Skip to main content

Investigating the physical activity, health, wellbeing, social and environmental effects of a new urban greenway: a natural experiment (the PARC study)

Abstract

Background

Evidence for the health benefits of urban green space tends to stem from small, short-term quasi-experimental or cross-sectional observational research, whilst evidence from intervention studies is sparse. The development of an urban greenway (9 km running along 3 rivers) in Northern Ireland provided the opportunity to conduct a natural experiment. This study investigated the public health impact of the urban greenway on a range of physical activity, health, wellbeing, social, and perceptions of the environment outcomes.

Methods

A repeated cross-sectional household survey of adult residents (aged ≥16 years) who lived ≤1-mile radius of the greenway (intervention sample) and > 1-mile radius of the greenway (control sample) was conducted pre (2010/2011) and 6-months post implementation (2016/2017). We assessed changes in outcomes pre- and post-intervention follow-up including physical activity behaviour (primary outcome measure: Global Physical Activity Questionnaire), quality of life, mental wellbeing, social capital and perceptions of the built environment. Linear regression was used to calculate the mean difference between post-intervention and baseline measures adjusting for age, season, education, car ownership and deprivation. Multi-level models were fitted using a random intercept at the super output area (smallest geographical unit) to account for clustering within areas. The analyses were stratified by distance from the greenway and deprivation. We assessed change in the social patterning of outcomes over time using an ordered logit to make model-based outcome predictions across strata.

Results

The mean ages of intervention samples were 50.3 (SD 18.9) years at baseline (n = 1037) and 51.7 (SD 19.1) years at follow-up (n = 968). Post-intervention, 65% (adjusted OR 0.60, 95% CI 0.35 to 1.00) of residents who lived closest to the greenway (i.e., ≤400 m) and 60% (adjusted OR, 0.64 95% CI 0.41 to 0.99) who lived furthest from the greenway (i.e.,≥1200 m) met the physical activity guidelines - 68% of the intervention sample met the physical activity guidelines before the intervention. Residents in the most deprived quintiles had a similar reduction in physical activity behaviour as residents in less deprived quintiles. Quality of life at follow-up compared to baseline declined and this decline was significantly less than in the control area (adjusted differences in mean EQ5D: -11.0 (95% CI − 14.5 to − 7.4); − 30.5 (95% CI − 37.9 to − 23.2). Significant change in mental wellbeing was not observed despite improvements in some indicators of social capital. Positive perceptions of the local environment in relation to its attractiveness, traffic and safety increased.

Conclusions

Our findings illustrate the major challenge of evaluating complex urban interventions and the difficulty of capturing and measuring the network of potential variables that influence or hinder meaningful outcomes. The results indicate at this stage no intervention effect for improvements in population-level physical activity behaviour or mental wellbeing. However, they show some modest improvements for secondary outcomes including positive perceptions of the environment and social capital constructs. The public health impact of urban greenways may take a longer period of time to be realised and there is a need to improve evaluation methodology that captures the complex systems nature of urban regeneration.

Introduction

By 2050, 84% of Europe’s population is expected to live in cities (75% of the European population already do so) [1]. Urban ecosystems need to be improved to ensure that, above all, they are sustainable and can support a higher quality of life for growing populations. We know that the natural and built environments including urban green spaces (UGS), and their relationships to the social fabric, are critical determinants of physical and mental health, and the economic and environmental wellbeing of our cities [2,3,4,5].

However, in an increasingly urbanised world, UGS are often under threat and face extensive competition, especially from housing, business and transport demands. Improving urban health and reducing health and social inequalities can be achieved by policies and practices that effect changes to create and enhance green (e.g. parks, forests, greenways) and blue (e.g. rivers, canals, beaches) space and by creative urban design that ultimately supports populations and individuals, encouraging healthier behaviour. Various political frameworks underscore the need for UGS in our cities. For example, the New Urban Agenda [6] states that ‘green space can reduce urban poverty, including tackling urban regeneration and creating safe and social spaces for integration and interaction, and access to quality services’. Similarly, the 2030 Agenda for Sustainable Development [1] pledges to ‘provide universal access to safe, inclusive and accessible, green and public spaces, in particular, for women and children, older persons and persons with disabilities’ (SDG 11.7). Therefore, UGS can contribute to the achievement of a number of the UN’s Sustainable Development Goals.

A large body of evidence demonstrates the benefits of UGS (e.g. pocket parks, green roofs/walls, greening of vacant lots, urban trails/greenways) for better health and for mitigating inequalities [2, 3, 5, 7]. Evidence has demonstrated significant positive contributions to physical, psychological, social, economic and environmental wellbeing [8,9,10,11,12,13,14,15,16,17,18,19,20]. However, these benefits may not be equitably distributed across populations and some UGS have been associated with widening health and social inequalities [17]. There are plausible aetiological pathways between contact with nature and reduced risk of non-communicable diseases (NCDs) [7]. Natural environments can also provide co-benefits such as combatting air pollution, enhancing resilience to adverse weather [10], and promoting social inclusion [15]. However, some of the evidence is contested.

Much of the evidence for health benefits stems from small, short-term (quasi)experimental or cross-sectional observational studies, and to a lesser extent longitudinal observational studies, but the evidence from intervention studies is sparse [21, 22]. A previous systematic review involving 11 studies [23] suggested that there was promising evidence for UGS interventions that combined a change to the physical green space with a promotion/marketing programme (i.e. a dual approach) for increasing park usage and physical activity levels. However, that review solely focused on physical activity behaviour. A more recent study [21] extended this work to review the current evidence base of UGS interventions (involving 38 studies) for other health, social and environmental benefits to understand better the multi-functional nature and value of UGS. Hunter et al. [21] identified evidence supporting the use of certain UGS interventions for health, wellbeing (e.g., reduction in stress), social (e.g., reduction in crime, improved perceptions of safety) and environmental (e.g., increased biodiversity) benefits. Most research and policy assume that proximity and access to UGS are surrogates for use, and more UGS nearby is assumed to be good for all. However, substantial sections of the population (many with high risk of NCDs), do not visit or pay attention to UGS. Our ability to properly evaluate UGS, given the co-benefits, is limited. Researchers have yet to fully evaluate and economically value the complex, multi-functional nature of UGS and so the full potential of these spaces as public health assets has yet to be realised [21, 24]. Further, UGS tend to be viewed as discrete physical ‘assets’ without adequate appreciation of how health and other co-benefits (such as social, environmental and economic) rely on their wider integration with the surrounding urban fabric and the social environment.

The aim of this study is to investigate the public health impact of the development of an urban greenway (an UGS intervention)—the Connswater Community Greenway (CCG)—on a range of physical activity, health, wellbeing, social, and perceptions of the environment outcomes, to provide an understanding of the public health influence of a systems-level intervention. More specifically, outcomes include physical activity, general health, mental wellbeing, social capital and perceptions of environment, stratified by exposure to the CCG and deprivation.

Methods and materials

The Physical Activity and Rejuvenation of Connswater (PARC) Study is a before-and-after evaluation of the public health effects of the CCG on physical activity, health, mental wellbeing, social and perceptions of the environment outcomes in Belfast [25]. The study was developed in partnership with statutory, voluntary and community organisations, and comprised four main elements: 1) a quasi-experimental before-and-after survey of the local CCG population (repeated cross-sectional design); 2) assessment of change in the local built environment and walkability using data from geographic information systems (GIS); 3) semi-structured interviews with local residents, and a range of community stakeholders before and after the regeneration project; 4) an economic evaluation. The current study focuses on the results from the before and after household survey. Economic evaluations have been published elsewhere [24, 26]. The study was funded by the National Prevention Research Initiative (Medical Research Council) and approved by the Office for Research Ethics Committees, Northern Ireland (09/NIR02/66).

