3.1.1 Results focusing on Area Variables.
At the community level, four studies indicated that high-SES areas tended to have higher walkability than low-SES areas [25–28]. Jacobs et al. [29] did, however, find variations across their studies included in their review, with some studies highlighting that areas with higher SES tend to have superior walking infrastructure and greater amount of walking tracks, while other studies find the opposite. The presence and accessibility of walking facilities are generally identified as supporting walking [30, 31]. In contrast to studies suggesting higher walkability in high-SES areas, an inverse relationship was found in studies by Choi and Yoon [32] and Conderino et al. [2]. The last study reported that on average, low-income neighbourhoods had higher walking scores than high-income ones. Notably, most white neighbourhoods generally had lower walk scores than other racial/ethnic majority neighbourhoods, except for the majority of black neighbourhoods, where tracts in lower income tertiles had the lowest walkability.
Perceptions within neighbourhoods also affect objective walkability measurements. Higher-income areas are often perceived as more aesthetically pleasing, with higher quality, fewer physical barriers to walking, and lower levels of crime and traffic [33, 34]. Conversely, low-SES areas tend to have poorer perceived built environmental experiences [30, 31, 34]. Giles-Corti and Donovan [30] suggest that the quality of the built walking environment may be more important than the SES of the area of residence, as a correlate of walking behaviour. This is supported by findings concluding that built environmental factors, such as pedestrian bridges over large roads, well-maintain pavements, and illuminated walk-and-bike paths, are encouraging for walking [31, 35].
Another aspect related to built environmental opportunities for PA is the density of the area. Remote areas tend to have poorer walking and bicycle infrastructure, lower walkability, and less favourable structural attributes for PA [36]. Two studies found that areas with higher intersection density and connectivity, often urban, with multiple destinations and branched road networks, tend to promote walking and meeting PA recommendations [37, 38]. The same result was observed among low-SES adults in a study by Christie et al. [39]. In contrast, Boone-Heinonen and Gordon-Larsen [40] found that higher landscape diversity was associated with higher PA, and for females, higher street connectivity was linked to lower PA. Furthermore, Isiagi, Okop, and Lambert [41] observed a negative association between intersection density and PA regardless of group. Wang et al. [42] also inversely observed a positive association between the built environment and PA in neighbourhoods characterised by low housing density, low road coverage, less land-use diversity (e.g. single land use of residence), high car dependency, poor access to public transport, longer distances to the city, and more green space coverage. Similarly, Frost et al. [43] found positive associations between aesthetics, pathways, safety from crime and traffic, parks, walkability, and PA among adults in rural areas.
Furthermore, three studies found that living in high-SES areas is closely related to increased active transportation, higher PA levels, or more steps pr. day [35, 44, 45]. However, Seguin-Fowler et al. [44] found no association between the walk score and PA for those living in low-SES neighbourhoods. Isiagi, Okop, and Lambert [41] inversely found that residents in low-SES/high walkable neighbourhoods reported more transport-related PA compared to high-SES/low walkable neighbourhoods. Similar results were found by Besor et al. [46], who stated that areas characterised by lower-SES residents and a higher proportion of Arab minorities had better-performing health programmes (higher PA). Zang et al. [45] found that the PA of people living in low-SES areas was more dependent on the built environment, whereas the association was limited in high-SES areas. In studies of interventions in both high- and low-SES areas, a positive change in neighbourhood walkability was associated with increased PA, especially in adults in low-SES areas [27, 47, 48]. In a study by Clary et al. [27], improvements in walkability scores were mostly driven by increases in residential density and land-use mix. In contrast, Adkins et al. [49] concluded that the built environment has weaker effects on walking and physical activity in disadvantaged groups than in advantaged ones.
In summary, area-specific studies had different indications. Some studies found varying associations between area walkability and PA (including transportation walking) [50, 51], whereas others reported clear associations between higher walkability scores and increased PA across different SES areas [28, 52]. Finally, Hillsdon et al. [53] found that most people engage in PA beyond an 800-metre radius from their homes, suggesting that neighbourhood characteristics alone may not predict PA levels.
3.1.2 Results focusing on Individual Variables.
At the individual level, multiple studies have shed light on the interplay between determinants of the social environment, walkability variables, and PA. Gullon et al. [54] indicated that individuals with lower income levels tend to have more accessible walking destinations nearby. Furthermore, Christe et al. [47] revealed that the percentage change in walkability was positively associated with increased walking, particularly among those with lower income and education levels. Conversely, Cerin and Leslie [33] found that individuals with higher education and income may choose and afford to live in more PA-friendly built environments, including areas conducive for walking. Similarly, Andrade et al. [48] observed that individuals with higher incomes have better access to free or low-cost recreational facilities (including walking trails), a pattern that is also prevalent among those with higher education and more working hours. When examining the use of newly built walking and cycling infrastructure, Smith et al. [55] found that lower educational level and income, rather than ethnicity, were associated with reduced usage.
Dias et al. [56] explored the associations between built environmental factors (objectively and subjectively) and leisure walking among boys and girls with different SES backgrounds. For girls with low SES, access to services and shorter distance to parks and squares were positively associated with leisure walking. For boys, perceived environmental factors such as crime safety, land-use mix, neighbourhood recreation facilities, and places for walking are crucial factors for leisure walking. Another relevant study by Burton et al. [57] revealed that participants across income groups (low, intermediate, and high) place equal importance on similar factors, such as low crime, friendly neighbours, streetlights, and good paths, according to PA. Individuals with higher incomes only marginally emphasised these factors in their PA considerations. Similarly, Cleland et al. [58] found that individual factors, especially those of women with low SES, outweighed environmental factors. Specifically, higher PA levels among low-SES women were associated with interesting local walking opportunities and busy roads to cross during walking.