Background

Human health is influenced by a variety of determinants including factors related to the environment where people live [1]. Urbanization has become one of the global megatrends that characterize the current development of mankind. At the beginning of the 20th century, only about 10 percent of the world’s population were living in urban areas. In 2015, this percentage increased to about 54%, and is predicted to reach 60% in 2030 and 66% by 2050 [2].

Urbanization more often offers health advantages in comparison with rural areas, as the basic infrastructure relevant for health such as water, sanitation and housing are generally more developed. In addition, health services and facilities appear to be more available in cities than in rural areas. However, cities may also cause relative disadvantages for health, e.g. crowded living and stressing working conditions, higher rates of crime and violence, sedentary life styles, reduced physical activity, and, additionally, the urban food environment may contribute to the rise of non-communicable diseases [2].

However, many decisions that impact human health are made outside the health sector [3, 4]. For example, environmental changes resulting from the intensification of agriculture, industrialization and increasing energy use are considered as important sources of health problems [5]. Decisions about the quality of social services, housing, employment opportunities or public transport are among many others key influences on health [6], and are again usually made outside the health sector. This has led the WHO to extend the ideas of healthy public policy, already formulated in the Ottawa Charta of Health Promotion to the principle of “Health in All Policies” [7]. Yet, considering not only reduction of health risks, but also enhancing health promoting potentials in urban development and urban planning seems not to be implemented systematically on a large scale [6].

Health Impact Assessment (HIA) is an approach to bring health considerations into urban development and urban planning. HIA has been defined as “… a combination of procedures, methods and tools by which a policy, program or project may be judged as to its potential effects on the health of a population, and the distribution of those effects within the population” [4]. One aim is to produce recommendations for decision makers and stakeholders for “… maximizing the proposal’s positive health effects and minimizing its negative health effects” [8]. HIAs have been conducted in all regions of the world, and the majority of HIA practitioners expect an increased use in Australia, East Asia and Pacific, Europe and North America [9].

The impact of urban environment on physical activity has received some consideration during the last decades. Globally, physical inactivity has been accounted for the fourth leading cause of mortality, after high blood pressure, tobacco use and high blood glucose, contributing to 6 percent of worldwide deaths [10]. The Global Burden of Disease study is using a sophisticated hierarchical model of risk factors including physical inactivity as a level 2 risk factor [11]. Recently, results from this study indicated that, globally seen, close to 1 million deaths in 2019 were attributable to physical inactivity [11]. In addition, physical inactivity is considered as a major risk factor for non-communicable disease, particularly cardio-vascular diseases, diabetes mellitus type 2, and several types of cancer [12]. Moreover, physical activity contributes to the maintenance of healthy weight and to the prevention of overweight and obesity [12] which in turn is a major risk factor for the mentioned NCDs [13]. On the other hand, the beneficial effects of physical activity on all-cause mortality [14,15,16], the incidence of cardiovascular health, diabetes, several types of cancer [17], and mental health [18, 19] have been well documented.

In relation to physical activity two concepts for the urban environment have received considerable attention and generated some research: active transport and walkability. Active transport comprises walking or cycling for the purpose of reaching a destination such as school, workplace, or a shop [20]. Walkability summarizes attributes of the urban built environment that encourage and/or enable more walking [21,22,23]. The original concept of walkability was developed in the 1990s in US transportation research and has focused on walking for transportation [24]. As this concept was adopted by physical activity and public health researchers and practitioners, it was extended to include walking for transportation and recreational purposes as well as other types of physical activity, e.g. biking [24]. Hence, walkability has extended beyond walking to generally promoting physical activity in communities, urban neighbourhoods and larger urban areas [25]. A review [23] concludes that there is sufficient evidence that the proximity to potential destinations, aesthetic qualities – the attractiveness of the environment -, mixed land use, residential density within neighbourhoods, sidewalks and connectivity are attributes of the built environment that correlate with increased walking. Recently, a more comprehensive framework of walkability has been suggested [26, 27] that incorporates nine dimensions of the built environment, namely connectivity, diversity of land-use, residential density, traffic safety, surveillance (how well traveling in the street can be seen from surrounding houses and businesses), parking (less parking encourages more walking), experience (including e.g. aesthetics), greenspace and community (social interaction and participation).