The intervention

Context

The study population comprised those residents living within 29 electoral wards (i.e. the smallest unit of administrative geography in Northern Ireland with an average population of 4000) in the political constituency of the CCG with a total population of approximately 110,600 (see Fig. 1). Twenty-two of these wards (approximately 87,500 residents) have a geographical centroid at or within (≤) a 1-mile radius of the CCG (i.e. the intervention area/sample), and seven of the wards are within the top 25% most deprived wards in Northern Ireland, as determined by the Northern Ireland Multiple Deprivation Measure [27].

Fig. 1
figure1

The Connswater Communitty Greenway and PARC Study Sampling Area

Intervention description

The CCG (www.communitygreenway.co.uk) is a major urban regeneration project in Belfast, Northern Ireland, funded primarily by a Big Lottery Living Landmarks Award, and provision of other funding by local government departments and the local city council (totalling £40 m). The funding for the CCG was acquired by a local community organisation, and is jointly managed and maintained with the local authority. Specific aspects of the regeneration include: provision of a 9 km urban greenway along the course of 3 rivers; 5 km of remediated water courses; 16 km of new or improved foot and cycle paths; development of a new civic square; development of 8 tourism and heritage trails; 23 new or improved bridges or crossings; 22 new signage points; installation of public art; 13 ha of upgraded parks; 2 multi-use games areas; 2 new toilets. Many of the green spaces that became part of the CCG did exist prior to the intervention, but were mostly unconnected to each other with poor accessibility to surrounding neighbourhoods. The new paths, bridges and new access points to parks opened up these improved spaces to more of the local population. For example, one new park gate at Sir Thomas and Lady Dixon Park increased the number of households in its 5 min catchment by 59% (1702 households to 2701 households) and the 15 min catchment by 33% (12,992 households to 17,314 households). This was accompanied by extensive landscaping and enhanced biodiversity creating a wildlife corridor. Significant community engagement and involvement occurred throughout the course of the development of the CCG including in the design, provision of volunteering opportunities, and the naming of local bridges. Twenty-six schools and colleges are in close proximity to the CCG and two education officers were employed to engage the schools and colleges with the CCG. Another unique aspect was that the CCG had lighting columns along the whole route, and was lit 24 h-a-day, making it the first area of UGS to be available for use 24 h-a-day in Northern Ireland. The regeneration also involved a £11 m flood alleviation scheme, moving the course of a river away from a residential area prone to flooding. Social engagement and CCG promotion activities and events occurred in parallel with physical changes to the intervention site. A ‘bottom-up’ approach was applied which involved the employment of a full-time community support officer. This project recognised that UGS interventions are long-term investments as reflected by the 40-year management and maintenance plan for the CCG that was developed from the outset.

The original intention was for a 3-year design and build project. However, due to legal and contractual issues, several substantial delays had major implications on the research study. The original PARC Study was intended to be a 5-year study, but had to be extended by a further 3 years to accommodate these contractual delays that led to a delay in the development of the CCG.

The study design was underpinned by the socio-ecological model [28]. The survey content was informed by, and reflected the various levels of, the socio-ecological model comprising measures of individual, community and perceptions of the environment factors (Fig. 2). The aim of the CCG was to deliver positive health, mental wellbeing, social and perceptions of the environment outcomes for the local population. Primarily, the CCG was hypothesised to offer new opportunities for physical activity through the development of the linear corridor, new and improved cycle and footpaths, and promotional events and social activities to encourage usage and physical activity. In addition to the range of changes to the physical environment, other interventions to promote physical activity included neighbourhood walking groups and initiatives targeted to promote the use of the CCG in distinct population segments (e.g. young mothers, unemployed and senior citizens), schools-based initiatives and community-based social marketing initiatives. This dual approach (i.e., changes to the physical environment coupled with promotional events and programmes to encourage use) was highlighted by Hunter et al. [21, 23] as being an important component of effective UGS interventions. The environmental aesthetics of the area were improved through the provision of landscaping involving the planting of trees and shrubs, public art and remediation of water courses to improve the biodiversity of the area. Improving the perception of safety of the area was directly impacted through 24 h-a-day lighting, closed-circuit television (CCTV) and the presence of volunteer park wardens who monitor the use of the Greenway and report any problems as well as serving as community champions for the CCG.

Fig. 2
figure2

CONSORT diagram

Study design – natural experiment

Repeated cross-sectional household survey

A survey of a random sample of households was conducted in 2010–2011 (i.e. before development of the CCG and repeated in 2016–2017 (i.e. 6 months after the opening of the CCG). Surveys were conducted over a 12-month period to account for seasonality. Households in 29 electoral wards were identified as the target sampling area due to their proximity to the CCG. The sampling strategy ensured proportionality with the Northern Ireland population based on estimates of the number of residents aged 16 years or older provided by the 2001 and 2011 Census. A random probability sampling framework was constructed by a random selection of addresses from each of the selected electoral wards using the Royal Mail’s Postal Address File (PAF), stratified by the proportion of the overall population within each target electoral ward. An information sheet about the study was posted to each household, followed up by a visit approximately 1–2 weeks later from a trained interviewer. The interviewer called back a maximum of five times in order to achieve a completed interview. Data collection was undertaken by an independent survey company (Perceptive Insights). If there was no response to these call backs, the address was recorded as a ‘non-response’ and another address within the same area was selected using the same process. For each household, the ‘last birthday rule’ (i.e. person in the household aged 16 years and over who had the most recent birthday) was used to randomly select an individual within each selected household to complete the survey. After the selected individual provided written consent to participate, an interviewer-administered questionnaire was completed.

Sample sizes

The primary outcome for the survey was the proportion of the affected population achieving the recommended levels of (total) physical activity. The sample size required to detect differences in population proportions was 934 at both time points, assuming an effect size of 0.15 at 90% power (α = 0.05), estimated using the initial population proportions of those achieving the recommended levels of physical activity using alternative assumptions of 20, 30 and 50%. We surveyed 1037 and 968 individuals, who resided in the electoral wards whose geographic centroid was ≤1-mile of the CCG area (i.e. the intervention area/sample), before and after the intervention respectively. The ≤1 mile radius represents an approximate 15-min walk to access the CCG. This distance is commonly used as a rule of thumb in walkability literature as being accessible [29,30,31].

In addition, we surveyed a further 168 and 246 individuals who resided in seven other wards (see Fig. 1) in the wider area (representing a similar proportion of the population to the other areas), which facilitated the exploration of exposure to the CCG via distance decay, before and after the intervention respectively. The other seven wards have a geographical centroid greater than a 1-mile radius from the CCG (i.e., the comparator area), and immediately surround the ‘intervention’ area (i.e., the wards within a 1-mile radius of the CCG).

Outcome measures

Physical activity was measured using the Global Physical Activity Questionnaire (GPAQ) which assessed total physical activity time via the domains of work, active travel and recreational physical activities of moderate and vigorous-intensity in the last 7 days [32]. This measure has been validated for the Northern Ireland adult population and reliably captures change in physical activity [33]. The primary outcome was the change in proportion of individuals identified as regularly physically active, according to the UK recommendations of a minimum of 150 min of at least moderate-intensity physical activity per week (or 75 min of vigorous intensity physical activity per week, or a combination of the two) [34, 35].

Secondary outcome measures included: i) mental wellbeing using the Warwick-Edinburgh Mental Wellbeing Scale [36, 37]; ii) quality of life using the EQ-5D-3L instrument [38, 39]; iii) perceptions of the characteristics of the environment associated with active travel and physical activity, including attractiveness, traffic, access to amenities and safety (measured on a 1–5 scale; 1 = strongly disagree; 5 = strongly agree) [40]; iv) items relating to neighbourhood social capital, as reflected in civic engagement, neighbourliness, social networks and support, and perceptions of the local area using the instrument employed in the UK General Household Survey [41, 42]. See Appendices A and B for ranges of the scales and further details.

