Table 1 Summary of papers included that report original research
AUTHORS | SAMPLE SIZE | STUDY DESIGN | EXPERIENCE OF THE NATURAL ENVIRONMENT (NE) | TYPE OF PHYSICAL ACTIVITY (PA) | SUMMARY OF FINDINGS |
|---|---|---|---|---|---|
Akers et al. (2012) [21] | 14 young male adults | Experimental trial, with pre-test measurements and within-subjects, non-counterbalanced design. No control group. | Exposure to colour-manipulated videography of NEs. Standardized setting | Cycling. Standardized activity | The colour green was associated with greater improvements of post-exercise total mood and lower ratings of perceived exertion while exercising. |
Anderson et al. (2008) [91] | 446 adults, ‘proximate/resident’ and ‘distant/non-resident’ visitors | Cross-sectional study (random sampling) and onsite survey (purposeful sampling). | Visitation of naturalistic parks (lake area). Standardized setting | Outdoor recreation such as fishing, camping, and motor boating. Self-reported | Important benefit factors to NEs visitors were ‘enjoy nature’, ‘mental and physical health’, and ‘social interaction’. For people living nearby a NE, ‘solitude’ and ‘learning’ were also important benefit factors. |
Aspinall et al. (2013) [25] | 12 adults | Experimental, within-subjects, non-counterbalanced design. No control group. | Experience of PA in an urban green-area (park). Standardized setting | Walking. Standardized activity | Measurements of brain-waves activity by a mobile EEG showed stress recovery mechanisms in line with Attention-Restoration Theory (Kaplan & Kaplan, 1989). |
Bai et al. (2013) [65] | 893 adults living nearby parks | Cross-sectional study among residents within .5 miles from 60 parks | Availability and perceived quality of urban parks. Self-reported | Overall moderate to vigorous PA and park-based PA. Self-reported | There was a strong agreement among residents that having neighbourhood parks is a benefit. Perceived quality of parks was positively associated with overall and park-based PA. |
Barton & Pretty (2010) [48] | 1,252 adults overall | Meta-analysis of 10 studies with matching measurements. Individual studies used pre-measurements with non-randomized allocation (self-selecting). | Experience of PA in different NEs. Standardized setting | Various green exercise activities. Standardized activity | A dose–response effect of green-exercise on mood and self-esteem. All types of NEs elicited greater post-exercise improvements, with different patterns identified for subjects’ age, time of exposure and PA intensity. |
Berman et al., 2008 [52] | 37 young adults (study 1) | Experimental trial using mixed design with counterbalanced cross-over. Random allocation. | Experience of PA in an arboretum/park. | Walking. Standardized activity | When compare to walking in an urban setting, walking in a NE after a mental-fatiguing-task was associated with improved performance in an attention task, which was not driven by changes in mood. |
Bjork et al. (2008) [92] | 24,819 adults | Cross-sectional study using stratified random sampling design. | Availability of NEs with high recreational and restorative value in suburban and rural areas. Objective measure | Genial PA. Self-reported | Residential proximity to NEs was associated with neighbour satisfaction, time spent on PA and normal or low BMI. A positive effect on vitality in women was also found. |
Bodin & Hartig (2003) [44] | 12 experienced runners | Experimental trial with pre-test measurements and counterbalanced cross-over design (two treatments). Random allocation (method not described) | Exposure to a NE (natural reserve). Standardized setting | Running. Standardized activity | PA in the natural environment was associated with greater potential for restoration, while statistically significance was not achieved for affective responses and attention performance. The subjects reported to prefer running in the natural environment. |
Boone-Heinonen et al. (2010) [83] | 10,773 adolescents | Cross-sectional study (sampling technique reported elsewhere). | Amount of green spaces and availability of parks in urban areas.. Objective measure | Wheel-based activities, active sports and exercise. Self-reported | Associations between availability of NEs and PA were found, especially in girls. |
Butryn & Furts, 2003 [50] | 30 female experienced runners | Experimental trial using mixed design with counterbalanced cross-over. Random allocation not specified. | Experience of PA in an urban park | Running. Standardized activity | There were no differences in affective responses after running in the two environments, despite most of runners reported to prefer the NE. Safety issues may influence the experience of PA in the NE and consequent affective responses. |
