How Do Awareness, Perceptions, and Expectations of an Urban Greenway Influence Residents’ Visits and Recreational Physical Activity? Evidence From the Connswater Community Greenway, Belfast, Northern Ireland

Abstract

The development of an urban greenway as part of a major regeneration project may encourage usage and recreational physical activity (PA). However, how the awareness and perceptions of such a project influence the usage of the greenway and recreational PA after regeneration remains unclear. This study focused on these relationships in the context of a major urban regeneration project, the Connswater Community Greenway (CCG) in Belfast, Northern Ireland. Data were collected from residents living near the CCG at 6 months postimplementation (2016/2017). Awareness, perceptions, and expectations of the CCG were assessed using several self-reported questions. Respondents were split into intervention and control groups based on a 1-mile threshold distance to the CCG from their home address (intervention group ≤1 mile; control group >1 mile). Self-reported CCG usage included both frequency and duration of visits. Recreational PA was evaluated with the Global Physical Activity Questionnaire and PA stages of change measure. We used multilevel linear and logistic regressions to examine awareness, perceptions, and expectations of the CCG on visitation and recreational PA. Results showed that favorable awareness, perceptions, and expectations were positively associated with CCG visits and recreational PA. Frequent and longer durations of CCG visits were also important for encouraging recreational PA. These findings provide evidence that it is important to include awareness-raising programs and education regarding the urban greenway development for health benefits.

Keywords:

• awareness
• expectations
• perceptions
• recreational physical activity
• urban greenway

Introduction

Development of urban greenways is assumed to increase people’s access to nature and therefore can benefit people’s physical activity (PA; Hausmann et al., 2020; Hunter et al., 2021; Schwartz et al., 2019; Veitch et al., 2014). This typically involves improving the physical and natural environment, such as developing more facilities and vegetation, and social interventions, including organizing PA and social programs (Cohen et al., 2015; Duncan et al., 2021; Veitch et al., 2018; Vert et al., 2019). Due to these interventions, residents’ awareness, perceptions, and expectations of greenways may also change after urban regeneration (Slater et al., 2016; Tester & Baker, 2009). Residents usually receive information to raise awareness of new urban greenway development and associated hypothesized benefits as well as information regarding events and programs after the regeneration project, typically disseminated locally through various channels including social media among residents (Smyth, 2005). Hence, the local government also invests in making promotional videos and billboards to market the project, because this may help attract more visitors (Smyth, 2005). As for residents’ perceptions and expectations of urban greenways, change is usually associated with social interventions to increase the usage of the greenway (Razani et al., 2020). After the urban regeneration, there is usually more marketing information, and residents share personal experiences regarding the health benefits of visiting urban greenways (Razani et al., 2020). Also, the newly founded PA and social programs may help increase residents’ awareness of the benefits of visiting urban greenways (Hunter et al., 2021). Therefore, residents are more likely to have more positive perceptions of the urban greenway, especially regarding potential health benefits, postimplementation. The improvement of the physical and natural environment after the regeneration project is also important (Veitch et al., 2014). The regeneration project usually invests funding in improving current facilities and developing new facilities, so residents may have more positive perceptions and expectations of the new or improved facilities (Tester & Baker, 2009; Veitch et al., 2012). Therefore, it is necessary to combine both physical interventions (e.g., new infrastructures, vegetation) and social interventions (e.g., social programs) in urban greenways to boost their health benefits (Hunter et al., 2015, 2019; World Health Organization, 2017).

Favorable awareness, perceptions, and expectations of urban greenways have been found to influence the frequency and duration of visits postimplementation (Tester & Baker, 2009). First, more sources of information regarding awareness of the urban greenway are positively associated with duration of information exposure, so when residents have more exposure to information regarding the urban greenway, they are more likely to be aware of the greenway and be willing to visit or take exercise in it (Hunter et al., 2014; Kirova & Thanh, 2019; H. Lee et al., 2019; M. K. Lee et al., 2017; Ramukumba, 2018). Also, more positive awareness about the urban greenway can give residents a better impression of the greenway and make them more willing to visit (K. Lee, 2011). Second, people’s perceptions and expectations of a particular place may influence their intention to visit, because such perceptions reflect their belief regarding whether visits to the place may benefit them (Oviedo et al., 2022). For example, Chen et al. (2020) found that when people believe that visiting green space is important for health promotion, they are more likely and willing to visit, and Geng et al. (2021) suggested that visits to urban parks declined during the COVID-19 pandemic because people believed that visiting a park would increase the odds of being infected. Third, people’s perceptions and expectations of a place postimplementation refer to how they think the project has changed the place, especially the physical environment (Ferretti & Grosso, 2019; Korkmaz & Balaban, 2020). The more positive perceptions of urban greenways postimplementation, the more significant the likely impact on the area (Grilli et al., 2020). Therefore, if people have more positive perceptions and expectations of urban greenways, they may see the sites as more functional and better equipped and be more likely to visit (Grilli et al., 2020).