Comparator data

There are two sources of comparator data: i) electoral wards whose geographical centroid was greater than 1 mile from the CCG (see section 2.2.2) (control area); ii) exposure to the intervention using distance decay analysis investigating outcomes related to distance from the CCG at 0-400 m, > 400 m–800 m, > 800 m-1200 m and > 1200 m (for the intervention area sample only where data was available).

Intervention exposure was defined as proximity to the CCG infrastructure, with less-exposed people living farther from the CCG acting as a comparison group for the more-exposed people living closer to the CCG. Proximity was operationalised as the distance from the participant’s home address to the nearest accessible point (e.g., accessible footpath, park entrance) to the CCG. Distance was calculated in ArcGIS 9 (Esri, Redlands, CA) using the purposefully digitised footpath network (see [43] for further details). Briefly, this included mapping the complete footpath network for the ‘intervention area’ surrounding the CCG. As no comprehensive network data had been collected on footpaths in Northern Ireland, this had to be mapped afresh, using the existing map base provided under a research agreement by the Land and Property Services of Northern Ireland cross-checked against aerial photographs, other online sources and field visits. Distances of 400 m, 800 m and 1200 m were chosen as they represent an approximate 5, 10 and 15-min walk respectively, from the participants home to the CCG. These distances have been used in previous physical activity and urban environment literature [44]. In particular, the limit of the ‘neighbourhood of opportunity’ has been described as being situated within 1000 m of home with 400 m being particularly influential [45,46,47].

We originally intended to compare our survey findings to a Northern Ireland population survey of physical activity (The Northern Ireland Sport and Physical Activity survey (SAPAS)) commissioned by Sport Northern Ireland [25], which also used the GPAQ [48]. This survey was collected in 2010 and was due to be repeated in 2015. However, owing to austerity in the public sector this survey was not repeated as intended. Therefore, we were unable to compare our household survey findings directly with the general Northern Ireland population as a whole. However, we have accessed other physical activity data (from a different government-funded survey), albeit using a different physical activity measurement tool (International Physical Activity Questionnaire; IPAQ [49]) and different sampling strategy, to illustrate the trend in population physical activity behaviour in adults in Northern Ireland between 2010 and 2017.

Statistical analyses

The main aim of the repeated cross-sectional analysis was to assess the effect of the CCG on a range of outcomes (physical activity behaviour, mental wellbeing, social environment (including social capital), perceptions of environment, exposure to the CCG). Pre–post changes in outcomes between baseline (wave 1; 2010) and follow-up (wave 2; 2017) were investigated.

For each outcome, linear regression was used to calculate the mean difference (in minutes of physical activity) at baseline (in 2010) compared with post-intervention (in 2017) (and 95% confidence interval) after adjusting for age, season, education, car ownership and area deprivation. Multi-level models were fitted using a random intercept at the super output area (individuals within super output areas) to account for clustering within areas. Analyses were repeated with physical activity category (as per groups defined by the UK government guidelines) as the outcome using logistic regression. The analyses were stratified by distance from the CCG (exposure to the intervention) and deprivation. Interaction tests were conducted by fitting interaction terms within regression models.

Where comparable outcome data were available for the region from the Health Survey for Northern Ireland (2010 and 2017), effects in the CCG intervention sample were compared with any parallel trends for the Northern Ireland adult population. Due to the different physical activity measurement instruments used for data collection, we have not undertaken any formal statistical testing but rather show trends over time (Appendix D).

To investigate effects on health inequalities, we undertook a stratified analysis to assess whether any impacts on the primary outcome were socially patterned. We assessed whether the differences in the social patterning of outcomes changed over time compared to baseline. Such analyses adopted the approach of Hunter et al. [50] using logistic regression to make model-based outcome predictions across strata. All analyses were undertaken in Stata 13 (StataCorp. Stata Statistical Software: Release 13. College Station, TX: StataCorp LP. College Station, Texas (TX): StataCorp LP; 2013).

Results

Population characteristics

The intervention area sample at baseline (n = 1037) and post-intervention (n = 968) had similar characteristics (Table 1). The mean ages of the intervention area samples were 50.3 (SD 18.9) years at baseline and 51.7 (SD 19.1) years at follow-up. More males were recruited at follow-up (44.5%) compared to baseline (41.0%). Samples were similar in terms of the number of participants from the most deprived quintile (22.8% at baseline versus 23.3% at follow-up).

Table 1 Sample Characteristics (intervention sample only (i.e. ≤1 mile radius from the greenway)

Figure 2 shows the participant flow diagram. For the baseline sample (2010), 2594 addresses were issued from which 1205 completed interviews were achieved (46.6%). However, where contact was made with a household, 65.7% participated (628 refused to participate or were not available during survey period). For the follow-up sample (2017), 2502 addresses were issued from which 1214 completed interviews were achieved (48.5%). However, where contact was made with a household, 69.7% participated (529 refused to participate or were not available during survey period).

Effect of the CCG on physical activity behaviour

Table 2 presents the difference in mean primary and secondary outcomes (and proportions meeting physical activity guidelines) between baseline and follow-up for the intervention area (i.e., those living in wards ≤1 mile of the CCG). Table 3 presents the difference in mean primary and secondary outcomes (and proportions meeting physical activity guidelines) between baseline and follow-up for the intervention area versus control areas (i.e., those living in wards > 1 mile from the CCG).

Table 2 The difference in mean outcome between baseline and follow-up (intervention sample only)
Table 3 The difference in mean outcome between baseline and follow-up (intervention sample and control sample)

There was a significant decline in the proportion of the local population meeting the UK physical activity guidelines. At baseline, 68% of participants met the physical guidelines, which declined to 61% at follow-up (adjusted OR 0.65; 95% CI 0.52 to 0.82; p = < 0.001 (Table 2)). This decline is broadly in line with the Northern Ireland population which has seen a decline of 6% of adults meeting the UK physical activity guidelines over a similar time period (see Appendix D [51];).

For the control area, 64% met physical activity guidelines at baseline versus 67% at follow-up (adjusted OR 1.04; 95% CI 0.61 to 1.79; p = 0.877 (Table 3), a change which was not statistically significant, nor different statistically to the change seen in the intervention area (p-value for interaction of 0.119).

Similarly, the mean minutes of total physical activity reduced from 89.9 min per day before the intervention to 72.6 min per day after the intervention, corresponding to a mean reduction of 16 min per day (adjusted difference in mean − 16.0, 95% CI − 26.0 to − 6.0; p = 0.002) (Table 2). For the control area, the mean minutes of total physical activity reduced from 119 before to 78 min per day after the intervention corresponding to a mean reduction of 44 min per day (adjusted difference in mean − 44.4, 95% CI − 82.2 to − 6.6; p = 0.024)), but again this is not markedly different to that seen in the intervention area (p-value for interaction of 0.122) (Table 3).

Table 4 presents the results of the analysis of exposure, based on distance to the CCG for the intervention sample. For the primary outcome, 64–70% of the intervention population met the physical activity guidelines before the intervention. However, post-intervention 65% (adjusted OR 0.60, 95% CI 0.35 to 1.00; p = 0.051) of those living closest to the CCG (i.e. ≤400 m) and 60% (adjusted OR, 0.64 95% CI 0.41 to 0.99; p = 0.044) of those living furthest from the CCG (i.e. ≥ 1200 m) met the physical activity guidelines.