Cerin et al. (2008) [56] | 2,650 adults | Cross-sectional study using two-stage stratified sampling design. | Availability of different types of NEs in urban areas, including beaches/coasts trails and open spaces. Self-reported | Leisure time physical activity. Self-reported | PA was positively associated with perceived access to beaches/rivers/lakes, while significance was not achieved for open spaces such as parks and trails. Self-efficacy did not mediate the relationship. |
Cerin et al. (2013) [75] | 484 older adults | Cross-sectional using stratified random sampling design. | Presence of nature elements within the neighbourhood in urban area. Self-reported | Leisure time walking and other PA Self-reported | Availability of NEs within the neighbourhood was associated with PA. Other attributes of the environment such as noise/air pollution, availability of spaces for walking and safety, were also predictors of PA. |
Cohen et al. (2007) [76] | 1,318 adults | Direct observation and onsite interviews. | Availability of urban parks. Objective measure (onsite observation) | Park-based PA. Assessed through direct observation | The NE was an important source for PA opportunity for the users. Furthermore residential proximity to NEs was associated with frequency visitation the NE and PA. Men used the NE for PA purposes more often than women. |
Coogan et al. (2009) [93] | 20,354 Afro-American women | Prospective cohort study. Recruitment based on subscription to a magazine. | Availability of urban parks. Objective measure | Utilitarian- and, exercise-walking. Self-reported | A weak association was found between distance to NEs and PA. Other factors associated with PA related to housing density, bus availability and access to transit. |
Coombes et al. (2010) [94] | 6,821 adults | Cross-sectional study using single stage sampling based on electoral wards (equal size populations selected). | Availability of urban green-spaces. Objective measure | General PA. Self-reported | Residential proximity to NEs, especially those classified as ‘formal green spaces’, was associated with higher PA and lower probability to be overweight or obese. |
Coutts et al. (2013) [95] | 67 counties | Cross-sectional study. Sampling technique not reported. | Availability of green-spaces within counties. Objective measure | Moderate to vigorous PA. Self-reported | Overall amount of NE within the county was positively associated with PA. |
Cummins & Fagg (2012) [96] | 79,136 adults | Cross-sectional study over two time-periods. Sampling technique not reported. | Availability of nature elements within the neighbourhood in urban and rural. Objective measure | General PA. Self-reported | Mixed results were found in the relationship between obesity/overweight and residential proximity to NEs, with PA not mediating the relationship. Living in urban or rural areas was instead a relevant factor. |
Day (2008) [90] | 45 older adults | Qualitative, case study. Purposeful snowball sampling design. | Availability of nature elements (e.g. presence of gardens, sea view, trees, etc.) within the living environment. Self-reported and direct observation | Walking and general PA habits. Self-reported | NE was an important motivational factor to engage in PA for pleasure. Cleanliness, aesthetic, and restorative potential were environmental characteristics encouraging older people to go outdoors, for walking and socializing. Among the physical barriers, quality of the pavement was the most important. |
Duvall & Young (2013) [89] | 62 experienced walkers | Observational study using a purposeful sample. Subjects recruitment through announcement. | Use of nearby nature. Self-reported | Walking. Self-reported | To set health goals and using good walking paths, especially using nearby nature, were the most useful strategies to sustain walking routines among experienced walkers. Social support was the least useful strategy, although it was associated with other types of PA. |
Fan et al. (2011) [23] | 1,544 adults | Cross-sectional study (three-stage random sampling technique) and definition of a conceptual model. | Availability of vegetation and parks within the neighbourhood. Objective measure | Moderate to vigorous PA. Self-reported | PA was directly associated with availability of NEs and indirectly associated with reduced stress via improved social support. NE was positively associated with reduced stress, although different components of neighbourhood green had different influences on stress mitigation. |
Foster el al. (2004) [77] | 4,157 adults | Cross-sectional study using multi-cluster random sampling design. | Availability of parks and open spaces within the neighbourhood (urban and rural areas). Self-reported | Walking. Self-reported | The availability of NEs was associated with PA in men, while walking behaviour in women was more influenced by perceived safety. |