A large body of evidence demonstrates that the visits to a natural environment (e.g., green space and blue space) are important for facilitating physical activity (Cohen et al., 2012; Flowers et al., 2016; Temple et al., 2011; Uebel et al., 2022). First, the improvement of physical environments can make urban greenways more pleasant places for people to undertake PA (Floyd et al., 2011). Frequent and longer durations of urban greenway visits can maximize the restorative effects of the natural environment, such as relaxation (Jiang et al., 2016). When people feel more relaxed, they are more likely to undertake recreational PA (Stults-Kolehmainen & Sinha, 2014). Second, existing literature has documented that recreational PA in natural environments may generate more health benefits than recreational PA in other environments (Stults-Kolehmainen & Sinha, 2014). Frequent and longer durations of urban greenway visits can increase the amount of PA, including walking, so people are more likely to shift from being physically inactive to being regularly physically active in natural environments (Stults-Kolehmainen & Sinha, 2014). Third, according to the theory of planned behavior, frequent and longer durations of exposure to social activities may strengthen people’s normative beliefs about the activities and encourage them to continue (Ajzen, 1991). People usually see others performing recreational PA when visiting urban greenways, which may thus facilitate their involvement in more recreational PA behavior (Chatzisarantis & Hagger, 2005). Last, there are usually more facilities (e.g., signage, promotional material) in urban greenways postimplementation to increase awareness regarding the benefits of PA in natural environments (Evenson et al., 2016; Hunter et al., 2021). Frequent and longer durations of urban greenway visits may increase people’s awareness of such information and have an impact on people’s recreational PA.

Despite the growing evidence for the positive effect of urban green and blue space interventions including urban greenways, we are unaware of studies on how awareness, perceptions, and expectations of an urban greenway impact residents’ visits and PA postimplementation. Therefore, this study aimed to explore how awareness, perceptions, and expectations of an urban greenway are related to residents’ visits and recreational PA after a major urban regeneration project in Belfast, Northern Ireland. The conceptual framework of this study is shown in Figure 1. The conceptual framework shows the interconnection among awareness, perceptions, and expectations of an urban greenway. First, people receive information from an urban greenway and become aware of it. Second, people’s awareness of the greenway may further lead to different perceptions of it. Lastly, their awareness and perceptions of the greenway will help them develop expectations regarding the health benefits of the greenway. Therefore, the awareness, perceptions, and expectations of the greenway should influence their visits to the greenway and physical activity.

Figure 1. The conceptual framework.

We investigate the following null hypotheses:

(H₀1) There are no associations between awareness, perceptions, and expectations of an urban greenway and residents’ visits to it.

(H₀2) There are no associations between awareness, perceptions, and expectations of an urban greenway and residents’ recreational PA.

(H₀3) There are no associations between residents’ visits to an urban greenway and recreational PA.

Data and methods

Study design and setting

Connswater Community Greenway

The Connswater Community Greenway (CCG; www.communitygreenway.co.uk) is an urban regeneration project in Belfast that was funded by local governments and a Big Lottery Living Landmarks Award. The goal of this project is to facilitate more PA and encourage more use of the CCG through the construction of 19.4 km of new walkways along the concourse of three rivers and the adjacent walkways and cycle paths. The regeneration project involves the improvement of the built and natural environments coupled with programs of awareness and social activities. For example, the intervention included improvements to the esthetics of shared open spaces, and more trees and shrubs were planted. Also, community-based social marketing initiatives (e.g., brochures to introduce the CCG) were launched and PA groups (e.g., walking groups) were organized to encourage more people to visit the CCG. To increase people’s sense of security and use of CCG during the night, 24-h lighting and closed-circuit television were added after the regeneration project. Figure 2 displays some of the changes in the CCG postimplementation.

Figure 2. CCG (a) before and (b) after the regeneration project.

Source: The Connswater Community Greenway Trust, EastSide Partnership, 2022.

The Physical Activity and the Rejuvenation of Connswater

The Physical Activity and Rejuvenation of Connswater (PARC) Study used a repeated cross-sectional study design to evaluate the public health effects of the CCG (Hunter et al., 2021; Tully et al., 2013; Wang et al., 2023). Detailed information for the rejuvenation is provided in the supplemental material. The baseline (preimplementation) data were collected in 2010–2011 and the follow-up wave was conducted in 2016–2017 (6 months after the completion and opening of the CCG). We only use data from wave 2 (6 months postimplementation) because the CCG was opened after the implementation and thus the wave 1 data do not contain information regarding the usage of the CCG. Participants were recruited from a random sample of households living in electoral wards with a geographic centroid within a 1-mile radius of the CCG. In the first stage, 29 electoral wards (the primary administrative unit in Northern Ireland with 4,000 residents on average) in the political constituency of the CCG were selected (supplementary material, Figure S1). In the second stage, households were randomly chosen from the electoral wards using a random probability sampling method, which selected housing addresses based on the Royal Mail’s Postal Address File, stratified by the proportion of the population of each ward, to ensure representativeness for the general Northern Ireland population. At the final stage, one participant was chosen from each household using the “last birthday rule” (Oldendick et al., 1988). Details of the sampling procedure and protocol of the PARC study are available in Tully et al. (2013). In total, 1,214 participants were recruited for wave 2, and 892 samples were included for the final analysis after excluding missing relevant data. We conducted several Student’s t-tests or chi-square tests to examine whether target outcomes and distance to the CCG varied between excluded and included samples. The results indicated that there was no significant evidence that target outcomes and distance to the CCG varied between samples (p > .1). All participants provided written consent to participate and completed an interviewer-administered questionnaire. The study was approved by the Office for Research Ethics Committees, Northern Ireland (09/NIR02/66).

