Socio-Ecological Correlates of Active School Travel in Primary School Children: A Cross-Sectional Analysis

Objectives: To counteract the decline in active school travel (AST), this study applies a socio-ecological framework to examine individual, family, and environmental correlates of AST in primary school children, with a particular focus on parental perceptions.

Methods: Data on anthropometry, physical activity (PA), family and environmental factors of first graders (n=471, 53% male) were collected through screening and parent-reported questionnaires. Associations between socio-ecological variables and school travel behavior (active vs. non-active) were analyzed using hierarchical binary logistic regression. Variables were entered in blocks (individual, family, environment). Model fit was assessed using the Hosmer-Lemeshow test and Nagelkerke’s R².

Results: Perceived walking time to school showed the strongest negative association with AST (OR=0.91, p<0.001). Several family-level correlates, including maternal education level, household size, family health climate for PA, parental walking, and parental moderate-to-vigorous PA, were significantly associated with AST (p<0.05). At the individual level, higher body fat percentage was negatively associated with AST (p<0.05).

Conclusions: The findings highlight the complex interplay of individual-level characteristics, family factors, and perceived environmental conditions in shaping children’s AST. Parental perceptions of walking distance emerged as a key correlate, alongside family-level factors such as parental attitudes toward PA and role modeling of PA. These findings underscore the need for interventions that extend beyond infrastructural improvements by engaging parents in promoting active lifestyles and supporting children’s independent mobility.

Key Words: Active Commuting, Physical Activity, Family, Parental Perception, Physical Environment

Physical activity (PA) in childhood is essential for health and well-being (5, 23). However, insufficient PA levels among children have become a growing public health concern, contributing to an increased prevalence of overweight and obesity, as well as related metabolic and cardiovascular risks (7, 41).

ongitudinal research indicates that PA behaviors established in childhood often track into adulthood (38), highlighting the need for early interventions to promote active behaviors. Notably, the transition to primary school is associated with a rise in sedentary time, further exacerbating declines in PA (25).

Among the various settings that can support children’s daily PA, the school commute offers a particularly promising and accessible opportunity. Active school travel (AST), such as walking or cycling, has been identified as a promising strategy to increase PA in children. A substantial body of evidence links AST to higher overall PA levels (10, 13), as well as to a reduced risk of overweight and obesity (15). In addition, AST is positively associated with cardiovascular fitness and well-being (24). Beyond the individual health benefits, active travel modes contribute to sustainability by reducing traffic-related emissions and fostering more livable urban spaces (14).

Despite the known benefits, studies from various countries report a decline in the number of children actively traveling to school (22, 26, 29). In Germany, primary school catchment areas are intended to place most children within reasonable walking or cycling distance of their school. However, studies report that only half to two thirds of German primary school children actively commute to school (12, 32, 39). 
To understand children’s school travel behavior, the socio-ecological model offers a useful framework by highlighting the complex interplay of correlates at multiple levels to shape an individual’s behavior (30). AST is similarly influenced by a range of correlates at the individual, family, and physical-environment levels (13, 29).

At the individual level, factors such as age, sex, socioeconomic status (SES), and migration background have been examined in relation to AST, but results vary. While boys are generally more likely to engage in AST than girls (3, 33), associations with SES and migration background remain inconclusive, with studies reporting both positive and negative correlations (11, 12, 26, 29).

The family level underscores the influence of parental support (18, 21) and role modeling (37), with parents taking on the role of ‘gatekeepers’ (27), yet the extent of this impact remains debated. In young children, the family plays a crucial role in shaping PA behavior, whereas parental influence tends to decline during adolescence (1). The choice of travel mode likely results from both parental and child perceptions, with children’s participation in the decision-making process increasing with age (20). In addition to parental support, broader social influences such as the presence of peers or siblings to walk with can also facilitate AST, as social companionship increases children’s perceived safety and independence (2, 21, 27).

At the physical-environment-level, distance to school is consistently identified as the strongest predictor of AST (26, 28, 31, 33, 34). In addition, parental perceptions of neighborhood characteristics are crucial correlates, as concerns about traffic and safety often influence whether children are allowed to walk or cycle to school (2, 26, 31). For example, parental perceptions of the environment, particularly safety concerns, have been found to be important correlates of children’s active travel (9).