Foster el al. (2009) [97] | 13,927 middle-aged and older adults | Cross-sectional study. Recruitment through approach by general practitioner. | Availability of urban green-spaces (nature reserve, river walk, or public park). Objective measure | Different types of PA, for leisure or transportation purposes. Self-reported | There was no association between availability of NEs and PA. Traffic intensity had a negative impact on cycling. |
Gatersleben & Andrews, 2013 [60] | 34 adults (study 2) | Experimental trial with counterbalanced cross-over design and control group (indoors with no PA). Random allocation not specified. | Exposure to actual and ‘virtual' NEs with different characteristics. | Walking. Standardized activity | Compared to a NE with low prospect/accessibility and high refuge characteristics, a NE with high prospect/accessibility and low refuge were associated with greater improvement of mood,. On the other hand low prospect/high refuge certain types of NEs may be responsible for more stress due to increased fear. |
Gomez et al. (2010) [78] | 1,966 older adults | Multi-level cross-sectional study using a two-stage randomized sampling design. | Availability of urban parks within the neighbourhood. Objective measure | Walking Self-reported | Associations between availability of NEs and PA were found. Other factors influencing PA behaviour were safety from traffic, pavement slope and connectivity. |
Gomez et al. (2010) [84] | 1,315 adults | Multi-level cross-sectional study using stratified and cluster sampling design. | Availability of urban parks within the neighbourhood. Objective measure | Leisure time PA. Self-reported | The availability of NEs predicted irregular and regular PA. Other environmental factors predicting PA were better connectivity and smaller pavement slope. |
Harte & Eifert (1995) [40] | 10 male adults | Experimental trial with counterbalanced cross-over design (three exercise-treatments and a control). Randomized allocation (method not described) | Garden of a University Campus and indoor reproduction of outdoors sounds. Standardized setting | Running. Standardized activity | PA in an outdoor NE was associated with an external focus of attention, greater likability, improved mood and lower production of stress hormones as compared with PA indoors. |
Hartig et al. (1991) [42] | 102 adults | Two studies: Quasi-experimental with pre-test measurements on three groups (one control). Experimental trial with pre-test measurements and between-subjects design (two treatments and a control). Randomized allocation (method not described) | Experiences in wilderness environment and a natural reserve. Spontaneously chosen environment and standardized setting | Vacation experiences and walking. Self-reported and standardized activity | Experiences in NEs were associated with greater restorative effects, overall happiness and improved cognitive performances as compared to the other experiences. Some effects on positive affect were also found. No effects were found on physiological indicators of stress (blood pressure and heart rate) |
Hartig et al. (2003) [43] | 112 students | Experimental trial with between-subjects design (two treatments). Random allocation stratified by gender (method not described). | Natural reserve adjacent to a forest. Standardized setting | Passive contemplation and walking. Standardized activity | As compared to PA in a urban setting, PA in a NE elicited a reduction of the blood pressure, an improvement of attention and cognitive performance, and greater positive psychological responses (positive affect and reduced anger). |
Hillsdon et al. (2006) [98] | 4,950 middle-aged adults | Cross-sectional study. Recruitment through approach by general practitioner. | Availability of urban green-spaces. Objective measure | General PA. Self-reported | No statistical evidence of a relationship between availability of NEs and PA was found. |
Hoehner et al. (2010) [64] | 7 parks implementing interventions targeting different audiences (general park visitors, tourists, employees from nearby businesses and youths) | Synthesis of findings from seven quasi-experimental cases, each of which used pre- measurements. | National and urban parks. Standardized settings | Park-based activities, such as walking, hiking, biking on trails and kayaking. Self-reported, onsite observation and existing interventions | The interventions showed some effects on awareness of PA benefits on health and in encouraging ‘active’ use of NEs. Parks offer important sources of PA, and relatively simple and low-cost interventions can efficiently promote PA. |