Measures

PA

Self-reported PA intensity

PA was measured by the Global Physical Activity Questionnaire, which assesses total moderate to vigorous PA time during work, active travel, and recreation in the last week (Bull et al., 2009). The Global Physical Activity Questionnaire has been validated for the Northern Ireland population for its reliability and capturing change over time (Cleland et al., 2014). We focused on recreational PA because most PA occurring on the CCG is mainly for recreational purposes (Hunter et al., 2021). Following existing studies (Ainsworth et al., 2011), the metabolic equivalent of task (MET) for weekly moderate-intensity recreational PA is equal to total moderate PA minutes * 4.0, and MET for weekly vigorous-intensity recreational PA is equal to total moderate PA minutes * 8.0.

Self-reported PA psychosocial outcomes

The PA stages of change measure (Carnegie et al., 2002) was used to define PARC participants’ intentions of being out of physically inactive (self-assessment) in the next 6 months. Participants were asked to answer “yes” or “no” to the following behavioral questions: “I am currently physically active,” “I intend to become more physically active in the next 6 months,” “I currently engage in regular physical activity,” “I have been regularly physically active for the past 6 months.” Following previous studies (Carnegie et al., 2002), participants’ PA stages of change were defined as precontemplation (inactive and no intention to become physically active in the next 6 months), contemplation (inactive but intend to become more physically active in the next 6 months), preparation (inactive but intend to become more physically active in the next month), action (active but for less than 6 months), and maintenance (active for at least 6 months). Precontemplation was defined as “no intention to become more physically active,” and the remaining statements were defined as “having the intention to become more physically active or staying physically active.”

The PA outcome expectation score, which is a 5-point Likert-type scale (Prodaniuk et al., 2004), was used to measure participants’ belief in the health benefits of PA. Participants were asked to answer the degree to which they agree or disagree with the following questions (1 = strongly agree, 5 = strongly disagree): “PA makes me feel better physically,” “PA makes my mood better generally,” “PA helps me feel less tired,” “PA makes my muscles stronger,” “PA is an activity I enjoy doing,” “PA gives me a sense of personal accomplishment,” “PA makes me more alert mentally,” “PA improves my endurance in performing my daily activities,” “PA helps to strengthen my bones.” We determined the participants’ mean scores (reversed) on these items, so the higher the score, the higher the participants’ belief in the health benefits of PA.

PA self-efficacy, which is also a 5-point Likert-type scale (Prodaniuk et al., 2004), was used to measure participants’ self-efficacy in conducting PA. Participants were asked to answer how confident are they were that they could be physically active in each of the following situations (1 = not at all confident, 2 = slightly confident, 3 = moderately confident, 4 = very confident, 5 = extremely confident): “When I am tired,” “When I am in a bad mood,” “When I feel I do not have time,” “When I am on holiday,” “When it is raining.” We determined the participants’ mean scores on these items, so the higher the score, the higher the participants’ self-efficacy in conducting PA.

Awareness, perceptions, and expectations of the CCG

Awareness about the CCG

First, participants were asked, “Have you seen, read, or heard advertising or messages promoting the CCG as a means of encouraging local people to be more physically active?” Responses to the question included different options for sources of such awareness (supplementary material, Table S1). The response was dichotomized as a binary variable (1 = at least one “Yes” answer, 0 = otherwise). Also, the number of sources of awareness (numbers) was calculated. Second, participants were also asked, “What do you know about the CCG?” Responses to the question included options about the functions of the CCG such as “developing and creating a safe space” and “providing green areas/open spaces” (supplementary material, Table S2). The total amount of positive awareness regarding the functions of CCG (numbers) was calculated.

Positive perceptions about the CCG

Participants were also asked, “What difference (if any) does the Connswater Community Greenway make?” Responses included “healthy lifestyle/fitness/for exercise” and “enjoyment and recreation” (supplementary material, Table S3). The total number of positive evaluations about CCG (numbers) was calculated.

Positive expectations about the CCG

Participants were asked, “To what extent would you agree or disagree with the statement: Using the Connswater Community Greenway … to be more physically active would be enjoyable,” “What extent would you agree or disagree with the statement: Using the Connswater Community Greenway … will help me to be more physically active,” “To what extent would you agree or disagree with the statement: Using the Connswater Community Greenway … will help me engage more with friends,: “What extent would you agree or disagree with the statement: Using the Connswater Community Greenway … will help me feel better.” Responses to the statements included 1 = strongly disagree, 2 = disagree, 3 = neither agree nor disagree, 4 = agree, 5 = strongly agree. The responses were dichotomized as agree (1 = 4 and 5) and disagree (0 = 1, 2, and 3).

CCG visit frequency and duration

We evaluated self-reported CCG visit frequency and duration through two questions. Participants were asked, “How often do you usually visit the parks along the Connswater Community Greenway?” Responses to the statement included 1 = daily, 2 = few times per week, 3 = once per week, 4 = couple times per week, 5 = monthly, 6 = rarely, 7 = never. The response was dichotomized as frequent visits (1 = 1 and 2) and infrequent visits (0 = 3, 4, 5, 6, and 7). They were also asked, “On a typical day when you go to any of the parks, how long do you usually stay?” Responses included 0 = 0 min, 1 = <15 min, 2 = 15–30 min, 3 = 31–60 min, 4 = >1 h but <2 h, 5 = 2–3 h, 6 = >3 h but <5 h, and 7 = 5 h or more. The response was dichotomized as long duration of visits (1 = 3, 4, 5, 6, and 7) and short duration of visits (0 = 1 and 2).

