from reviews that included international literature. The majority of the international literature was from English-speaking countries. We were unable to compare results from US and non-US countries.
Other limitations include publication bias that favors positive findings and inclusion of grey literature that may be more biased.
We expect there will be some objections to our summary scores. Though we do not encourage the score to be interpreted literally as the actual strength of evidence, we believe the scores do provide a rough indication of the extent of evidence-pro, con, and neutral. There are certainly limitations since reviews that receive a score of 4.5 may summarize 10 studies or 100 studies. However, if the net scores for level of evidence high, there is reason to have high confidence in the finding for a connection between a feature and an outcome. Please consider the summary scores rough indicators, not absolute or true measures.

Recommendations
The most important conclusion of this review is that creating communities, transportation systems, schools, and buildings that make physical activity attractive and convenient also produces a wide range of other benefits for communities. Benefits were found for green/environment, economic, and multiple dimensions of health outcomes. It is often said that a good solution solves multiple problems, so building places that are good for physical activity should be considered a superlative solution.
Rates of chronic disease are high in countries at all income levels, and these rates are increasing fastest in low and middle income countries (UN, 2011). Health care costs are unsustainable and climbing in most countries. The mental health disorder of depression creates the highest burden of disease worldwide. Injuries are the biggest cause of death among youth people worldwide. Every country and city is looking for ways to improve economic growth. Pollution from automobiles is a major threat to health in most countries and is making many cities unlivable. The consequences of climate change are expected to be the worst human-made disasters in history. It seems inconceivable that making cities better for physical activity can contribute to solving all of these problems. However, more parks and trails, walkable community designs, more sidewalks and bike paths, better public transit, and schools and workplaces within walking and biking distance of students and workers could improve the functioning of cities in many ways.
We urge mayors, other city officials, and staff in multiple departments to consult these findings as an aid in decisionmaking. There are many reasons to make a city better for physical activity, so it is reasonable to consider the full range of effects.
One of the more surprising findings was that all five physical activity settings could be designed so they have positive effects on economic outcomes. Based on the specific studies we found, many groups could enjoy economic benefits of activity-friendly environments, including reduced government spending on infrastructure, home owners, real estate developers, health insurance companies, employers, retailers, commercial property owners, and taxpayers. This is an extremely broad coalition, and we suspect most of them are not aware of the economic benefits of activity-friendly environments.
There is no one environmental feature that will solve the problem of physical inactivity. Some features support active transportation and others support active recreation. An active living city supports both types of physical activity. An international study clearly showed that the more activity-friendly features in a neighborhood, the more active people are (Sallis, 2009). Thus, it is the pattern of multiple features that seems to be important. Several features need to work together. The combination of walkable urban designs and transportation environments that support walking and biking is needed to facilitate active transportation. If people can walk or bike safely to well-designed parks, they may get twice the physical activity. Children's physical activity is likely to be maximized when they can walk or bike safely to nearby schools and stay after school to use high-quality recreation facilities. Physical activity of workers is likely to optimized when the workplaces are in walkable communities served well by public transit, the building has attractive and accessible stairs, and there are facilities for showers and exercise.
There are multiple features of each setting that can be designed well to yield the multiple co-benefits (Table 15). Rather than thinking that designing one feature of a transportation system or school is sufficient, we encourage decisionmakers and designers to consider how all features of a setting can be optimized for physical activity and multiple other benefits. Explicit and conscious design of the best possible physical activity-promoting settings can be expected to pay off with multiple benefits, so additional costs are likely to be excellent investments. Changing policies about built environments is usually a politically difficult process, opposed by vested interests and vocal citizens. Evidence of the benefits of change may not be sufficient to overcome voices of negativity. Fortunately, we can expect many powerful voices to speak in favor of creating activity-friendly environments. Real estate developers can make more profits. Homeowners will likely see their properties appreciate. Retailers can expect more business. Most importantly, the majority of people, at least in the US, support more walkable communities, more walking and bicycling facilities, and more and better parks and recreation facilities. People want to live in walkable communities but cannot because zoning laws in the US favor single-use, automobile-dependent developments. Though there is opposition to every proposal, support for activity-friendly environments appears to be widespread and growing.
The studies showing that people with low incomes have less activity-friendly and less-safe neighborhoods should stimulate action. Low-income status is one of the biggest drivers of poor health, and neighborhood design and physical activity amenities may be contributing to the poor health. Thus, targeting improvements to low-income communities could help reduce health disparities and may produce additional co-benefits in economic development. However, targeted improvements to low-income neighborhoods should be planned so the current residents benefit and are not priced out as the neighborhood is enhanced.
The evidence in this report should reduce resistance to making multiple environmental changes to promote physical activity. Creating more activity-friendly features produces additional co-benefits, including economic ones, which will be experienced by a broad range of stakeholders. Thus, a reasonable conclusion based on the evidence is that the more we invest in helping residents be active, the better physical/mental/social health, environmental health, and economic health of the city. People want more activity-friendly neighborhoods, so policy changes to create them are likely to be popular. It would make sense for city leaders to compete with other cities in creating the best activity-friendly environments.