However, research on the correlates of AST in young primary school children remains inconsistent (18, 34), particularly regarding family-related socio-ecological factors. In addition, most AST studies focus on children aged 9 years or older (13), leaving a gap in research on younger primary school children. Building on these findings, this study aims to examine the complex interplay of socio-ecological correlates of AST in first-grade primary school children, with a particular focus on parental perceptions, behaviors and attitudes influencing school travel behavior. A deeper understanding of these influences is crucial for developing targeted interventions to promote healthier and more sustainable school travel modes. 

Study Design and Participants
This analysis is based on cross-sectional baseline data from the second wave of the KOMPASS(2) study, a longitudinal school-based survey conducted in Leipzig, Germany – a city of approximately 630,000 inhabitants. The study was approved by the local research ethics committee in 2014 (Medical Faculty of Leipzig University, file number: 253-14-14072014).

Children and one of their parents or legal guardians were recruited from 27 primary schools during the 2018-19 school year. The selection of schools was based on participation in the first study wave and established school contacts, with participation being voluntary. Before participating in the study, all parents or legal guardians of children were informed about the content and the measurement procedures of the study. Written informed consent was subsequently obtained. Reasons for the exclusion of a child included congenital or chronic cardiovascular disease, acute infectious disease, or acute musculoskeletal injury (36). 

Measure and Instruments
Data relevant for this analysis were obtained through a parent-reported questionnaire and anthropometric screening. The questionnaire assessed children’s PA behavior as well as family-level and physical-environment-level correlates. 

Outcome Variable
The child’s most frequent mode of travel to school (walking, cycling, public transport, or private motorized transport) was reported by parents. Travel modes were dichotomized into active (walking, cycling) and non-active travel (public transport, private motorized transport).

Individual-Level Correlates
Parents reported children’s age, sex, and migration background. Anthropometric data were collected using a portable digital stadiometer to measure height (m) and a SC 240 digital body composition scale (Weightcheckers®) to measure weight (kg) and body fat percentage. Body Mass Index (BMI, kg/m²) and age- and sex-specific BMI percentiles were calculated. For analysis, body fat percentage was used instead of BMI, as it more directly reflects adiposity and has greater health relevance in childhood. Unlike BMI, which does not distinguish between fat and lean mass, body fat percentage has been shown to be more strongly associated with cardiometabolic risk factors and obesity-related health outcomes in children (40).

Family-Level Correlates
The number of people in the household and parental characteristics – maternal and paternal education level, parental moderate-to-vigorous physical activity (MVPA), parental walking and family health climate for PA (FHC-PA) – were assessed using the parental questionnaire. MVPA and walking were each measured as the number of physically active days per week on a 0 to 7-day scale. For walking, only activities lasting at least 10 minutes without interruption were counted. The FHC-PA scale captured PA values, behavior routines and interaction patterns regarding PA in everyday family life (19). For analysis, education level (low vs. high), MVPA, and walking (< 5 vs. ≥ 5 days/week) were dichotomized.

Physical-Environment-Level Correlates
Neighborhood perceptions were assessed using selected items adapted from the validated European environmental questionnaire ALPHA (35), including perceived residential density, neighborhood quality for PA (attractiveness, safety, amenities, and location) and perceived walking time from home to school and to the nearest playground. These items aimed to capture parents’ subjective evaluations of environmental conditions potentially influencing children’s AST.

Data Analysis
Prior to statistical analysis, missing values and extreme values were addressed using established data cleaning procedures to ensure overall data quality. To examine the association between socio-ecological factors and AST, a hierarchical binary logistic regression model was employed. The dependent variable was AST (0=no, 1=yes). Independent variables were entered into the model in three blocks based on theoretical considerations reflecting the socio-ecological framework.

- Model 1 - Individual-level correlates: children’s sex, age, body fat percentage
- Model 2 - Family-level correlates: maternal and paternal education level, parental MVPA, parental walking, FHC-PA, household size
- Model 3 - Physical-environment-level correlates: neighborhood density, neighborhood quality for PA, walking time to school, walking time to playgrounds.

The enter method was used, including all variables per block simultaneously based on theoretical reasoning, rather than selected through statistical criteria. This approach allowed for the evaluation of the incremental contribution of each block to the overall model. Model fit was assessed using the Hosmer-Lemeshow goodness-of-fit test, Nagelkerke’s R², and classification accuracy. Multicollinearity was evaluated using the variance inflation factor (VIF), with values above 5 considered problematic (8). Odds ratios (OR) with 95% confidence intervals (CI) were reported. Analyses were conducted using SPSS (Version 29, IBM Corp.), with statistical significance set at p<0.05.