Hug et al. (2009) [59] | 319 members of fitness centres | Onsite survey (subjects approached for interview) | Outdoor NE with features for PA and exercise. Standardized setting | Gym-based PA. Self-reported | The NE was reported to provide greater potential for restoration than the indoor-exercise setting. Different restorative qualities predicted exercise frequencies in the different environments, with compatibility predicting frequency of PA in the NE. The use the NE was influenced by season and weather. |
Kaczynski & Mowen (2011) [99] | 585 adults | Cross-sectional study using random sampling design (respondents selected from a property list). | Availability of urban parks. Objective measure (method not described) | Park based PA. Self-reported | Access to NEs was positively associated with PA. The association appear to not be determined by a self-selection phenomenon. E.g. people giving greater importance to NEs did not necessarily reside in areas with greater access to it. |
Kaczynski et al. (2008) [85] | 380 adults | Cross-sectional study. Sampling through systematic selection of household in four neighbourhoods. | Availability and quality of urban parks. Objective measure | Park-based PA. Self-reported | Number of features and, to a lesser extent, size and residential proximity were significantly associated with use of a NE for PA purposes. Paved trials, unpaved trials and wooded areas were the stronger predictors of PA. Natural parks interconnected by trails may be effective for PA promotion. |
Kaczynski et al. (2009) [79] | 384 adults | Cross-sectional study. Sampling through systematic selection of household in four neighbourhoods. | Availability of urban parks within the neighbourhood. Objective measure | Moderate to vigorous PA (general, neighbourhood- and park-based). Self-reported | Availability of NEs predicted the residents meeting minimum recommended PA levels, both general and park-based. Women, younger and older individuals were more likely to meet minimum recommended PA when NEs were available around home. |
Karusisi et al. (2012) [63] | 7,290 adults | Cross-sectional study. Recruitment during health check-up (without a priori sampling). | Availability of urban green and open spaces within the neighbourhood. Objective measure | Jogging. Self-reported | Availability of NEs predicted PA behaviour (frequency and location), with neighbourhood experiences and attitudes towards health being only a modest mediator. Other factors predicting PA behaviour were socio-economic status, perceived neighbourhood’s social cohesion, neighbourhood-related stress and having friends. |
Kerr et al. (2006) [41] | 44 young adults | Two experimental trials with counterbalanced cross-over design (two treatments). Random allocation not specified. | Garden of a University Campus. Standardized setting | Running. Standardized activity | PA in a NE vs. PA indoors showed different effects on mood, with different patterns for competitive or recreational runners. The former reported greater rating of tension and effort, while the latter yield greater ‘pride’. |
King et al. (2012) [100] | 2305 adults | Multi-level cross-sectional study. Cluster and stratified random sampling design was used. | Availability of urban parks both referring to smaller and larger parks (type of park not assessed). Objective measure | Walking. Self-reported | Total NE area and proximity to NEs were respectively not associated and negatively associated with PA. |
Kouthouris & Spontis (2005) [32] | 329 young adults | Observational and intervention (invitation to participate in an outdoor program). Convenience sample with recruitment through announcements. | Different setting for outdoors recreations. Standardized setting | Outdoor recreation program, including lake canoe/kayak, orienteering, and archery. Existing intervention | Theory of planned behaviour well predicted PA behaviour in the NE. PA was predicted by intention, which was in turn predicted by attitudes and perceived behavioural control. Subjective norm did not weight in the applied model, while an effect was observed for the added variable ‘role identity’. |
Lackey & Kaczynski (2009) [69] | 574 adults | Cross-sectional study using random sampling design (respondents selected from a property list. | Availability of urban parks. Objective and perceived measure | Park- and neighbourhood-based PA. Self-reported | A poor match was found between perceived and objective distance to NEs. Some associations between proximity to NEs and PA were found. The match between perceived and objective proximity to parks was a predictor of park-based PA, with self-efficacy not being a predictor of the match. |
Lee & Moudon (2008) [101] | 608 adults | Cross-sectional study using spatial randomized sampling design. | Presence of nature elements and views within the neighbourhood (urban area). Objective measure | Walking, cycling, moderate and vigorous PA. Self-reported | Presence of NEs was more commonly reported by sufficiently active individuals. Factors such as poor lighting, distance to destinations, hilly terrain, traffic and dangerous crossing conditions were barriers to PA. |
