Table 2 Study characteristics and summary of findings for studies including moderation or stratified analysis
Authors (year), study | Study setting & study design | N & age & % girls | Assessment of built enviornment | Assessment of socioeconomic position | Outcome | Additional variables included in the statistical model | Type of analysis | Findings |
|---|---|---|---|---|---|---|---|---|
Babey 2008 [50] California Health Interview Survey | USA Cross-sectional | 4010 Age range: 12–17 % girls not stated | Perceived environment Access to recreational facilities | Parental-reported income | Self-reported Total physical activity | Age, sex, race/ethnicity, urbanicity, housing types, neighborhood safety | Formal test of interaction | No significant interaction. Models of regular activity and inactivity that tested interactions between access to a safe park and SEP were not significant. → |
Bringolf-Isler 2014 [24] Pooled analysis of 7 Swiss studies | Switzerland Cross-sectional | 1742 Age range: 4–17 498% girls | GIS Accessibility Aesthetics Safety | Area-level index | Accelerometery Total physical activity Moderate-to-vigorous physical activity | Age, sex, season, accelerometer model, study-cluster, + moderate-to-vigorous physical activity for wear-time | Formal test of interaction | Significant interaction for total physical activity and moderate-to-vigorous physical activity. Associations between physical activity and accessibility were stronger in low SEP neighborhoods. ↑ |
Bringolf-Isler 2019 [25] Swiss children’s Objectively measured Physical Activity (SPOHYA) | Switzerland Cross-sectional | 1306 Age range: 6-16 9% girls | Perceived environment Safety Access to recreational facilities GIS Accessibility Safety Aesthetics | Area-level index | Accelerometery Moderate-to-vigorous physical activity | Age, sex, household income, language region, urbanicity, individual SEP | Formal test of interaction | Significant interaction for moderate-to-vigorous physical activity. Associations between physical activity and accessibility were positive in low SEP areas, but negative in high SEP areas. ↑ |
Clennin 2019 [26] Transitions and Activity Changes in Kids | USA. Prospective 2 year of follow-up | 660 Mean (SD) age: 10.6 (0.1) 4% girls | GIS Access of recreational facilities | Area-level index. | Accelerometery Total physical activity | Age, sex, race/ethnicity, parent education, and weight-status | Formal test of interaction | No significant interaction. Access to recreational facilities did not moderate the association between deprivation and change in physical activity over two years. → |
da Silva 2018 [55] Brazilian part of the International Physical Activity and Environment Network (IPEN) Study. | Brazil Cross-sectional | 495 Age range: 12–17 years 1% girls | Perceived environment Safety | Parental-reported SEP Index | Self-reported Active transportation | Age, sex, perception of time spent to walk to school, leisure time moderate-to-vigorous physical activity, period spent in school + car ownership for analysis of parent reported built environment | Stratified analysis | Poisson regression showed largely similar weak associations with active transportation across SEP-categories for both adolescent and parent reported built environment. → |
De Meester 2012 [23] Belgian Physical Activity study in Youth (BEPAS-Y) | Belgium Cross-sectional | 513 Mean (SD) age: 14.6 (0.9). 0% girls | GIS Accessibility | Area-level household income | Accelerometry Total physical activity MVPA | Parental employment and educational attainment of mother and father, neighborhood-clustering | Formal test of interaction | Significant interaction for moderate-to-vigorous physical activity and total physical activity. Accessibility was positively associated with physical activity among children living in low SEP neighborhoods only. ↑ |
D’Haese 2014 [22] Belgian Physical Activity study in Children (BEPAS-Child) | Belgium Cross-sectional | 494 Mean (SD) age: 10.1 (0.9) 55% girls | GIS Accessibility | Area-level household income | Self-reported Active transportation Accelerometry Moderate-to-vigorous physical activity | Age, sex, mother or father with college/university education | Formal test of interaction | Significant interaction for walking for transportation during leisure time. Accessibility was associated with more active transportation during leisure time among children living in low SEP neighborhoods only. No interaction for moderate-to-vigorous physical activity and active transportation to school was found. ↑ |
Diaz 2019 [46] No study name | Brazil Cross-sectional | 1130 Mean (SD) age: 16.5 (1.1) 3% girls | Perceived environment Accessibility Safety Aesthetics GIS Accessibility | Self-report SEP index | Self-reported Active transportation | Age, sex, SEP, class cluster | Stratified analysis | The associations between thirteen different built environment exposures and active transportation were largely similar and non-significant across SEP tertiles. Perceived land-use mix and perceived access to recreational facilities were associated with active commuting to school among adolescents in middle SEP tertile only. →. |
Hunter 2020 [39] Recording and Evaluating Activity in a Modified Park (REVAMP) | Australia Cross-sectional | 1212 Mean (SD) age: 8.5 (3.6). 483% girls | Perceived environment Accessibility Access to recreational facilities | Area-level index | Parental reported Total physical activity | Sex, typical behaviour of child during last 7 days, park recruitment area, year of data-collection | Formal test of interaction | Significant interaction for total physical activity. Park quality was positively associated with physical activity among children living in high SEP neighborhoods, but not in low and medium SEP neighborhoods. ↓ |