Covariates

Following previous studies (Hunter et al., 2021; Payne et al., 2002; Veitch et al., 2014, 2018; Zanon et al., 2013), we controlled for a series of socioeconomic covariates: sex (female vs. male), age, weekly household income (≤£230 per week vs. >£230 per week), education level (tertiary or equivalent vs. secondary school/none or other), marital status (married vs. others), and accommodation condition (owned outright vs. mortgage/co-ownership or rented). Area-level deprivation was measured by the Northern Ireland Multiple Deprivation Measure (least deprived [4th and 5th quintiles] vs. others [1st, 2nd, and 3rd quintiles]; Ijpelaar et al., 2018). Because dog ownership (yes vs. no) and functional restriction status (“Do you have any long-term illness, health problem, or disability which limits your daily activities or the work you can do? Include problems which are due to old age”: yes vs. no) influence green space visits (Temple et al., 2011), they were also controlled for. Existing studies have pointed out that there may be seasonal variations in people’s use of greenways (Sun et al., 2024), so we also controlled for this. Safety perception is another important determinant of greenway visits (Paneerchelvam et al., 2020), so we used a two-item 5-point Likert scale (supplementary material, Table S4) to measure it and included it in our analysis.

Distance to the CCG was calculated using the distance from the participant’s home address to the nearest CCG access point (e.g., accessible footpath or entrance). The distance was calculated using the footpath network, which was constructed by using the map provided by the Land and Property Services of Northern Ireland. More details can be found in Ellis et al. (2016). The distance from the CCG was then classified as 0–400 m, >400–800 m, >800–1,200 m, and > 1,200 m, because these thresholds represent an approximate 0 to 5-, 5 to 10-, 10 to 15-, and over 15-min walk, respectively, from the participant’s home to the CCG (Hunter et al., 2021). Table 1 summarizes the characteristics of the participants.

Table 1. Descriptive statistics of PARC samples.

Variables	Proportion/mean (SD)
Outcome
Vigorous-intensity recreational PA (MET/week)	263.14 (731.00)
Moderate-intensity recreational PA (MET/week)	335.90 (686.29)
Stage of change PA (%)
Precontemplation	29.48
Contemplation, preparation, maintenance, or action	70.52
PA outcome expectation score (1–5)	4.13 (0.87)
PA self-efficacy (1–5)	2.39 (0.95)
Frequent CCG visits (%)
Yes	17.94
No	82.06
Long duration of CCG visits (%)
Yes	36.10
No	63.90
Predictors
Distance to the CCG (%)
<400 m	16.48
400–800 m	18.61
>800–1,200 m	16.93
>1,200 m	47.98
Positive awareness about the CCG
Have seen, read, or heard advertising or messages promoting the CCG as a means of encouraging local people to be more physically active (%)
Yes	67.94
No	32.06
Number of sources of awareness (numbers)	1.89 (1.83)
Number of positive awareness regarding the functions of CCG (numbers)	1.38 (0.92)
Positive perceptions of the CCG (numbers)
Positive expectations about the CCG
Agree that using the CCG will make PA enjoyable (%)	79.26
Yes	20.74
No
Agree that using the CCG will help people to be more physically active (%)	57.85
Yes	42.15
No
Agree that using the CCG will help people engage more with friends (%)	39.13
Yes	60.87
No
Agree that using the CCG will help people feel better (%)	61.55
Yes	38.45
No	1.54 (1.23)
Individual covariates
Sex (%)
Male	45.40
Female	54.60
Age (years)	50.60 (17.96)
Weekly household income (%)
≤£230 per week	24.10
>£230 per week	75.90
Educational level (%)
Tertiary or equivalent	45.40
Others (secondary school/none or other)	54.60
Marital status (%)
Married	43.83
Other	56.17
Accommodation (%)
Owned outright	32.40
Mortgage/co-ownership or rented	67.60
Area-level deprivation (%)
Least deprived (5th quintile)	45.40
Others (1st, 2nd, 3rd, and 4th quintiles)	54.60
Dog ownership (%)
Yes	24.78
No	75.22
Functional restriction (%)
Yes	28.48
No	71.52
Safety perception (1–5)	3.72 (0.71)
Season (%)
Spring	26.57
Summer	26.01
Autumn	24.89
Winter	22.53

Analysis

To identify the association between awareness, perceptions, and expectations of the CCG; CCG visit frequency and duration; and recreational PA, we used multilevel linear and logistic regressions (Tom et al., 1999). The model included a random effect to account for the clustering of individuals within super output areas (SOAs, secondary geographical units for collecting primary census information). Variance inflation factors (VIF < 3) suggested no multicollinearity among the predictors. First, we estimated the effect of awareness, perceptions, and expectations of the CCG on CCG visit frequency and duration using multilevel logistic regression models (Models 1 and 2). Second, we estimated the awareness, perceptions, and expectations of the CCG and CCG visit frequency and duration on moderate- and vigorous-intensity recreational PAs using multilevel linear regression models (Models 3 and 4). Third, we estimated the awareness, perceptions, and expectations of the CCG and CCG visit frequency and duration on PA stages of change using multilevel logistic regression models (Model 4). Last, we estimated the awareness, perceptions, and expectations of the CCG and CCG visit frequency and duration on PA outcome expectation scores and PA self-efficacy using multilevel linear regression models (Models 5 and 6). Stata v15.1 (StataCorp LLC, College Station, TX) was used for statistical analyses.