Li (2010) [22] | 49 adults overall | Monographic review of a series of studies. The individual studies were experimental trials with non-counterbalanced cross-over design (two treatments). | Experiences of Shinrinyoku (forest bathing trips). | Walking in forest environment. | As compared with trips in urban settings, Shinrinyoku was associated with reduced stress hormones and improved immune resources, possibly via reduced allostatic load. Improvements in mood were also reported. |
Li et al. (2011) [61] | 16 male adults | Experimental trial with pre-test measurements and non- counterbalanced cross-over design (two treatments). | Forest environments. Standardized setting | Walking (day trip). Standardized activity | PA in the NE, as compared to walking in urban environment, and reduced blood pressure, stress hormones and improve the profile of metabolic parameters. |
Librett et al. (2006) [102] | 4,345 adults | Cross sectional using a stratified random sampling design. | Use of trails. Self-reported | General moderate to vigorous PA and trails visitation. Standardized activity | Trail-users were more likely to achieve minimum recommended levels of PA. Presence of NEs was reported to be important for choosing a place where to live. |
Maas et al. (2008) [103] | 4,899 representing the overall population | Cross-sectional study using random sampling. | Availability of green-spaces within the neighbourhood (urban) Objective measure | Walking, cycling, sport activities and gardening Self-reported | In general, there was no significant association between availability of NEs and overall PA, e.g. meeting minimum recommendations. A negative relationship was found for cycling, while only gardening was positively associated with availability of NEs. Age-related differences were observed. |
Mao et al., 2012 [55] | 20 young males | Experimental trial with between-subjects design and pre-test measurements. Random allocation. | Shinrin-yoku (forest bathing trips) | Walking. Standardized activity | As compared with PA in an urban setting, PA in the NE was associated with improved mood and profile of physiological indicators of stress (cortisol and immune parameters). |
Mason, Kearns & Bond (2011) [68] | 5,657 adults | Cross sectional study using random stratified sampling design. | Availability of urban parks and open spaces within the neighbourhood. Self-reported | Walking. Self-reported | Use and quality of NEs, along with general shops, were associated with more PA. Physical (perceived safety) and social aspects were the strongest predictors of PA. |
Mayer et al., 2009 [53] | 232 young adults overall | Multiple experimental study using between-groups design with pre-test measurements. Random allocation not specified. | Experience of PA in natural settings and exposure to ‘virtual reality. | Walking. Standardized activity | As compared with urban settings, walking in a NE was associated with improved connectedness to nature and positive emotions, with connectedness to nature mediating the effects on positive emotions. Experiences of real nature led to greater effects than experiences of virtual nature did. |
McGinn et al. (2007) [104] | 1,659 adults | Cross-sectional study. Random digit dialled phone survey. | Perception of urban neighbourhood physical characteristics including natural elements as barriers to PA. Objective and self-reported measures | Leisure and transportation PA. Self-reported | There was little agreement between objective and perceived measurements of the environment. Perception of environmental barriers to PA (e.g. presence/lack of trees) was associated with different outcomes of PA. |
Michael et al. (2010) [80] | 513 older men | Prospective cohort study. Recruitment through clinical sites and stratified sampling. | Availability of urban parks and trials within the neighbourhood. Objective measure | Walking. Self-reported | Older men living closer to a NE had increased probability to maintain or increase amounts of PA as compare to men living farther from NEs. Though, socio gradient was observed. |
Michimi & Wimberly (2012) [105] | 931,116 adults overall | Two cross-sectional studies. Random digit dialled phone survey. | Proximity to outdoor recreational opportunities and availability of natural amenities in non-metropolitan areas. Objective measure | General PA. Self-reported | There was a positive association between availability of NE and PA, which was also associated with lower risk for obesity. |
Mitchell (2012) [57] | 3,750 adults | Cross-sectional study. Sampling design not described. | Experiences of PA in different NEs (forest, woodland, open space, or park). Self-reported | General PA. Self-reported | PA in quality NEs as compared to PA in other environments was associated with a lower risk for poor mental health. |