Isgor 2011 [47] Child Development Supplement of the Panel Study of Income Dynamics Study | USA Cross-sectional | 634 Mean (SD) age: 15.8 (1.22) 506% girls | GIS Access to recreational facilities | Self-reported family income | Self-report Total physical activity | Age, sex, ethnicity, marital status of family head, mother’s education, mother’s work status, zip-code level urbanization, zip-code level median household income | Stratified analysis. | No significant difference by SEP in the association between facility availability and PA among male adolescents. Statistically significant differences by SEP in the associations between facility availability and physical activity among female adolescents. Additional facilities being more important in low SEP but no impact in high SEP. ↑ |
Johanson 2012 [51] Health Behaviour in School-aged Children + Stockholm county | Sweden Cross-sectional | 5423 5th − 9th grade 50% girls | Perceived environment Accessibility | Self-reported employment type | Self-report Active transportation | Car ownership | Stratified analysis. | In urban areas, children of low SEP were more likely to use active transportation compared with children of high SEP. In rural areas, only children of self-employed persons (including farmers) were less likely to use active commuting compared to those of intermediate- to high-level salaried employees. ↑ |
Kerr 2006 [52] Neighborhood Quality of Life Study | USA Cross-sectional | 259 Mean (SD) age 11.3 (4) 486% girls | GIS Accessibility | Area-level household income | Parental reported Active transportation | Sex and age | Formal test of interaction | Significant interaction for active transportation. In low-walkability neighborhoods, income was not related to active transportation. In high-walkability neighborhoods, children in high-income neighborhoods were more likely to actively commute. ↓ |
Kim 2020 [53] Healthy Communities Study | USA Cross-sectional | 4114 Age range: 4–15 50% girls | GIS Accessibility | Parental reported household income | Self-report Moderate-to-vigorous physical activity | Age, sex, race/ethnicity, parent education, neighborhood SEP history | Stratified analysis | No significant difference by SEP in the association between urbanicity and PA among children. → |
Liu 2008 [54] National Survey of Children’s Health | USA Cross-sectional | 44,631 Age range: 10-17 49% girls | GIS Accessibility | Parental reported family poverty status | Parental reported Total physical activity | Age, sex, race/ethnicity, overall health status, screen time, household poverty status/ parental education (mutual adjustment), number of children in household, parental physical activity, Perceived neighborhood safety | Stratified analysis | No consistent and meaningful difference by SEP was observed in the association between urbanicity and PA among children. → |
McCrorie 2020 [28] Studying Physical Activity in Children’s Environments across Scotland | Scotland Cross-sectional | 774 Mean (SD) age: 11.1 (0.3) 53% girls | GIS Accessibility | Household income | Accelerometry Total physical activity, Light intensity physical activity Moderate-to-vigorous physical activity | Sex, wear-time, number of valid days, season, distance to school, data zone walkability | Formal test of interaction | No significant interaction. The relationship between urban/rural settlement did not vary by household income quintile. → |
Molina-Garcia 2017 [40] International Physical Activity and Environment Network (Spain) | Spain Cross-sectional | 310 Mean (SD) age: 11.2 (0.5). 49% girls | GIS Accessibility | Parental reported education | Self-report Active transportation | Unclear as variables removed if p > 0.15. Considered variables are age, gender, weight-status, adult accompaniment to school, number of motor-vehicles per licensed driver, distance to school, environmental/safety and planning (psychosocial barriers, participant clustering within neighborhoods | Formal test of interaction | No significant interaction. Socioeconomic position did not influence the strength of association between the built environment and physical activity. → |
Molina-Garcia 2017 [48] International Physical Activity and Environment Network (Spain) | Spain Cross-sectional | 325 Mean (SD) age: 16.4 (0.8) 542% girls | GIS Accessibility | Parental reported education | Accelerometry Moderate-to-vigorous physical activity Self-report Active transportation | Unclear as variables removed if p > 0.15. Considered variables are age, gender, days per week living at primary address, distance to school, driver license, number of children < 18 years old living at household, number of motor vehicles per licensed driver, number of years living at current address, and workout equipment in or around home, participant clustering within neighborhoods | Formal test of interaction | Significant interaction for moderate-to-vigorous physical activity during weekends. In low-walkable neighborhoods, moderate-to-vigorous physical activity during weekends were higher with among low SEP children, while in high-walkable neighborhoods, moderate-to-vigorous physical activity during weekends were higher in high SEP. ↓ |