The health impacts of active transport have been intensively studied and a systematic review provides strong evidence that active transport provides substantial net health benefits even if negative health impacts like accidents and exposure to air pollution are taken into account [14, 20].

To the best of our knowledge up to now no systematic review is available to summarize the evidence on the health impacts of walkability conceptualized as detailed above as characteristics of the urban built environment. We hypothesize that the walkability of urban environments may affect health outcomes via several pathways (see Fig. 1): walkability may result in more physical activity either by improving active transport or by encouraging recreational activities including deliberate exercising. In addition, green space improves health by encouraging more physical activity and by other effects, e.g. lower distress and better mental health [28] and the better walkability of the built environment could promote and result in improved social relationships [29] that in turn impact human health.

Fig. 1
figure 1

Potential major pathways of the health impact of walkability, own presentation

Full size image

The objective of this paper is to conduct and report on a systematic review of health impact assessments (HIAs) of projects, policies or programmes that aim to improve the walkability of urban built environments.

Particularly, we aim at answering the following research questions related to such HIAs:

  • Which types of projects, policies or programmes related to walkability in urban development have been investigated in HIAs?

  • Which methods were used to assess the health impacts? Which data sources were used, and which analytical models were applied to assess health impacts?

  • Which health impacts (e.g. changes in mortality, incidence or prevalence of diseases, quality of life) have been identified related to improvements of walkability?

  • How and by whom are HIAs of walkability implemented in practice?

Methods

This systematic review was designed based on the PRISMA 2020 guidelines [30]. We considered the following definition of walkability as the dividing line for identification of eligible articles in the current systematic review: Walkability summarizes attributes of the urban built environment that encourage and/or enable more walking or other types of physical activity in communities, urban neighbourhoods and larger urban areas.

In order to include a wide range of studies, no specific preference for a definition of Health Impact Assessment was considered.

Data sources and search strategy

We searched PubMed and ScienceDirect databases with the purpose of incorporating international studies as well as the websites of two German associations “Stadt-und-Gesundheit” (City and Health; http://stadt-und-gesundheit.de) and “Akademie für Raumentwicklung in der Leibniz-Gemeinschaft” (ARL – Academy for Territorial Development in the Leibniz Association; https://www.arl-net.de) to identify research reports focusing on spatial planning particularly in the German context. The databases were searched thoroughly in November 2020, and an additional search of PubMed and Science Direct was also conducted in the end of 2021 to update the study pool. We operated the advanced search in PubMed and ScienceDirect using the search terms in Table 1. The two German databases did not offer an advanced search tool. Therefore, an adoption of the strategy for these two sources was necessary. This involved a title screening of all listed papers and reports. We included all papers that mentioned “Health Impact Assessment” (the German equivalent term “Gesundheitsfolgenabschätzung” respectively) or “Walkability” in the title. As the term Walkability was rarely used in the titles of the reports on the two German databases, we accepted the German terms for mobility/mobile, physical activity, walking (distance), transport (ation), walkable, and pedestrian as potential equivalents. For the same reason, we accepted papers with a title suggesting a health outcome related to walkability (e.g. increased walking). The details of identified articles and search terms for each of databases are provided in the Table 1. In addition, reference lists of papers included during the selection process were screened for additional papers that might be relevant.

Table 1 Search strategy and identified articles in each database
Full size table

Inclusion and exclusion criteria

Regarding selection criteria, any peer-reviewed publications evaluating a real or projected (modelled) health impact or health outcomes of a policy, programme or project that intended to change an aspect of walkability in an urban environment were eligible for the review, with the exception of short communications, published abstracts and conference contributions. Furthermore, the eligible articles were published in English or German and after the year 2010. We decided not to include grey literature because we believe that the peer-review process provides an important quality assurance and therefor enhances the credibility of the research findings. In addition, identification of relevant articles via databases of the peer-reviewed scientific literature seems more transparent and replicable than a comprehensive open search using a wide variety of search engines. Details of inclusion and exclusion criteria are available in Table 2.