Results

Multilevel regression results

Associations of awareness, perceptions, and expectations of the CCG and CCG frequency and duration of visits

Model 1 in Table 2 shows the associations between the awareness, perceptions, and expectations of the CCG and the odds of frequent visits. The results indicate that respondents living farther from the CCG were less likely to visit frequently (>800 m–1,200 m: odds ratio [OR] = 0.26, 95% CI [0.13–0.54]; >1,200 m: OR = 0.17, 95% CI [0.09–0.32]). Also, positive perceptions about the CCG (OR = 1.29, 95% CI [1.07–1.55]) were positively associated with the odds of frequent CCG visits. Respondents who agreed that using the CCG would help people engage more with friends were more likely to visit frequently (OR = 2.27, 95% CI [1.39–3.71]), and a similar finding was reported for those who agreed that using the CCG would help people feel better (OR = 2.94, 95% CI [1.30–6.64]).

Table 2. Multilevel model results of associations of information, perceptions, and expectations and perceptions of the CCG and frequency and duration of CCG visits.

	Model 1 (DV: Frequent CCG visits)	Model 2 (DV: Long duration of CCG visits)
	OR (95% CI)	OR (95% CI)
Fixed effects
Distance to the CCG
400–800 m	0.56* (0.30–1.02)	0.73 (0.42–1.26)
>800–1,200 m	0.26*** (0.13–0.54)	0.74 (0.42–1.30)
>1,200 m	0.17*** (0.09–0.32)	0.62* (0.37–1.04)
Positive information about the CCG
Have seen, read, or heard advertising or messages promoting the CCG as a means of encouraging local people to be more physically active
Yes (ref: No)	1.17 (0.59–2.35)	1.75** (1.05–2.91)
Number of sources of information	1.15* (0.99–1.33)	1.14** (1.01–1.29)
Number of positive information regarding the functions of CCG	1.03 (0.81–1.33)	0.93 (0.75–1.14)
Positive perceptions of the CCG	1.29*** (1.07–1.55)	1.24*** (1.07–1.44)
Positive expectations about the CCG
Agree that using the CCG will make PA enjoyable
Yes (ref: No)	0.42* (0.18–1.05)	1.03 (0.59–1.78)
Agree that using the CCG will help people to be more physically active
Yes (ref: No)	1.88* (0.92–3.86)	1.81** (1.09–3.03)
Agree that using the CCG will help people engage more with friends
Yes (ref: No)	2.27*** (1.39–3.71)	1.21 (0.82–1.77)
Agree that using the CCG will help people feel better
Yes (ref: No)	2.94*** (1.30–6.64)	1.72** (1.02–2.91)
Random effects
Var (SOAs)	0.25***	0.12***
Log-likelihood	−314.61	−479.95
AIC	681.22	1,011.91

Note. Models adjusted for covariates. DV = dependent variable; AIC = Akaike information criterion.

*p < .10. **p < .05. ***p < .01.

Model 2 in Table 2 shows the associations between awareness, perceptions, and expectations of the CCG and the odds of having longer duration of CCG visits. The results indicate that respondents who had seen, read, or heard advertising or messages promoting the CCG as a means of encouraging local people to be more physically active were more likely to have longer duration of visits to the CCG (OR = 1.75, 95% CI [1.05–2.91]). Also, the number of sources of awareness about the CCG (OR = 1.14, 95% CI [1.01–1.29]) and positive perceptions about the CCG (OR = 1.24, 95% CI [1.07–1.44]) were positively associated with the odds of having longer duration of visits to the CCG. Participants who agreed that using the CCG would help people to be more physically active were more likely to have a longer duration of CCG visits (OR = 1.81, 95% CI [1.09–3.03]), and a similar finding was reported for those who agreed that using the CCG would help people feel better (OR = 1.72, 95% CI [1.02–2.91]).

Associations of awareness, perceptions, and expectations of the CCG; CCG visit frequency and duration; and PA

Model 3 in Table 3 shows the associations between the awareness, perceptions, and expectations of the CCG; CCG visit frequency and duration; and moderate recreational PA. The results indicate that participants living farther from the CCG had more moderate recreational PA behavior (>1,200 m: Coef. = 0.68, SE = 0.30). Also, positive perceptions of the CCG were negatively associated with moderate-intensity recreational PA behavior (Coef. = −0.20, SE = 0.09). In contrast, the number of sources of awareness was positively associated with moderate recreational PA behavior (Coef. = 0.21, SE = 0.07). Hence, participants who engaged in frequent CCG visits (Coef. = 0.76, SE = 0.29) or longer duration of CCG visits (Coef. = 0.85, SE = 0.23) had higher levels of moderate-intensity recreational PA.

Table 3. Multilevel model results of associations of information, perceptions, and expectations of the CCG; CCG visit frequency and duration; and PA.