Table 4 The difference in mean outcome between baseline and follow-up (intervention sample) based upon distance to the greenway

Effect of the CCG on mental wellbeing, quality of life, social and perceptions of the environment outcomes

There was little evidence of a difference in mean WEMWBS before (mean = 50.6) compared with after (mean = 51.2) the intervention (adjusted difference in mean = − 0.3; 95% CI − 1.1 to 0.6; p = 0.524) (Table 3). In the control area, there was evidence of a slight decline in mental wellbeing from before (mean = 52.2) to after (mean = 51.3) the intervention (adjusted mean difference − 2.4; 95% CI − 4.3 to − 0.5; p = 0.018), though this was not markedly different from the intervention area (p-value for interaction =0.075). This decline is less than the best estimates of meaningful change which have been reported as being between 3 and 8 points [52].

Table 4 shows that, for those living closest to the CCG (i.e. ≤400 m) the mean WEMWEBS at baseline was 50.4 and at follow-up was 52.6 (adjusted difference 0.6 (95% CI − 0.9 to 2.1; p = 0.447)) and in those living furthest from the CCG (i.e. ≤1200 m) but within the intervention sample (and intervention area) the mean WEMWEBS was 51.6 at baseline and 51.3 at follow-up (adjusted difference in mean compared with baseline of − 0.7 95% CI − 2.0 to 0.5; p = 0.239).

There was evidence of a decline in mean quality of life from before (EQ5D mean = 73.3) to after (mean = 63.2) the intervention (adjusted difference in mean = − 11.0; 95% CI − 14.5 to − 7.4; p < 0.001). In the control area, there was evidence of a greater decline in quality of life, comparing before (mean = 76.2) versus after (mean = 54.3) the intervention (adjusted mean difference − 30.5; 95% CI − 37.9 to − 23.2; p = < 0.001), which was markedly different from the intervention area (p for interaction = < 0.001) (Table 3). Compared with the mean before of 71.9, the mean EQ5D score in those living closest to the CCG (i.e., ≤400 m) at follow-up was 63.2 (adjusted difference − 11.4 (95% CI − 15.7 to − 7.1; p = < 0.001)); for those living furthest from the CCG (i.e., ≤1200 m) mean EQ5D score changed from 75.2 to 65.7 (adjusted difference − 10.7 (95% CI − 14.5 to − 6.8; p = < 0.001 (Table 4)).

There were mixed findings for social capital. Table 3 shows a small improvement in local area trust between baseline and follow-up (adjusted mean difference 0.1; 95% CI 0.1 to 0.2; p = < 0.001) and a small decline in social networks (i.e., contact with friends, family, neighbours) (adjusted mean difference − 0.1; 95% CI − 0.2 to 0.0; p = 0.003) for those in the intervention area. Similar trends were found for the control area (see Tables 2 and 3) but no significant interaction was found. The exposure analysis in Table 4 shows no distinct trend for social capital constructs.

Perceptions of the environment for attractiveness (adjusted mean difference 0.3; 95% CI 0.2 to 0.3; p = < 0.001), traffic (adjusted mean difference 0.2; 95% CI 0.1 to 0.2; p = < 0.001) and safety (adjusted mean difference 0.2; 95% CI 0.1 to 0.3; p = < 0.001) improved over time in the intervention area, but not for access to amenities (adjusted mean difference 0.1; 95% CI 0.0 to 0.1; p = 0.04) (Table 2). Similar trends were found for the control area (Table 3), though with no evidence for any significant interaction between the intervention and control area. The exposure analysis in Table 4 shows no distinct trend for constructs of perception of the environment in the intervention sample.

Equity impact of the CCG

Table 5 presents the results for the primary outcome by area level deprivation to investigate the impact of the CCG on equity. Results show that those in the most deprived quintiles had a similar reduction in physical activity behaviour compared to those in the lesser deprived quintiles.

Table 5 Physical activity category by area deprivation and year (intervention sample)

Discussion

Our study has shown declines in physical activity behaviour in the CCG area at the follow-up period compared to baseline in the intervention area. This is broadly in line with the decrease in physical activity levels in the adult population in Northern Ireland. Results also showed a decline in quality of life at follow-up compared to baseline in the intervention area, albeit this decline was significantly less than in the control area. There was little evidence of a change in mental wellbeing in the intervention area. There was however evidence of improvements in some constructs of social capital, namely local area trust, but not for increased contact with social networks in the intervention area. Results also showed increased perceptions of the local environment for attractiveness, traffic and safety, but not for access to amenities in the intervention area. Overall, our analyses showed no evidence for a pattern of impacts for those living closer to the greenway (than those living farther away) and that these impacts were broadly similar for those living in deprived and affluent neighbourhoods.

Compare/contrast with previous literature

Our study adds to the relatively scant literature on assessing the public health impact of urban greenways. The recent review by Hunter et al. [21] identified only six previous studies that evaluated the public health impact of urban greenways and trails. All studies that employed a dual approach (i.e., those that combined a change to the physical environment with promotion activities) demonstrated a significant intervention effect. For example, Fitzhugh et al. [53] showed significant effects for total physical activity and cycling when investigating the impact of a greenway in an urban area in Tennessee, USA. The intervention involved retrofitting 2.9 miles of urban greenway to enhance pedestrian connectivity costing $US 2.1 m. The follow-up period was 14 months after construction of the urban greenway was completed.

The IConnect study by Ogilvie et al. [54], investigated the impact of building or improving walking and cycling routes in three cities in the UK. This large quasi-experimental study (n = 1796 participants) conducted follow-up surveys at 1 year and 2 years. Results demonstrated that proximity to the new walking/cycling infrastructure was associated with greater usage. At 1-year follow-up, a 32% increase in usage was reported, which increased to 38% at 2-year follow-up. At 2-year follow-up, the study also showed that those living nearer the new infrastructure significantly increased walking and cycling levels (by 15 mins/week/km) and overall physical activity levels (by 12.5 mins/week/km) [55].

It is important to note that both studies [53, 55] had follow-up periods > 1 year post-construction and, in the case of Goodman et al. [55], the post-intervention effect was not realised until the 2-year follow-up period. Therefore, we could hypothesise that, given our post-intervention evaluation started only 6 months after completion of construction, it might be too early to see population-level changes in our study. Albeit, we acknowledge that our data collection took place over 12 months so some respondents could potentially have experienced the greenway for up to 18 months. The findings by Fitzhugh et al. [53] and Goodman et al. [55] suggest that it takes at least 1 year after completion of construction for the local population to become aware of the new infrastructure and to become regular users. However, these studies largely focused on physical activity behaviours and so we do not know the impact of these interventions on health, wellbeing, social or perceptions of the environment outcomes.

Importance of context

Population level interventions do not operate in a vacuum, and so it is important to note the context in which the Connswater Community Greenway natural experiment took place [56]. The construction period and post-completion of the greenway was at a time of significant economic austerity, resulting in uncertainty regarding job security and job location; a time when health and wellbeing was at its most vulnerable. During much of this period there was also no Government Executive in place in Northern Ireland and so many policy initiatives were stalled which could have had wider effects on investment and other interventions in the area. Contextual issues of security and wider political aspects have been the subject of discussion in our previous work in this area [57].

“Natural experiments” are challenging to evaluate because this sort of added social complexity makes it difficult to control for the many factors and “moving parts” that generate community level outcomes. Population level data from the Northern Ireland Statistics and Research Agency (NISRA) [58] shows that east Belfast (i.e., the part of the city in which the greenway was developed) has worsening profiles for health and education over the 8-year period of our study. Data from other government surveys (see Appendix D) also showed a decline in levels of physical activity across the country.

Following the theory of the socio-ecological model, changes in major structural effects like land uses should drive more significant changes in behaviour (e.g., increased local destinations, more footfall, more social interaction) and land markets, particularly in a deprived area like east Belfast are very slow to evolve and stimulate change. These effects may not be discernible for many years to come. Given that the primary focus of the CCG was on connecting and improving existing greenspace, the CCG largely depends on induced land use change in the surrounding areas rather than the intervention itself producing such changes, as would be the case with a totally new greenway or park, or conversion of previous private land to public access.