Table 2 Inclusion and exclusion criteria for article selection
Full size table

Screening, data extraction and analysis

Title screening was done by one reviewer. In case of any uncertainty, the decision was made after discussion with another author. For the remaining records, abstract screening was done independently by two reviewers. Any disagreement between the reviewers was resolved by discussion, in some rare cases by including additional authors of the present paper.

The final selection step was the review of the full text of remaining articles by two authors. Every step of the review process was discussed during regular meetings in order to clarify uncertainties and challenges related to selection process. In case of disagreement of authors on the eligibility of articles, a third reviewer conducted an additional review, and the final selection was approved by discussion.

During data extraction stage, two authors independently reviewed the full text and tabulated the extracted data. The third reviewer extracted the data of articles that were the subject of disagreement of the first two reviewers. Extracted data included author, year of publication, HIA definition and method that was used, operationalization of walkability (if any), which NCDs were considered in HIA, aim, setting, study population of the project, policy or program, dependent and independent variables, measuring instruments, statistical/analytical methods applied, and results. In addition, conductive conditions and resources, barriers and challenges, and recommendations were extracted, if mentioned. Regarding the HIA itself, we extracted, if mentioned, who initiated the HIA, who conducted the HIA, other actors involved in the HIA, and how HIA was integrated into existing planning instruments or processes. The extracted data were captured in an Excel-Sheet that is available as supplemental material.

We conducted a qualitative analysis and summary of the extracted data to answer our research questions. The development of categories and classification of study reports was done following the principles of inductive thematic analysis [31, 32], and was done in consensus of three authors (JW, JB, EN).

Results

Identification and selection of relevant studies

As shown in Fig. 2, a total of 946 records was identified. Database search resulted in 817 records. Additional 129 records were identified through reference tracking. 21 duplicate records were removed before title and abstract screening. Title screening excluded 682 records; abstract screening excluded additional 133 records. Since 8 reports could not be retrieved, the full text of 102 reports were assessed according to inclusion and exclusion criteria. 62 reports were excluded, most often because they did not address the effect of a policy, programme, or project on walkability (see Fig. 2). Finally, 40 study reports were included in the present review.

Fig. 2
figure 2

Prisma flow chart of study identification and selection

Full size image

An overview of the included HIA reports is presented in Table 3. The complete data extraction sheet is available as supplementary material (Additional File 2).

Table 3 Source, place, aim of project, main results of included HIAs
Full size table

Location, type of projects, and type of HIAs

Most of the identified published HIAs addressed projects or policies in the USA (n = 18), followed by Australia (n = 6), UK (n = 5), and Canada (n = 3). Moreover, there were two reports from India, one from New Zealand, one that covered several European countries, one that covered several European and Non-European Countries, and another three reports each covering one European country. Particularly, Germany was only addressed in one HIA as one of several European countries (see Additional file 1: supplementary table S1 for details).

Most of the HIAs investigated the impact of improving or extending the infrastructure to facilitate active transport or public transport (n = 13). This refers to more concrete projects like extending cycling networks or better sidewalks. Among them, five HIAs were about improving bicycle and pedestrian infrastructure, and another five addressed bicycle infrastructure alone. Respectively one HIA studied improving public transport infrastructure alone, non-motorized transport plus public transport infrastructure, or extending sidewalks. Additional 11 HIAs assessed the impact of more general scenarios or policies to support active transport (see Additional file 1: suppl. table S2 for details).

Six HIAs examined the development of new suburbs; the redevelopment, revitalisation, or regeneration of a city or abandoned areas: Another five HIAs examined the redesign of urban neighbourhoods.

33 HIAs were clearly quantitative HIAs, that aimed at quantifying at least one primary health outcome, four HIAs were clearly qualitative HIAs and three reports included quantitative and qualitative reports (see Additional file 1: suppl. table S3 for details).