Oyeyemi 2014 [49] No study name | Nigeria Cross-sectional | 1006 Mean (SD) age: 15.6 (1.7) 50% girls | Perceived environment Accessibility Access to recreational facilities Safety Aesthetics | Area-level income | Self-report Active transportation | Age, sex, school grade, parental education and occupation, all BE factors | Formal test of interaction | No significant interaction. Neighborhood income did not influence the associations of environmental perception and active transportation. Significant interaction is observed for leisure time in boys, however, as leisure time is restricted to sports participation, we have chosen to disregard this association. → |
Page 2010 [41] Personal and Environmental Associations with Children’s Health | England Cross-sectional | 1300 Age (range): 10-11 49% girls | Perceived environment Accessibility Safety Aesthetics | Area-level deprivation | Self-report Active transportation | Sex, daylight hours, pubertal stage, BMI, school clustering | Formal test of interaction | No significant interaction between neighborhood deprivation and accessibility in the association with active transportation (results not shown). → |
Rodrigues 2018 [42] No study name | Portugal Cross-sectional | 834 Mean (SD) age: 8.1 (1.2). 49% girls | GIS Accessibility | Parental reported education | Self-report Active transportation | Bivariate: none, Multivariable: not applied for relevant variables in boys( SEP not < 0.05 in bivariate model). Girls: age, sex, obesity, abdominal obesity, perceived safety, facilities close to home, distance to school, mother’s transportation, father’s transportation | Stratified analyses | No association between SEP and active transportation in urban setting. Bivariate analysis shows higher odds of active transportation with low SEP in rural setting. The association is attenuated after multivariable adjustment. → |
Sallis 2018 [29] Teen Environment and Neighborhood Study | USA Cross-sectional | 928 Mean (SD) age: 14.1 (1.4) 50.4% girls | GIS Accessibility | Area-level median household income | Accelerometery Moderate-to-vigorous physical activity Self-report Active transportation | Potential variables were (removed if p > 0.15): age, sex, race/ethnicity, has driver’s license, attends school away from home, days/week living at current address, works outside the home, number of children in household, number of motor vehicles per licensed driver, years at current address, walkability-related reasons for moving here, study site, census block clustering (+ for accelerometry: acc model and wear-time) | Formal test of interaction | Significant interaction for active transportation to school. In low walkable areas, low SEP children had higher prevalence of active transportation than high income children. No difference in active transportation between low and high SEP in high walkable areas. No significant interaction for moderate-to-vigorous physical activity. ↓ |
Shams-White 2021 [30] Family Life, Activity, Sun, Health and Eating Study | USA Cross-sectional | 1295 Age range: 12-17 50.4% girls | GIS Accessibility | Area-level index | Self-report Moderate-to-vigorous physical activity | Age, gender, ethnicity, parent education, neighborhood urban/rural location | Formal test of interaction | Significant interaction for moderate-to-vigorous physical activity. Strongest association between built environment and physical activity in middle SEP. No association with moderate-to-vigorous physical activity in lowest and highest SEP (Q2-Q4 have slopes) → |
Stone 2014 [43] Project BEAT | Canada Cross-sectional | 856 Mean (SD) age: 11.0 (0.6). 55% girls | GIS Accessibility | Area-level median household income | Accelerometry Total physical activity Light intensity physical activity Moderate-to-vigorous physical activity | None | Stratified analysis | Main effect of SEP-built environment combination for weekday total, MVPA and LPA in boys and girls are presented. These are further stratified by independent mobility, precluding a clear interpretation of the results. → |
Su 2013 [44] Southern California Children’s Health Study | USA Cross-sectional | 4338 Mean (SD) age: 6.6 (0.7) 48.2% girls | GIS Accessibility | Parental reported education | Parental reported Active transportation | Age, sex, ethnicity, distance to school, traffic density, percent government and institutional, percent free and reduced-price meals program | Formal test of interaction | Significant interaction for active transportation. Accessibility associated with more frequent active transportation among high SEP children only. ↓ |
Uys 2016 [45] The International Study og Childhood Obesity, Lifestyle and the Environment (South Africa) | South Africa Cross-sectional | 258 Mean (SD) age: 10.2 (0.6). 59% girls | GIS Access to recreational facilities Perceived environment Access to recreational facilities Safety | Self-reported household income | Accelerometery Total physical activity Moderate-to-vigorous physical activity | Age, sex | Formal test of interaction | Significant interaction for moderate-to-vigorous physical activity. Physical activity facilities positively associated with physical activity in low SEP children, but not in high SEP children. High traffic risk was associated with less activity among low SEP children, but not in high SEP children. ↑ |
Uzochukwu 2017 [56] National Survey of Children’s Health | USA Cross-sectional | 25,092 Mean (SD) age: 13.6 range (10–17) 48% girls | GIS Accessibility Perceived environment Accessibility Aesthetics | Parental reported household income | Parental reported Total physical activity | Race/ethnicity, SEP, physical amenities, social cohesion, physical detractions, social detractions. | Stratified analysis Formal test of interaction | Stratified analysis suggested stronger correlation between SEP and physical activity in rural areas. No significant interaction. → |