As emphasised earlier, there was a seven-year gap between when the baseline and follow-up surveys were conducted. It is likely that societal trends during this extended period could have largely confounded the effect of the intervention. For example, we see slight change in the demography of the local population over this timeframe, such as, slightly wealthier, and more highly educated people living in rental properties.

Our findings also demonstrate that a relatively small amount of physical activity behaviour (variance/R2 value) was explained by our measured variables. The proportion of the variance explained was estimated to be 11% at baseline and 20% post-intervention. There are other factors affecting physical activity behaviour that we have not captured. For example, based on the socio-ecological model, genetic factors, psychological variables such as attitudes, intention to exercise, self-efficacy, physical activity history and transport environment have been identified as known correlates and determinants of adult physical activity [59]. In addition, contextual factors have been highlighted as important for the effects of built environment interventions [22].

Despite the modest improvements in some secondary outcomes, the primary outcome (proportion meeting UK physical activity guidelines) declined significantly at post-intervention follow-up. These results pose several scientific and real-world implementation challenges that are too infrequently exposed in public heath intervention trials [60], including how to balance positive and negative results when primary and secondary outcomes are discordant [61]. Some largely qualitative approaches (with interpretations contingent on programme theory) have been proposed for learning from multiple outcomes, such as Contribution Analysis with Process Tracing [62] and Ripple Effects Mapping [63], analogous to Cost Consequence analysis in health economics, while more quantitative approaches adopting a Bayesian decision framework have also been advocated [64]. There is a long-standing belief that positive results are favoured by scientific journals and that this may contribute to “publication bias”. On the other hand, some journals claim now to select articles for publication based on their contribution to the literature and welcome null results that challenge conventional wisdom or prior expectations [65]. The results from our study certainly challenged prior expectations. However, it is notoriously hard to disprove any hypothesis, and so negative studies must have the precision and strength of design to be reasonably persuasive.

Strengths and limitations

This study took advantage of a rare natural experiment opportunity to investigate the public health impact of a new urban greenway. To the best of our knowledge, this study is one of the first to evaluate the multi-functional nature of UGS interventions, viewed through a health, wellbeing, social, perceptions of the environment and equity lens [21].

The study adds to the rather sparse evidence base investigating the impact of urban greenways which has largely been ‘opportunistic’ studies undertaking process evaluations focussing on usage, using uncontrolled pre/post study designs and collecting data using observational, intercept or e-counter methods [66]. We have also addressed a number of key limitations and sources of bias identified by Benton et al. [66] in a review detailing the risk of bias inherent in built environment interventions for promoting physical activity. For example, we published our study protocol which outlined our analytical approach a priori [25]. We also employed a number of well-matched comparator analyses including sampling survey respondents in electoral wards greater than 1 mile from the intervention site (albeit the sample size for the control group was much less that the intervention area) and investigating the impact of the greenway using distance from greenway as a measure of exposure. Since we employed a repeated, cross-sectional design to investigate population-level impacts, we had minimal missing data (as we might have anticipated with a longitudinal design). Our sampled population was also representative of both the east Belfast population and wider Northern Ireland population in terms of age, gender and deprivation profiles, variables which we adjusted for in our analyses.

Our study did have a number of limitations including reliance on the use of self-report measures in the household survey. Validated instruments were used and standardised scoring protocols employed where possible. However, it is important to note the susceptibility of self-reported measures, in particular physical activity behaviour, to social desirability and recall bias [67,68,69], though we do not see why this might have differed before and after the greenway construction. If this bias was greater after the construction of the greenway, this would not explain our negative findings. Our study only sampled adults aged 16 years and over, so we do not know the impact of the greenway on children or adolescents. We were unable to compare the impact of the greenway on physical activity levels of the Northern Ireland population as originally intended, as the SAPAS survey was not re-commissioned due to austerity measures. However, based on data from another population-wide Northern Ireland household survey (see Appendix D), we did report similar trends in terms of a 6% decline in adult physical activity levels across the country. As we previously highlighted, our follow-up household survey commenced approximately 6 months after completion of the greenway which may have been too short a timeframe to see population-level changes based on the findings from Fitzhugh et al. [53] and Goodman et al. [55]. However, this is similar to findings by West and Shores [70] who showed no significant change in physical activity following the construction of a new greenway. We were unable to extend this timeframe due to funding restrictions: our originally planned 5-year study became an 8-year study. Despite these limitations, our study contributes to the evidence base on UGS interventions for population health outcomes, providing data on physical, mental, social and perceptions of the environment impacts, and highlights key methodological considerations, such as the selection of suitable control samples, for the evaluation of large-scale natural experiment studies.

Implications of findings

Our study adds to the evidence base on UGS interventions by the WHO Regional Office for Europe [3] and Hunter et al. [21] and our findings highlight the importance of viewing UGS interventions through the multi-functional lens of health, wellbeing, social, perceptions of the environment and equity impacts. To view UGS only through one lens under-evaluates the public health impact of such interventions. Our study has a particularly important message for policy-makers – such large-scale investments must be given time to mature and realise their true impact. From our qualitative data and discussions with local stakeholders, the CCG is having a significant impact on the local area in terms of increased number of visitors, attracting housing developments and business investments. The CCG is highly regarded as a vanguard project, led by the local community, stimulating interest in future greenway investments across Northern Ireland [71]. However, these effects have not, to date, been translated into the variables that were collected in this study.

Future research

In order to address our hypothesis that the follow-up survey was conducted in too short a timeframe following completion of the greenway to realise population level changes, there is a need to conduct a further wave of the study. We also need to interrogate our longitudinal qualitative data from household interviews to further explore the reasons for our findings. This will be the subject of future publications and is beyond the scope of the current manuscript. We need new methodological approaches that can help us better understand mixed results from a range of outcomes, including those with conflicting and contrasting findings [61], enabling us to reflect on and update pre-defined logic models. Little is known about non-use of UGS, but the limited evidence suggests culture (including nature orientation), values, and capability (i.e., not wanting to or not feeling able to visit, for example, due to time poverty) are greater determinants of non-use than lack of nearby UGS which requires further exploration.

Conclusion

Our findings showed no intervention effect for improvements in population level physical activity behaviour or health but did show some modest improvements for secondary outcomes including perceptions of environment and some constructs of social capital. Based on the results of previous similar studies, we hypothesise that the public health impact of the new urban greenway may take a longer period of time to be realised. The Connswater Community Greenway management team should continue to engage with the local communities to promote greater usage and leverage the capacity of communities for potential longer-term gains. Our findings illustrate the major challenge of evaluating complex urban interventions and the difficulty of capturing and measuring the network of potential outcomes and explanatory variables and the most appropriate time to evaluate impact.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Abbreviations

aOR:

Adjusted Odds Ratio

CCG:

Connswater Community Greenway

CCTV:

Closed-circuit Television

CI:

Confidence interval

EQ-5D-3L:

EuroQol-5 Dimensions-3 Levels

GIS:

Geographic Information Systems

GPAQ:

Global Physical Activity Questionnaire

IPAQ:

International Physical Activity Questionnaire

km:

Kilometre

mins:

Minutes

NCD:

Non-communicable disease

NISRA:

Northern Ireland Statistics and Research Agency

PAF:

Postal Address File

PARC:

Physical Activity and the Rejuvenation of Connswater

SD:

Standard deviation

SDG:

Sustainable Development Goals

UGS:

Urban green space

UK:

United Kingdom

UN:

United Nations

USA:

United States of America

WEMWBS:

Warwick Edinburgh Mental Wellbeing Scale

WHO:

World Health Organisation

References

  1. 1.

    United Nations Sustainable Development Goals, 2015. https://www.un.org/sustainabledevelopment/sustainable-development-goals/.

  2. 2.