Data sources and analysis

A variety of data sources was used as basis for HIAs, this included primary data collection, secondary use of existing data, measurement of built environment variables via geographic information systems (GIS), and analysis of existing reports, inventories, or other similar data (see Addirional file 1: suppl. table S5 for details). Primary data collection through surveys or questionnaires was used in eight HIAs, interviews and/or focus groups and group discussions in seven HIAs, structured observations and audits in three HIAs, and accelerometer measurement in one HIA.

The secondary use of existing surveys was applied in 14 HIAs. The secondary use of data from group discussions, interviews and travel diaries was each mentioned in one HIA respectively.

Literature reviews as basic data source were used in five HIAs, the measurement of built environment variables by GIS was applied in seven HIAs, and the analysis of existing reports, inventories or other types of data was applied in seven HIAs.

Health impacts

Cardiovascular diseases were the health endpoint that was investigated most often, in 16 HIAs. This was followed by diabetes in 12 HIAs, Cancer (8 HIAs), mental illness (6 HIAs), premature death (5 HIAs), all-cause mortality (5 HIAs), respiratory diseases (5 HIAs), traffic accidents (4 HIAs) and obesity (4 HIAs) (see Additional file 1: suppl. table S4).

Most of the HIAs (n = 31) reported the improvement of health or health behaviour resulting from the investigated project or policy. However, three HIAs reported a lack of improvement or even a decrease of health status. In parallel, 13 HIAs reported a gain in economic value, whereas one reported a lack or loss of economic effects. Moreover, three HIAs reported on social effects and six HIAs gave additional recommendations for policies or the implementation of projects or HIAs (see Additional file 1: suppl. table S6 for details).

A closer look at those HIAs that reported negative health impacts or failed to find a positive impact yields the following results. One HIA that reported negative health impacts was on a comprehensive transit-oriented district redevelopment plan, that would result in increased exposure to air pollution and increased rate of asthma in children on one hand, but increased physical activity from walking and reductions in type 2 diabetes and high blood pressure on the other hand [43]. A HIA of different scenarios of urban development on air quality concluded that compact development increases particulate matter PM2.5 concentrations and PM2.5 attributable mortality [55]. Finally, a qualitative HIA using focus groups of a neighbourhood transformation project failed to find noticeable increases of physical activity [41].

Among the social effects reported, there was the potential for more alcohol outlets in mixed-use developments of a rezoning project in Baltimore, and associated increasing violent crime [65]. In contrast, greening of a vacant urban space was associated with reductions in gun assaults, vandalism and stress, and more safety [37]. Finally, in a newly developed neighbourhood that was designed to improve walkability, residents reported safety, aesthetics, and sense of community as factors that facilitated walking [49].

Implementation of HIAs

Regarding the implementation of the HIAs, none of the reports mentioned explicitly who was doing the work, but it reasonable to assume that this was done by the listed authors.

Most of the studies (32 of 40) reported external funding of the HIA and few studies explicitly mentioned that the HIA was initiated by a particular institution.

Only two study reports mentioned other institutions or stakeholders that were actively involved in the HIA: the city where the project was located, the respective Metro company and the NIH in one study, a number of local organizations, the police department, the church priest and local farmers and ranchers in the other study.

None of the study reports mentioned whether or how the HIA was integrated or associated with other planning instruments or procedures.

Discussion

The present review aims at summarizing the peer-reviewed literature on health impact assessments of walkability in urban development. The vast majority of studies reported beneficial health impacts particularly reductions of non-communicable diseases like cardiovascular diseases, diabetes, cancer, and mortality. As most HIAs examined the impact of projects, plans or scenarios that aimed at increasing active transport and/or public transport these results are in line with prior findings, because the beneficial consequences of active transport, walking and cycling have been well established [14, 20]. Negative health impacts were only reported in two HIAs and were related to increased exposure to air pollution that may result from more walking and cycling [43, 55]. However, it has been shown that beneficial impacts of walking and cycling clearly outweigh the potential negative impacts from air pollution and traffic accidents.