	Model 3 (DV: Moderate PA in log form)	Model 4 (DV: Vigorous PA in log form)	Model 5 (DV: Stage of change PA)	Model 6 (DV: PA outcome expectation score)	Model 7 (DV: PA self-efficacy)
	Coef. (SE)	Coef. (SE)	OR (95% CI)	Coef. (SE)	Coef. (SE)
Fixed effects
Distance to the CCG
400–800 m	−0.12 (0.33)	−0.42 (0.29)	1.07 (0.53–2.16)	−0.04 (0.09)	−0.15 (0.09)
>800–1,200 m	0.61* (0.34)	−0.23 (0.30)	0.73 (0.36–1.47)	0.03 (0.09)	−0.08 (0.10)
>1,200 m	0.68** (0.30)	−0.16 (0.27)	0.47** (0.24–0.92)	−0.00 (0.09)	−0.15 (0.09)
Positive information about the CCG
Have seen, read, or heard advertising or messages promoting the CCG as a means of encouraging local people to be more physically active
Yes (ref: No)	−0.58* (0.30)	0.46* (0.26)	2.33*** (1.25–4.31)	0.17** (0.07)	0.02 (0.08)
Number of sources of information	0.21*** (0.08)	−0.08 (0.07)	0.84** (0.71–0.99)	0.00 (0.02)	0.06** (0.02)
Number of positive information regarding the functions of CCG	−0.03 (0.12)	−0.16 (0.10)	1.02 (0.80–1.30)	−0.06** (0.03)	−0.07** (0.03)
Positive perceptions of the CCG	−0.20** (0.09)	0.04 (0.08)	1.05 (0.87–1.25)	0.02 (0.02)	−008*** (0.02)
Positive expectations about the CCG
Agree that using the CCG will make PA enjoyable
Yes (ref: No)	0.49* (0.29)	0.55** (0.25)	0.78 (0.46–1.33)	0.41*** (0.07)	0.19** (0.08)
Agree that using the CCG will help people to be more physically active
Yes (ref: No)	0.53* (0.32)	0.36 (0.27)	0.62 (0.34–1.16)	0.11 (0.08)	0.19** (0.09)
Agree that using the CCG will help people engage more with friends
Yes (ref: No)	−0.32 (0.26)	−0.23 (0.22)	1.07 (0.53–2.16)	0.10 (0.06)	0.06 (0.07)
Agree that using the CCG will help people feel better
Yes (ref: No)	0.38 (0.32)	−0.68*** (0.27)	0.41*** (0.23–0.75)	−0.06 (0.08)	−0.01 (0.09)
Frequent CCG visits
Yes (ref: No)	0.76*** (0.29)	0.54** (0.25)	0.92 (0.48–1.76)	0.20*** (0.07)	0.16** (0.08)
Long duration of CCG visits
Yes (ref: No)	0.85*** (0.23)	−0.18 (0.20)	0.32*** (0.20–0.54)	0.09 (0.06)	−0.03 (0.06)
Random effects
Var (SOAs)	0.04***	0.04***	0.11***	0.02***	0.01***
Var (Individuals)	8.07***	5.91***		0.49***	0.63***
Log-likelihood	−2,196.91	−2,061.16	−335.79	−959.20	−1,065.97
AIC	4,451.834	4,180.331	727.59	1,976.40	2,189.95

Note. Models adjusted for covariates. DV = dependent variable; AIC = Akaike information criterion.

*p < .10. **p < .05. ***p < .01.

Model 4 in Table 3 shows the associations between awareness, perceptions, and expectations of the CCG; CCG visit frequency and duration; and vigorous-intensity recreational PA. The results indicate that respondents who agreed that using the CCG would help people feel better had less vigorous recreational PA (Coef. = −0.68, SE = 0.27). In contrast, respondents with frequent CCG visits had more vigorous recreational PA (Coef. = 0.54, SE = 0.25). Similar findings were observed among those who agreed that using the CCG would make PA enjoyable (Coef. = 0.55, SE = 0.25).

Model 5 in Table 3 shows the associations between the awareness, perceptions, and expectations of the CCG; CCG visit frequency and duration; and the odds of staying at the PA stage of precontemplation. The results indicate that respondents living farther from the CCG were less likely to stay at the PA stage of precontemplation (>1,200 m: OR = 0.47, 95% CI [0.24–0.92]). Also, the number of sources of awareness was negatively associated with the odds of staying at the PA stage of precontemplation (OR = 0.84, 95% CI [0.71–0.99]). Respondents who had seen, read, or heard advertising or messages promoting the CCG as a means of encouraging local people to be more physically active were more likely to stay at the PA stage of precontemplation (OR = 2.33, 95% CI [1.25–4.31]). In contrast, respondents with longer duration CCG visits were less likely to stay at the PA stage of precontemplation (OR = 0.32, 95% CI [0.20–0.54]). Similar findings were observed among those who agreed that using the CCG would help people feel better (OR = 0.41, 95% CI [0.23–0.75]).

Model 6 in Table 3 shows the associations between the awareness, perceptions, and expectations of the CCG; CCG visit frequency and duration; and PA outcome expectation score. The results indicate that participants who had seen, read, or heard advertising or messages promoting the CCG as a means of encouraging local people to be more physically active had higher PA outcome expectation scores (Coef. = 0.17, SE = 0.07). The amount of positive awareness regarding the functions of CCG was negatively associated with the PA outcome expectation score (Coef. = −0.06, SE = 0.03). Also, participants who agreed that using the CCG would make PA enjoyable (Coef. = 0.41, SE = 0.07) or who had frequent CCG visits (Coef. = 0.20, SE = 0.07) had a higher PA outcome expectation score.