    WHO Regional Office for Europe. Urban green spaces and health. Copenhagen: WHO Regional Office for Europe; 2016. https://www.euro.who.int/en/health-topics/environment-and-health/urban-health/publications/2016/urban-green-spaces-and-health-a-review-of-evidence-2016

    Google Scholar 

  3. 3.

    WHO Regional Office for Europe. Urban green space interventions and health: a review of impacts and effectiveness. Copenhagen: WHO Regional Office for Europe; 2017a. https://www.euro.who.int/en/health-topics/environment-and-health/urban-health/publications/2017/urban-green-space-interventions-and-health-a-review-of-impacts-and-effectiveness.-full-report-2017

    Google Scholar 

  4. 4.

    WHO Regional Office for Europe. Action brief on urban green space interventions. Copenhagen: WHO regional Office for Europe; 2017b. https://www.euro.who.int/__data/assets/pdf_file/0010/342289/Urban-Green-Spaces_EN_WHO_web3.pdf%3Fua=1

    Google Scholar 

  5. 5.

    World Health Organization. Global action plan on physical activity 2018–2030: more active people for a healthier world. World Health Organization. 2018. https://apps.who.int/iris/handle/10665/272722. License: CC BY-NC-SA3.0 IGO

  6. 6.

    United Nations New Urban Agenda, 2016. http://habitat3.org/the-new-urban-agenda/. Accessed 26 Oct 2021.

  7. 7.

    Twohig-Bennett C, Jones A. The health benefits of the great outdoors: a systematic review and meta-analysis of greenspace exposure and health outcomes. Environ Res. 2018;166:628–37. https://doi.org/10.1016/j.envres.2018.06.030.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  8. 8.

    Bedimo-Rung AL, Mowen AJ, Cohen DA. The significance of parks to physical activity and public health: a conceptual model. Am J Prev Med. 2005;28(2 Suppl 2):159–68. https://doi.org/10.1016/j.amepre.2004.10.024.

    Article  PubMed  Google Scholar 

  9. 9.

    Bowler DE, Buyung-Ali LM, Knight TM, Pullin AS. A systematic review of evidence for the added benefits to health of exposure to natural environments. BMC Public Health. 2010a;10:456. https://doi.org/10.1186/1471-2458-10-456.

    Article  PubMed  PubMed Central  Google Scholar 

  10. 10.

    Bowler DE, Buyung-Ali L, Knight TM, Pullin AS. Urban greening to cool towns and cities: a systematic review of the empirical evidence. Landsc Urban Plann. 2010b;97(3):147–55 https://doi.org/10.1016/j.landurbplan.2010.05.006.

    Article  Google Scholar 

  11. 11.

    Bragg R and Atkins G. A review of nature-based interventions for mental health care. Natural England Commissioned Reports (number 204); 2016.

    Google Scholar 

  12. 12.

    Husk K, Lovell R, Cooper C, Stahl-Timmins W, Garside R. Participation in environmental enhancement and conservation activities for health and well-being in adults. Cochrane Database Syst Rev. 2016;5:CD010351. https://doi.org/10.1002/14651858.CD010351.pub2.

    Article  Google Scholar 

  13. 13.

    Lachowycz K, Jones AP. Greenspace and obesity: a systematic review of the evidence. Obes Rev. 2011;12(5):e183–9. https://doi.org/10.1111/j.1467-789X.2010.00827.x.

    CAS  Article  PubMed  Google Scholar 

  14. 14.

    Lee AC, Maheswaran R. The health benefits of urban green spaces: a review of the evidence. J Public Health. 2011;33(2):212–22. https://doi.org/10.1093/pubmed/fdq068.

    CAS  Article  Google Scholar 

  15. 15.

    Maas J, van Dillen SME, Verheij RA, Groenewegen PP. Social contacts as a possible mechanism behind the relation between green space and health. Health Place. 2009a;15(2):586–95. https://doi.org/10.1016/j.healthplace.2008.09.006.

    Article  PubMed  Google Scholar 

  16. 16.

    Maas J, Verheij RA, de Vries S, Spreeuwenberg P, Schellevis FG, Groenewegen PP. Morbidity is related to a green living environment. J Epidemiol Community Health. 2009b;63(12):967–73. https://doi.org/10.1136/jech.2008.079038.

    CAS  Article  PubMed  Google Scholar 

  17. 17.

    Mitchell R, Popham F. Effect of exposure to natural environment on health inequalities: an observational population study. Lancet. 2008;372(9650):1655–60. https://doi.org/10.1016/S0140-6736(08)61689-X.

    Article  PubMed  Google Scholar 

  18. 18.

    Mitchell RJ, Richardson EA, Shortt NK, Pearce JR. Neighborhood environments and socioeconomic inequalities in mental well-being. Am J Prev Med. 2015;49(1):80–4. https://doi.org/10.1016/j.amepre.2015.01.017.

    Article  PubMed  Google Scholar 

  19. 19.

    Panter J, Guell C, Humphreys D, Ogilvie D. Can changing the physical environment promote walking and cycling? A systematic review of what works and how. Health Place. 2019;58:102161. https://doi.org/10.1016/j.healthplace.2019.102161.

    Article  PubMed  PubMed Central  Google Scholar 

  20. 20.

    Song C, Ikei H, Miyazaki Y. Physiological effects of nature therapy: a review of the research in Japan. Int J Environ Res Public Health. 2016;13(8). https://doi.org/10.3390/ijerph13080781.

  21. 21.

    Hunter RF, Cleland C, Cleary A, Droomers M, Wheeler BW, Sinnett D, et al. Environmental, health, wellbeing, social and equity effects of urban green space interventions: a meta-narrative evidence synthesis. Environ Int. 2019;130:104923. https://doi.org/10.1016/j.envint.2019.104923.

    CAS  Article  PubMed  Google Scholar 

  22. 22.

    Panter J, Guell C, Prins R, Ogilvie D. Physical activity and the environment: conceptual review and framework for intervention research. Int J Behav Nutr Phys Act. 2017;14(1):156. https://doi.org/10.1186/s12966-017-0610-z.

    Article  PubMed  PubMed Central  Google Scholar 

  23. 23.

    Hunter RF, Christian H, Veitch J, Astell-Burt T, Hipp JA, Schipperijn J. The impact of interventions to promote physical activity in urban green space: a systematic review and recommendations for future research. Soc Sci Med. 2015a;124:246–56. https://doi.org/10.1016/j.socscimed.2014.11.051.

    Article  PubMed  Google Scholar 

  24. 24.

    Hunter RF, Dallat MAT, Tully MA, Heron L, O’Neill C, Kee F. Social return on investment analysis of an urban greenway. Cities Health. 2020. https://doi.org/10.1080/23748834.2020.1766783.

  25. 25.

    Tully MA, Hunter RF, McAneney H, Cupples ME, Donnelly M, Ellis G, et al. Physical activity and the rejuvenation of Connswater (PARC study): protocol for a natural experiment investigating the impact of urban regeneration on public health. BMC Public Health. 2013;13:774. https://doi.org/10.1186/1471-2458-13-774.

    Article  PubMed  PubMed Central  Google Scholar 

  26. 26.

    Dallat MAT, Soerjomataram I, Hunter RF, Tully MA, Cairns KJ, Kee F. Urban greenways have the potential to increase physical activity levels cost-effectively. Eur J Pub Health. 2014;24(2):190–5. https://doi.org/10.1093/eurpub/ckt035.

    Article  Google Scholar 

  27. 27.

    National Statistics: Northern Ireland multiple deprivation measure; 2005. https://www.nisra.gov.uk/statistics/deprivation/northern-ireland-multiple-deprivation-measure-2005-nimdm2005.

    Google Scholar 

  28. 28.

    Sallis JF, Cervero RB, Ascher W, Henderson KA, Kraft MK, Kerr J. An ecological approach to creating active living communities. Annu Rev Public Health. 2006;27:297–322. https://doi.org/10.1146/annurev.publhealth.27.021405.102100.