Walkability in a wider sense was studied in eleven HIAs, by either investigating the redevelopment of cities or abandoned areas, the development of new suburbs or the redesign of urban neighbourhoods. With one exception these HIAs reported beneficial health impacts. Social effects were rarely addressed in the included HIAs and differed fundamentally among the examined projects or plans. Mixed-use neighbourhoods could result in more alcohol outlets and more violent crime [65], whereas greening of vacant urban space could result in reduced violence and vandalism and increased perceived safety of the residents [37]. A newly developed neighbourhood designed to improve walkability was associated with improved perceived safety, aesthetics and sense of community [49].

In summary, the present review clearly shows that improving the walkability of neighbourhoods or cities yields positive health impacts and has only limited negative effects, depending on the project or policy. Moreover, improving infrastructure and opportunities for active transport and public transport play a major role for the beneficial effects. However, the research reports included in this review did not address aspects of inequality and equity, i.e. whether beneficial or harmful effects of projects, policies or programmes are equally distributed among different population groups. There is a need for future research to address the social inequalities of health impacts as well. Otherwise, it could happen that walkability is primarily improved in more affluent neighbourhoods, whereas more deprived neighbourhoods are even more neglected.

This review has some limitations that have to be considered when drawing conclusions. Most importantly, due to our search and selection strategy we may have missed several relevant HIA reports. First, we limited our search on peer-reviewed papers that were identified through two databases and two websites of German organizations and follow-up reference tracking. Thus, HIA reports in the grey literature were not included. It is likely that several HIA reports exist either unpublished or available only on local or regional websites. Indeed, this has been confirmed by a UK expert in health impact assessment (personal communication Dr. Fischer, Liverpool).

Secondly, we explicitly included walkability in our search terms which may cause restriction of the identified results. For example, the article by Mueller et al. [58] examines the health impacts of the superblock design in Barcelona, Spain. The superblock design is clearly improving the walkability of neighbourhood blocks as it “… aims to reclaim space for people, reduce motorized transport, promote sustainable mobility and active lifestyles, provide urban greening …” [58]. However, the whole article does not even mention the term “walkability”. Nevertheless, we hope that we have covered the most important HIAs related to walkability through our reference tracking which identified the mentioned article on the Barcelona superblocks.

Having the limitations in mind, the following conclusions seem justified. Health impact assessments related to the walkability of urban environments are more established in English speaking countries, including US, UK, Canada, Australia, and New Zealand. Other European countries, particularly Germany, clearly lack behind. Despite the 40 HIA reports that were included in our review, there is a need for more HIAs published in peer-reviewed journals, given the important role urban environments play in determining human health. Peer-reviewed publication would provide some basic quality assurance and hence increase the credibility of the findings, and in addition would help to identify relevant reports more easily through standard scientific literature data bases. This in turn could motivate more research on the health impacts of walkability and underline the importance of considering and improving walkability in urban planning processes. If not considerably more unpublished HIAs exist, we have to conclude that health impacts of urban planning and urban development need a considerable push in local administrations and planning agencies.

A major strength of this review is going beyond active and public transport; it also includes more general characteristics of the urban built environment which are subsumed under the concept of walkability.

Conclusions

While the beneficial effects of active transport and public transport which are important components of walkability are well documented and established, the contribution of other features or components of walkability to health are less well understood. Particularly, there is a need to establish more quantitative associations of the different dimensions of walkability as suggested in the walkability framework [27] to increased physical activity, social interaction and perceived safety and stress. Such quantitative associations would allow to predict the health impact of urban design features more accurately.

Future research would benefit from explicitly mentioning the key terms like “health impact” or “walkability” in the title of their publications, and from providing more clear definitions of such constructs in their report. With regard to public health practice, reports should provide more detail about who initiated HIAs, who conducted them, who participated and how they were implemented in the planning processes. Finally, it is desirable, that access to existing HIAs is facilitated by publishing them in peer-review, preferably open access scientific journals. This could promote the consideration of health impacts in urban planning, particularly in countries and regions were this is not well-established practice.