Morita et al., 2007 [51] | 498 adults | Experimental/Quasi-experimental trial using a mixed design with repeated measurements. Probable self-selection. Random allocation not specified. | Shinrin-yoku (forest bathing trips) | Walking. Non-standardized activity | As compared to a non-NE location, walking in a NE was associated with improvements of mood and anxiety. |
Mowen et al. (2007) [74] | 1,515 older adults | Cross-sectional study with recruitment of subjects on different sites. | Availability of urban parks. Objective measure | PA ‘status’ (i.e. sedentary/active). Self-reported | Proximity to NEs was associated with more frequent visits to NEs and PA. The duration of visits which was longer in subjects living farther away, was not associated with PA. |
Mytton et al. (2012) [106] | 54,296 adults | Cross-sectional study. Cluster sampling with selection of respondents to be representative of national population. | Availability of nature elements and green-spaces within the neighbourhood. Objective measure | General PA and specific outdoor PA. Self-reported | Residing in neighbourhood with greater availability of NEs was a predictor of meeting recommended levels of PA, although this association was not explained by type of PA typically taking place in NE. |
Nelson & Woods (2010) [81] | 2,159 adolescents | Cross-sectional study. Sampling technique not described. | Presence of natural elements along the way to/from school. Self-reported | Active commuting to school. Self-reported | Among other characteristics of the physical environment, there were some associations NEs and PA, although mixed results were found. Different patterns were observed across genders. |
Orsega-Smith et al. (2004) [58] | 100 older adult users of the district parks | Cross-sectional study. Onsite recruitment. | Use of urban parks. Self-reported | Park based PA ‘status’ (sedentary/active). Self-reported | An indirect effect, although weak, of NE-based leisure on stress and health was observed. The relationship between NE-use, PA and stress was not clear. |
Park et al. (2010) [47] | 280 young males overall | Multiple experimental trials, with pre-test measurements and counterbalanced cross-over design. Random allocation (method not described). | Forest environment. Standardized setting | Walking. Self-reported | As compared to urban environments, being in a NE was associated with an improvement of mood and physiological indicators of stress such as heart rate, blood pressure, salivary cortisol and indicators of cardiac autonomic control |
Pate et al. (2008) [82] | 1,506 girls | Cross-sectional study. Subjects recruited in schools (all students invited), which were chosen with the goal of providing balanced sample. | Availability of parks within the neighbourhood. Objective measure | Moderate and vigorous PA. Self-reported | Along with other PA/recreational facilities, availability of NEs was positively associated with girls’ participation in PA, although social-gradients (based on ethnicity) were observed. |
Pretty et al. (2005) [45] | 100 adults | Experimental trial, with pre-test measurements and between-subjects design (four treatments and a control). Random allocation (method not described). | Exposure to images of NEs on a screen. Standardized setting | Light-intensity treadmill exercise. Self-reported | PA whilst viewing images of pleasant NEs was associated with reduced blood pressure, and improvements in self-esteem and mood profile, as compared with PA whilst viewing scenes of built environment. Images of unpleasant NEs had the most dramatic effect on psychological responses to PA. |
Pretty et al. (2007) [46] | 263 adults | Quasi-experimental, with pre-test measurements. Cluster sampling design used to select random sample of cases. | Different setting for ‘green-exercise’/outdoor recreations activities in the countryside. Standardized setting | Different ‘green-exercise’ (PA in natural environment) e.g. walking in NE, horse-riding, and fishing, canal boating and conservation activities. Existing interventions | Experiences of green exercises were associated with improved mood and self-esteem. |
Prince et al. (2011) [70] | 3,883 adults | Cross sectional study. Random digit dialled phone survey, applying standard survey weights. | Availability of urban green-spaces and parks within the neighbourhood. Objective measure | General PA. Self-reported | No associations or even some negative ones were found between availability of NE and PA. Association between PA and food environment, social cohesion, socio-economic status of the neighbourhood and were observed. Different patterns across genders were observed. |