    Article  PubMed  Google Scholar 

  29. 29.

    Agrawal AW, Schlossberg M, Irvin K. How far, by which route and why? A spatial analysis of pedestrian preference. J Urban Des. 2008;13(1):81–98. https://doi.org/10.1080/13574800701804074.

    Article  Google Scholar 

  30. 30.

    Colabianchi N, Dowda M, Pfeiffer KA, Porter DE, Almeida MJ, Pate RR. Towards an understanding of salient neighborhood boundaries: adolescent reports of an easy walking distance and convenient driving distance. Int J Behav Nutr Phys Act. 2007;4:66. https://doi.org/10.1186/1479-5868-4-66.

    Article  PubMed  PubMed Central  Google Scholar 

  31. 31.

    Yang Y, Diez-Roux AV. Walking distance by trip purpose and population subgroups. Am J Prev Med. 2012;43(1):11–9. https://doi.org/10.1016/j.amepre.2012.03.015.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  32. 32.

    Bull FC, Maslin TS, Armstrong T. Global physical activity questionnaire (GPAQ): nine country reliability and validity study. J Phys Act Health. 2009;6(6):790–804. https://doi.org/10.1123/jpah.6.6.790.

    Article  Google Scholar 

  33. 33.

    Cleland CL, Hunter RF, Kee F, Cupples ME, Sallis JF, Tully MA. Validity of the global physical activity questionnaire (GPAQ) in assessing levels and change in moderate-vigorous physical activity and sedentary behaviour. BMC Public Health. 2014;14:1255. https://doi.org/10.1186/1471-2458-14-1255.

    Article  PubMed  PubMed Central  Google Scholar 

  34. 34.

    Department of Health: UK physical activity guidelines; 2011. https://www.gov.uk/government/publications/uk-physical-activity-guidelines.

    Google Scholar 

  35. 35.

    Department of Health: UK chief medical officers’ physical activity guidelines; 2019. https://www.gov.uk/government/publications/physical-activity-guidelines-uk-chief-medical-officers-report.

    Google Scholar 

  36. 36.

    Lloyd P, Devine P. Psychometric properties of the Warwick-Edinburgh Mental Well-Being Scale (WEMWBS) in Northern Ireland. J Mental Health. 2012;21(3):257–63. https://doi.org/10.3109/09638237.2012.670883.

    Article  Google Scholar 

  37. 37.

    Tennant R, Hiller L, Fishwick R, Platt S, Joseph S, Weich S, et al. The Warwick-Edinburgh Mental Well-Being Scale (WEMWBS): development and UK validation. Health Qual Life Outcomes. 2007;5:63. https://doi.org/10.1186/1477-7525-5-63.

    Article  PubMed  PubMed Central  Google Scholar 

  38. 38.

    Brooks R. EuroQol: the current state of play. Health Policy. 1996;37:53–72. https://doi.org/10.1016/0168-8510(96)00822-6.

    CAS  Article  PubMed  Google Scholar 

  39. 39.

    The EuroQol Group. EuroQol-a new facility for the measurement of health-related quality of life. Health Policy. 1990;16:199–208. https://doi.org/10.1016/0168-8510(90)90421-9.

    Article  Google Scholar 

  40. 40.

    Ogilvie D, Mitchell R, Mutrie N, Petticrew M, Platt S. Perceived characteristics of the environment associated with active travel: development and testing of a new scale. Int J Behav Nutr Phys Act. 2008;5:32. https://doi.org/10.1186/1479-5868-5-32.

    Article  PubMed  PubMed Central  Google Scholar 

  41. 41.

    Babb P. Measurement of social capital in the UK. London: Office for National Statistics; 2005.

    Google Scholar 

  42. 42.

    Foxton F, Jones R. Social capital indicators review. London: Office of National Statistics; 2011.

    Google Scholar 

  43. 43.

    Ellis G, Hunter RF, Tully MA, Donnelly M, Kelleher L, Kee F. Connectivity and physical activity: using footpath networks to measure the walkability of built environments. Environ Plan B Plann Des. 2016;43(1):130–51. https://doi.org/10.1177/0265813515610672.

    Article  Google Scholar 

  44. 44.

    Sallis JF, Cerin E, Conway TL, Adams MA, Frank LD, Pratt M, et al. Physical activity in relation to urban environments in 14 cities worldwide: a cross-sectional study. Lancet. 2016;387(10034):2207–17. https://doi.org/10.1016/S0140-6736(15)01284-2.

    Article  PubMed  Google Scholar 

  45. 45.

    Boer R, Zheng Y, Overton A, Ridgeway GK, Cohen DA. Neighbourhood design and walking trips in ten U.S. metropolitan areas. Am J Prev Med. 2007;32:298–304. https://doi.org/10.1016/j.amepre.2006.12.012.

    Article  PubMed  PubMed Central  Google Scholar 

  46. 46.

    Learnihan V, Van Niel KP, Giles-Corti B, Knuman M. Effect of scale on the links between walking and urban design. Geogr Res. 2011;49(2):183–91. https://doi.org/10.1111/j.1745-5871.2011.00689.x.

    Article  Google Scholar 

  47. 47.

    Milward H, Spinney J, Scott D. Active-transport walking behaviour: destinations, durations, distances. J Transp Geogr. 2013;28:101–10. https://doi.org/10.1016/j.jtrangeo.2012.11.012.

    Article  Google Scholar 

  48. 48.

    Sport Northern Ireland: The Northern Ireland Sport and Physical Activity Survey (SAPAS). A baseline report; 2010. https://www.scribd.com/document/73014459/The-Northern-Ireland-Sport-and-Physical-Activity-Survey-2010-SAPAS.

    Google Scholar 

  49. 49.

    Hagstromer M, Oja P, Sjostrom M. The international physical activity questionnaire (IPAQ): a study of concurrent and construct validity. Public Health Nutr. 2006;9:755–62. https://doi.org/10.1079/PHN2005898.

    Article  PubMed  Google Scholar 

  50. 50.

    Hunter RF, Boeri M, Tully MA, Donnelly P, Kee F. Addressing inequalities in physical activity participation: implications for public health policy and practice. Prev Med. 2015b;72:64–9. https://doi.org/10.1016/j.ypmed.2014.12.040.

    Article  PubMed  Google Scholar 

  51. 51.

    Department of Health, Northern Ireland. Physical activity statistics; 2018. https://www.health-ni.gov.uk/articles/physical-activity-statistics.

    Google Scholar 

  52. 52.

    Putz R, O’Hara K, Taggart F, Stewart-Brown S. Using WEMWBS to measure the impact of your work on mental wellbeing: a practice-based user guide. Coventry: Coventry City Council; 2012.

    Google Scholar 

  53. 53.

    Fitzhugh EC, Bassett DR, Evans MF. Urban trails and physical activity: a natural experiment. Am J Prev Med. 2010;39:259–62. https://doi.org/10.1016/j.amepre.2010.05.010.

    Article  PubMed  Google Scholar 

  54. 54.

    Ogilvie D, Bull F, Powell J, Cooper AR, Brand C, Mutrie N, et al. An applied ecological framework for evaluating infrastructure to promote walking and cycling: the iConnect study. Am J Public Health. 2011;101(3):473–81. https://doi.org/10.2105/AJPH.2010.198002.

    Article  PubMed  PubMed Central  Google Scholar 

  55. 55.

    Goodman A, Sahlqvist S, Ogilvie D, iConnect Consortium. New walking and cycling routes and increased physical activity: one- and 2-year findings from the UK iConnect study. Am J Public Health. 2014;104:e38–46. https://doi.org/10.2105/AJPH.2014.302059.

    Article  PubMed  PubMed Central  Google Scholar 

  56. 56.