Model 7 in Table 3 shows the associations between the awareness, perceptions, and expectations of the CCG; CCG visit frequency and duration; and PA self-efficacy. The results indicate that the number of sources of awareness of the CCG (Coef. = 0.06, SE = 0.02) was positively associated with PA self-efficacy, whereas positive awareness regarding the functions of the CCG (Coef. = −0.07, SE = 0.03) and positive perceptions of the CCG (Coef. = −0.08, SE = 0.02) were negatively associated with PA self-efficacy. Also, participants who agreed that using the CCG would help people to be more physically active (Coef. = 0.19, SE = 0.09) or who agreed that using the CCG would make PA enjoyable (Coef. = 0.19, SE = 0.08) had higher PA self-efficacy. Participants who had frequent CCG visits (Coef. = 0.16, SE = 0.08) had higher PA self-efficacy.

Discussion

This study extended previous research regarding how awareness, perceptions, and expectations of a new urban greenway influence urban greenway visits and recreational PA in several respects. First, it made a methodological contribution to the study of the effect of a major urban regeneration project by proposing several indicators for measuring residents’ awareness, perceptions, and expectations of the greenway. Second, it further developed a conceptual framework for understanding the association among awareness, perceptions, and expectations of a major urban regeneration project for the urban greenway, residents’ visits to it, and recreational PA.

Our findings indicated that positive expectations and perceptions of the CCG were generally associated with more frequent and longer durations of CCG visits, whereas such associations were not found for positive awareness about the CCG. First, we were not able to measure all kinds of awareness activities, and people are more likely to be influenced by awareness of specific interest to them (Snyder & Hamilton, 2002). Hence, awareness about the CCG in this study may not include all information that is of residents’ interest, which can explain its lack of association with CCG visits. Also, the results suggested that the number of sources of awareness about the CCG was positively associated with frequent and longer CCG visits. Existing evidence indicates that a greater number of sources of awareness exposure can increase awareness more believable and enhance credibility, thus influencing people’s behavior (Johnson & Kaye, 2013). Second, positive perceptions of an urban greenway measure how people perceive the changes postintervention (Masterton et al., 2020). Therefore, positive perceptions of the CCG can reflect how residents perceive positive changes in the CCG, which may make the CCG a more pleasant site to visit and encourage longer stays. This perception was self-reported by residents, so the higher the number of positive perceptions, the more changes that are of interest. These positive perceived changes in the CCG are especially important for residents’ visits because existing studies indicated that people pay more attention to the changes they care about in a certain regeneration project (Cheng & Chou, 2015). Third, expectations of the CCG reflect how residents believe visiting the CCG can benefit their well-being. The theory of planned behavior indicates that when people perceive the benefits of a certain kind of behavior, they are more willing and likely to engage in such behavior more frequently (Fishbein & Ajzen, 1977). Therefore, residents who agree that visiting the CCG can help them be more physically active, engage more with friends, or feel better tend to visit the CCG more frequently.

We also found that awareness, perceptions, and expectations of the CCG were associated with recreational PA and residents’ intention to stay inactive. First, when people receive more sources of awareness, this usually indicates a longer duration of exposure to certain information, which may make it more credible (Hay et al., 2009) and thus able to influence PA behavior. However, we found that the amount of positive awareness regarding the functions of the CCG was negatively associated with recreational PA. For example, people who had seen, read, or heard advertising or messages promoting the CCG as a means of encouraging local people to be more physically active were more willing to stay inactive. This may be because physically inactive people are more likely to pay more attention to the information and awareness regarding the health benefits of outdoor recreational PA, because they may need such information and awareness to motivate them to be active (Kahn et al., 2002). Second, positive perceptions about the CCG were surprisingly associated with less recreational PA. There are some potential explanations for such a finding. It is possible that people’s ideal PA environment varies by individual (Kirtland et al., 2003; Smith et al., 2015), and changes in CCG may not be related to any of the elements that matter for residents’ ideal recreational PA environment. Therefore, residents may not adapt to too many changes in the CCG and thus reduce the amount of PA time spent in the CCG. Also, people with more PA in the CCG may perceive fewer changes, because they spent more time in the CCG throughout intervention and witnessed the changes over time, whereas people with less PA in the CCG may perceive a more significant difference. Third, most of the positive expectations of the CCG were associated with more recreational PA. Previous studies indicated that belief can strengthen and encourage people’s actual behavior (Fishbein & Ajzen, 1977). Therefore, residents who agreed that visiting CCG is enjoyable are more likely to take PA in it. Also, one of the purposes of PA is social contact (Bertera, 2003), so residents who agree that visiting the CCG can help them engage more with friends are more likely to choose the CCG as a PA site. However, residents who agreed that using the CCG would help people feel better had less recreational PA. A possible explanation is that such a perception may strengthen their belief regarding the benefits of taking PA in the CCG, so they may think a shorter duration of PA in the CCG will provide the same amount of health benefits as elsewhere, thereby reducing it.