Prince et al. (2012) [71] | 4,727 adults | Cross sectional study. Random digit dialled phone survey, applying standard survey weights. | Availability of urban green-spaces and parks within the neighbourhood. Objective measure | Leisure time PA. Self-reported | There was some association between availability of NEs and PA. Associations were observed also for food environment, crime and season with different patterns across genders. |
Rhodes et al. (2006) [31] | 315 adults | Cross sectional study using random sampling. | Availability of attractive natural sight within the neighbourhood. Self-reported | Walking. Self-reported | Theory of planned behaviour [28] efficiently predicted the environment-walking relationship, with NEs and land-use-mix predicting PA via affective and instrumental attitudes and, to a lesser extent by subjective norms. Perceived behavioural control did not relevantly weighted in the model. |
Richardson et al. (2013) [107] | 12,488 youth and adults | Cross-sectional study. Sampling design not described. | Availability of vegetation and green-spaces within the neighbourhood. Objective measure (national databases) | Moderate to vigorous PA. Self-reported | Availability of NEs was associated with probability to meet recommended PA levels, and reduced risk for poor mental health and cardiovascular diseases. |
Ries et al. (2009) [66] | 329 adolescents (predominantly Afro-American) | Cross-sectional study. Non-randomized recruitment among two schools. | Availability of urban parks within the neighbourhood. Objective and self-reported measures | General PA and park-based PA. Self-reported and measured by accelerometry | Associations between the availability of NEs and PA were observed. Use of NEs by peers, age, gender and ethnicity/race also influenced the NE-PA relationship. |
Rodriguez et al. (2012) [108] | 293 adolescent girls | Prospective cohort study. Recruitment among from a previous study. | Availability of urban parks within the neighbourhood. Objective measure | Moderate to vigorous PA. Accelerometry and GPS | NE, along with presence of schools and population density, was positively associated with PA. Road length and number of food outlets was negatively associated with PA. |
Ryan et al. (2010) [49] | 66 young adults (study 2) | Experimental trial, with pre-test measurements and between-subjects design. Random allocation (method not described). | Experience of PA in a NE. Standardized setting | Walking. Standardized activity | Walking in a NE had greater impacts on subjective vitality than walking in an interesting and varied indoor setting. |
Scott & Jackson (1996) [67] | 1,054 adults | Cross-sectional. Random digit dialled phone survey. | Use of different types of public parks. Self-reported | Park visitation. Self-reported | ‘Lack of time’ was the most commonly reported reason for not visiting NEs. Older women were less likely to visit parks because of fear of crime, lack of companionship and poor health. Improvement of programming and information would encourage more visits to NEs. |
Sharpe et al. (2004) [3] | 1,936 adults | Cross-sectional. Random digit dialled phone survey. | Availability and use of parks and other outdoor recreations facilities (e.g. trails and routes for walking and cycling). Self-reported | Moderate to vigorous PA. Self-reported | Among other environmental and policy factors, use of NEs was associated with PA. However, knowledge and quality of the infrastructure were important factors in determining the use of NEs for PA purposes. |
Shores & West (2009) [88] | 139 young adults | Observational study using convenience sample (students recruited during classes). | Perception of public parks as a source of ‘leisure’ (“activity enjoyable for its own sake”). Self-reported | Leisure time PA. Self-reported | Most of PA perceived as ‘leisure’ was carried out in private fitness centres and dance clubs. NEs were only a small source of the leisure-time PA. Companionship and sociality appear to be important factors for engaging in leisure PA. |
Shores et al. (2008) [72] | 454 older adults | Cross-sectional study using stratified random sampling design. | Availability of parks in rural areas. Self-reported | PA status (active or inactive). Self-reported | Proximity to NEs was positively associated with PA. Other important variables predicting PA were access to social support, safety and reported ability to walk to a local park. |
Stigsdotter et al. (2010) [87] | 11,238 adults | Cross-sectional study using stratified random sampling design. | Proximity to green-spaces and motives to visit them. Self-reported | Various PA and outdoor recreations. Self-reported | PA was reported as the most important reason for visiting NEs from less stressed individuals; those who are more stressed visit NEs to relax, seek for quite places and engage in social activities. Proximity to NEs was associated with more frequent visits to NEs, better quality of life and less stress. |