    Craig P, Di Ruggiero E, Frohlich KL, Mykhalovskiy E, White M, on behalf of the Canadian Institutes of Health Research (CIHR)–National Institute for Health Research (NIHR) Context Guidance Authors Group. Canadian Institutes of Health Research-National Institute for Health Research. Taking account of context in population health intervention research: guidance for producers, users and funders of research. Southampton: NIHR Journals Library; 2018.

    Google Scholar 

  57. 57.

    Prior L, Scott D, Hunter RF, Donnelly M, Tully MA, Cupples ME, et al. Exploring lay views on physical activity and their implications for public health policy. A case study from East Belfast. Soc Sci Med. 2014;114:73–80. https://doi.org/10.1016/j.socscimed.2014.05.015.

    CAS  Article  PubMed  Google Scholar 

  58. 58.

    Northern Ireland Statistics and Research Agency. National Statistics: Northern Ireland Multiple Deprivation Measure; 2017. https://www.nisra.gov.uk/statistics/deprivation/northern-ireland-multiple-deprivation-measure-2017-nimdm2017.

  59. 59.

    Bauman AE, Reis RS, Sallis JF, Wells JC, Loos RJF, Martin BW, et al. Correlates of physical activity: why are some people physically active and others not? Lancet. 2012;380(9838):258–71. https://doi.org/10.1016/S0140-6736(12)60735-1.

    Article  PubMed  Google Scholar 

  60. 60.

    Tannahill A, Kelly MP. Layers of complexity in interpreting evidence on effectiveness. Public Health. 2013;127(2):164–70. https://doi.org/10.1016/j.puhe.2012.11.011.

    CAS  Article  PubMed  Google Scholar 

  61. 61.

    Lilford RJ, Chilton PJ, Hemming K, Girling AJ, Taylor CA, Barach P. Evaluating policy and service interventions: framework to guide selection and interpretation of study end points. BMJ. 2010;341:c4413. https://doi.org/10.1136/bmj.c4413.

    Article  PubMed  Google Scholar 

  62. 62.

    Befani B, Mayne J. Process tracing and contribution analysis: a combined approach to generative causal inference for impact evaluation. IDS Bull. 2014;45(6):17–36.

    Article  Google Scholar 

  63. 63.

    Washburn LT, Traywick LV, Thornton L, Vincent J, Brown T. Using ripple effects mapping to evaluate a community-based health program: perspectives of program implementers. Health Promot Pract. 2020;21(4):601–10. https://doi.org/10.1177/1524839918804506.

    Article  PubMed  Google Scholar 

  64. 64.

    Kavelaars X, Mulder J, Kaptein M. Decision-making with multiple correlated binary outcomes in clinical trials. Stat Methods Med Res. 2020;29(11):3265–77. https://doi.org/10.1177/0962280220922256.

    Article  PubMed  PubMed Central  Google Scholar 

  65. 65.

    Wilcox AJ. A positive approach to negative results. Epidemiology. 2014;25(2):165. https://doi.org/10.1097/EDE.0000000000000022.

    Article  PubMed  Google Scholar 

  66. 66.

    Benton JS, Anderson J, Hunter RF, French DP. The effect of changing the built environment on physical activity: a quantitative review of the risk of bias in natural experiments. Int J Behav Nutr Phys Act. 2016;13:107. https://doi.org/10.1186/s12966-016-0433-3.

    Article  PubMed  PubMed Central  Google Scholar 

  67. 67.

    Adams SA, Matthews CE, Ebbeling CB, Moore CG, Cunningham JE, Fulton J, et al. The effect of social desirability and social approval on self-reports of physical activity. Am J Epidemiol. 2005;161(4):389–98. https://doi.org/10.1093/aje/kwi054.

    Article  PubMed  Google Scholar 

  68. 68.

    Ferrari P, Friedenreich C, Matthews CE. The role of measurement error in estimating levels of physical activity. Am J Epidemiol. 2007;166:832–40. https://doi.org/10.1093/aje/kwm148.

    Article  PubMed  Google Scholar 

  69. 69.

    Sallis JF, Saelens BE. Assessment of physical activity by self-report: status, limitations, and future directions. Res Q Exerc Sport. 2000;71(2 Suppl):S1–14.

    CAS  Article  Google Scholar 

  70. 70.

    West ST, Shores KA. The impacts of building a greenway on proximate residents’ physical activity. J Phys Act Health. 2011;8(8):1092–7. https://doi.org/10.1123/jpah.8.8.1092.

    Article  PubMed  Google Scholar 

  71. 71.

    Department for Infrastructure, Northern Ireland. Exercise, explore, enjoy: a strategic plan for greenways, 2016. https://www.infrastructure-ni.gov.uk/sites/default/files/publications/infrastructure/exercise-explore-enjoy-a-strategic-plan-for-greenways-november-2016-final.pdf.

    Google Scholar 

Download references

Acknowledgements

The authors would like to acknowledge the partners and stakeholders involved in the PARC Study including the Eastside Partnership; Belfast City Council; Department of Health, Social Services and Public Safety; Department for Communities; Department of the Environment; Department for Infrastructure; Belfast Health and Social Care trust; East Belfast Community Development Agency; Sport Northern Ireland; Belfast Healthy Cities; Sustrans; Public Health Agency; Ordnance Survey NI and local residents of the Connswater Community Greenway population.

Funding

The PARC study is supported by a grant from the National Prevention Research Initiative (NPRI) (grant number: G0802045/1). The funding partners are (in alphabetical order): Alzheimer’s Research Trust; Alzheimer’s Society; Biotechnology and Biological Sciences Research Council; British Heart Foundation; Cancer Research UK; Chief Scientist Office; Scottish Government Health Directorate; Department of Health; Diabetes UK; Economic and Social Research Council; Engineering and Physical Sciences Research Council; Health and Social Care Research and Development Division of the Public Health Agency (HSC R&D Division); Medical Research Council; The Stroke Association; Welsh Assembly Government and World Cancer Research Fund. We wish to also acknowledge funding from the UKCRC Centre of Excellence for Public Health Northern Ireland. The authors declare that the funders had no role in any aspect of the PARC study, including the study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the article for publication.

Author information

Affiliations

Authors

Contributions

RH led the writing of the manuscript. FK (Principal Investigator) had the original idea for the study and led the design of the study and the application for grant funding. RH, MT, HMcA, GE, GH, LP, MD, MS, and MC (co-investigators) contributed to the design of the study and writing of the funding application and manuscript. CC led the statistical analysis. SF and BJ led the GIS data element. All authors contributed to the conduct of the research and to the critical revision of the paper and approved the final version.

Corresponding author

Correspondence to Ruth F. Hunter.

Ethics declarations

Ethics approval and consent to participate

The study was approved by the Office for Research Ethics Committees, Northern Ireland (09/NIR02/66). Participants provided written informed consent prior to taking part in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Frank Kee is the principal investigator.

Supplementary Information

Additional file 1

: Appendix A. Description of the primary and secondary outcomes and measures.

Additional file 2

: Appendix B. Household Survey.

Additional file 3

: Appendix C. Control Sample Characteristics (Control area sample (i.e. >1 mile radius from the greenway).

Additional file 4

: Appendix D. Population trends in meeting physical activity recommendations in 2010 and 2016 by sex and age [Northern Ireland Health and Wellbeing Survey].

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Hunter, R.F., Adlakha, D., Cardwell, C. et al. Investigating the physical activity, health, wellbeing, social and environmental effects of a new urban greenway: a natural experiment (the PARC study). Int J Behav Nutr Phys Act 18, 142 (2021). https://doi.org/10.1186/s12966-021-01213-9

Download citation

Keywords

  • Urban green space
  • Intervention
  • Natural experiment
  • Physical activity
  • Health
  • Mental wellbeing