The association between CCG visits (including both frequency and duration) and recreational PA was positive. There are several explanations for such a finding. First, frequent and longer duration of CCG visits after the regeneration project may increase people’s nature connectedness, which will promote their willingness to visit the site (Martin et al., 2020). Also, the environment of the CCG was significantly improved after the regeneration project, so the frequent and longer duration of CCG visits may help residents notice the changes and adapt to them. Second, existing evidence has highlighted the beneficial effect of taking PA in a natural environment (e.g., greenways and freshwater; Stults-Kolehmainen & Sinha, 2014). Therefore, frequent CCG visits may improve people’s PA performance and provide more motivation to persist in PA. Third, when visiting the CCG, residents are more likely to view others taking recreational PA. Existing literature suggests that viewing others taking PA can encourage one’s own PA behavior, because it may enhance PA as a social norm, thus strengthening one’s belief that taking PA is beneficial (J. J. Lee et al., 2020). The regeneration project also aims at organizing more PA groups to encourage PA, so when visiting the CCG residents are more likely to join different PA groups, which may help them take collective PA such as walking, running, or cycling in a group (Burke et al., 2006). Last, because the regeneration project on the CCG aims at facilitating recreational PA, there is a lot of advertising information in the CCG regarding the health benefits of taking PA. With frequent and longer duration visits to the CCG, residents are more likely to be exposed to such information, which increases their willingness to be active (Kahn et al., 2002).

Policy implications

Our findings have important implications for promoting healthy communities by implementing urban regeneration projects in urban green and blue spaces. First, because more positive information and awareness about the CCG is positively associated with visits to the CCG, policymakers should consider using a variety of media to advertise the CCG. Also, the content of the advertisement should focus on introducing the direct health benefits of visiting the CCG, not just the promotion of physical activity. Second, to increase people’s positive expectations of the health benefits of visiting the CCG, more regular visitors should be invited to share their positive comments through social media. Third, people’s perceptions about the CCG also matter, so policymakers should provide more ways for visitors to comment on their experience in CCG, which may help planners and designers to better understand visitors’ needs and wants. Fourth, the accessibility of the CCG also plays a role in attracting visitors, so more entrance points and public transportation stops can be planned near the CCG. Last, our study provides some evidence that frequent and longer duration CCG visits are positively associated with PA, so more efforts should be taken to attract new visitors. For instance, policymakers may consider providing gift coupons to visitors who share their location in the CCG a certain number of times through social media.

Limitations

This study has several limitations. First, it was based on a cross-sectional study design, so we cannot infer the causality among awareness, expectations, and perceptions of the CCG; CCG visits; and PA. Future studies should use a longitudinal study design, which may include information on an urban greenway pre- and postimplementation to calculate the changes in the use of the greenway as well as the difference in PA, which may improve the robustness of our estimation. Also, our results may be influenced by some potential self-selection bias (Boone-Heinonen et al., 2010). For example, we did not measure people’s nature connectedness, but this factor influences both nature visits and perceptions of nature (Martin et al., 2020). Second, CCG visits are self-reported, which may be influenced by measurement bias. Future studies could use more objective measures such as Global Positioning System in a subsample to better evaluate people’s actual exposure to the CCG. Third, participants’ recreational PA behavior was self-reported, which may lead to bias. However, the GPAW has been validated for measuring the change of PA in Northern Ireland. Fourth, CCG visits may have only a short-term effect on PA (Vert et al., 2020), but because our study measured PA behavior 6 months postimplementation, such an association was not explored. Fifth, an important variable that is missing from the model is mode of transportation. We did not include such a variable, because we included a variable reflecting the walking distance to the CCG and most participants were sampled within a 1-mile radius of the CCG (a walkable distance). Last, this study did not test whether there were mediating or moderating effects among variables, and future studies should consider such associations (Wang et al., 2024). For example, people’s use of green space may mediate the association between perceptions of an urban greenway and recreational PA.

Conclusion

This study of the CCG in Belfast, Ireland, found that positive awareness, perceptions, and expectations of the CCG were positively associated with CCG visits and recreational PA. Frequent and longer CCG visits were also associated with recreational PA. These findings showcase the importance of advertising and education regarding the health benefits of access to natural environments when implementing urban greenways.

Acknowledgments

The study was approved by the Office for Research Ethics Committees, Northern Ireland (09/NIR02/66). The authors thank the co-investigators of the PARC Study: Mark A Tully, Helen McAneney, Margaret E Cupples, Michael Donnelly, Geraint Ellis, George Hutchinson, Alberto Longo, Lindsay Prior, and Michael Stevenson. The authors also 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.

Disclosure statement

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article.

Supplemental Material

Supplemental data for this article can be found on the publisher's website.

Additional information

Funding

This work was undertaken by the GroundsWell consortium supported by funding from the UK Prevention Research Partnership (MR/V049704/1), which is funded by the British Heart Foundation, Cancer Research UK, Chief Scientist Office of the Scottish Government Health and Social Care Directorates, Engineering and Physical Sciences Research Council, Economic and Social Research Council, Health and Social Care Research and Development Division (Welsh Government), Medical Research Council, National Institute for Health Research, Natural Environment Research Council, Public Health Agency (Northern Ireland), The Health Foundation, and Wellcome. The work was also supported by the HSC Research and Development Office Northern Ireland (COM/5634/20). The PARC study was supported by a grant from the National Prevention Research Initiative (NPRI; Grant No. 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 Center of Excellence for Public Health Northern Ireland. The authors declare that the funders had no role in any aspect of the study, including the study design; the collection, analysis, and interpretation of data; in the writing of the article; or in the decision to submit the article for publication.