Sugiyama et al. (2008) [24] | 1,895 adults | Cross-sectional study using cluster random sampling design. | Perception of availability of nature and green-spaces within the neighbourhood (urban). Self-reported | Leisure and transport walking. Self-reported | The NE was a predictor of PA and social factors. PA explained the link between the NE and physical health. The relationship between NE and mental health was only partly accounted for by PA and social coherence. |
Thompson et al. (2012) [27] | 25 adults | Observational exploratory study, using convenience sampling design. | Availability of green-space within the neighbourhood (urban). Self-reported | General PA. Self-reported | Availability of NEs was associated with less perceived stress and, along with PA, with better cortisol profiles. A direct association between availability of NEs and PA was not found. |
Toftager et al. (2011) [109] | 21,832 adults | Cross-sectional study using stratified random sampling design. | Proximity to green spaces (beach, seaside, lake, park, urban green space, forest or other open green spaces excluding agricultural fields). Self-reported | Leisure time moderate to vigorous PA. Self-reported | Proximity to NEs was associated with more PA. People living closer to NEs had increased chances in using it for PA purposes. |
Van den Berg & Custers, (2011) [54] | 30 middle-aged adults | Experimental trial with between-subjects design and pre-test measurements. Random allocation. | Experience of gardening activity. | Gardening. Standardized activity | After an induced stress, both reading and gardening were associated with an improvement in stress parameters (mood and salivary cortisol), with greater improvements observed after gardening. The difference between the ‘stress recovery’ interventions was somewhat weak. |
Ward Thompson et al. (2012) [110] | 96/61 older adults | Cross-sectional and longitudinal cohort / quasi-experimental study, with pre- and post-intervention measurements. | Availability of urban parks and other natural environments. Self-reported / perceived | General PA & any outdoor activity. Self-reported and measured by accelerometry | Availability of clean and quite NEs with attractive, barrier-free routes to it was positively associated with more PA. No significant change of PA ratings was found after the neighbourhood improvement intervention. |
Wen & Zang (2009) [111] | 3,530/907 adults | Multilevel cross sectional study. Random digit dialled phone survey. | Availability of urban parks within the neighbourhood. Objective measure | Exercise behaviour. Self-reported | Access to NEs was not associated with PA. Predictors of PA were social capital and access to restaurants/bars. |
Wen et al. (2007) [73] | 41,545 adults | Cross-sectional study. Random digit dialled phone survey. | Availability of parks/open-spaces within the neighbourhood. Self-reported | Leisure and transport walking. Self-reported | Availability of NEs along with neighbourhood social cohesion was positively associated with PA. Individual socio-demographic and safety did not have significant effects on PA, while differences for race/ethnicity were observed. |
West et al. (2012) [112] | Adult respondents within 67 metropolitan statistical areas | Cross-sectional study using random sampling design. | Availability of urban parks within the neighbourhood. Objective measure | Moderate and vigorous PA. Self-reported | Availability of NEs was positively associated with probability of meeting PA recommendations, and negatively associated with risk of being overweight/obese. |
Wilson et al. (2011) [86] | 10,286 middle-age adults | Cross sectional study using a stratified two-stage cluster design. | Availability of different NEs. Objective and self-reported measure | Walking. Self-reported | Different NEs had different effects on PA. E.g. proximity to rivers and coasts was positively associated, while tree coverage was negatively associated and proximity to parks was not associated. Other environmental factors also predicted PA. |
Witten et al. (2008) [113] | 12,529 youths and adults | Cross-sectional study. Sampling design not reported. | Availability to different NEs. Objective measures | Brisk walking, moderate and vigorous PA. Self-reported | Different NEs had different effects on PA and BMI. E.g. parks were not associated with PA, while some associations were found for beaches/coasts. |
Yamaguchi et al. (2006) [62] | 15 young males | Experimental trial, with counterbalanced cross-over design (two treatments and a control). | Low threshold exercise in NE. | Walking. Standardized activity | As compared to urban environment, PA in NE was associated with an improved profile of indicators of sympathetic nervous activity (salivary amylase). |