Sample Characteristics
Data from n=471 children (7.39±0.39 years, 53.1% male) were included in the statistical analyses after exclusion of invalid values and outliers. Table 1 summarizes participants’ anthropometric, socio-demographic, and parental characteristics, stratified by active and non-active school travel. In total, 64.8% of the children were classified as active school travelers.

Correlates of AST
Results of the hierarchical binary logistic regression are presented in table 2.

Model 1 included individual-level correlates and served as the baseline. The analysis showed that fat mass (%) was significantly negatively associated with AST (OR=0.91, 95% CI: 0.86–0.96, p<0.001). Age and sex were not significant. This model explained 4% of the variance (Nagelkerke’s R²=0.04).

Model 2 added family-level correlates to the individual-level variables. The inclusion of these variables improved model fit (Nagelkerke’s R²=0.11). Significant positive associations with AST were found for parental MVPA on ≥ 5 days per week (OR=1.71, 95% CI: 1.00–2.92, p=0.049) and for parental walking on ≥ 5 days per week for at least 10 minutes (OR=1.63, 95% CI: 1.04–2.56, p=0.032). Other family-level variables, including maternal and paternal education level, household size, and FHC-PA, were not significantly associated with AST in this model.

Model 3 incorporated physical-environment-level correlates in addition to the individual- and family-level variables. This final model was statistically significant (χ²(13)=192.71, p<0.001) and demonstrated the highest explanatory power (Nagelkerke’s R²=0.46, classification accuracy=81.7%). Walking time to school (OR=0.91, 95% CI: 0.88–0.93, p<0.001) and walking time to playgrounds (OR=0.96, 95% CI: 0.92–0.99, p=0.013) were negatively associated with AST. Neighborhood density and neighborhood quality for PA were not significant. Compared to Model 2, additional family-level correlates became significant in the final model: maternal education, household size, and FHC-PA, alongside the already significant regular parental MVPA and parental walking. Paternal education remained non-significant.

Model diagnostics showed no indication of multicollinearity (VIF range = 1.03–1.36) and good model fit for all three models, as indicated by non-significant Hosmer-Lemeshow test results (p>0.05). 

This study examined socio-ecological correlates of AST in first-grade primary school children. In the present sample, 64.8% of children engaged in AST, aligning with previous German studies reporting AST rates between 50% and 70% (12, 32, 39). Similar rates were observed in southwestern Germany (12), while slightly lower rates were found in studies from Leipzig, with 59.3% (39) and 44.2% exclusively active travelers (33). The findings further revealed that body fat percentage, as an individual health characteristic, family characteristics, and parental perceptions of environmental conditions were significantly associated with AST. The final model demonstrated substantial explanatory power, highlighting the complex and multi-layered nature of children’s school travel behavior.

At the individual level, a higher body fat percentage was linked to lower odds of AST. This finding aligns with previous research from Germany reporting a negative association between overweight and AST in primary school children (12). It suggests that higher adiposity may reflect lower habitual PA levels and highlights the possibility of bidirectional relationships, as greater engagement in active travel could potentially help to reduce body fat over time (15). In contrast to previous studies reporting higher rates of AST among boys (3, 34), no significant sex differences were observed in this sample. This may be explained by the young age of the participants, as sex differences in active travel behavior tend to become more pronounced at older ages (6, 17). Consistent with the present findings, other studies conducted in Leipzig also reported no significant sex differences in children’s AST (33, 39). Similarly, no significant association was found for age, which is likely due to the homogeneous age structure of the sample, comprising only first-grade children.

Several family-level correlates were significantly associated with children’s AST. Regular parental MVPA and parental walking behavior were positively linked to children’s AST, highlighting the importance of parental role modeling. These findings are in line with previous research emphasizing that children are more likely to adopt active behaviors when they observe similar patterns in their parents (37). Moreover, a favorable FHC-PA was associated with higher odds of AST. Shared family values and routines likely support active travel choices. Maternal education also emerged as a significant correlate, while paternal education was not. This may reflect the particular importance of maternal decision-making in young children’s daily routines, including school travel. This may also indicate broader gendered patterns in caregiving, where mothers more often oversee daily logistics, including school travel decisions. In this context, both health-consciousness (12) and trust in the child’s skill level (27) may play a role. Furthermore, a larger household size was associated with higher odds of AST. While this may indicate that children have older siblings to accompany them, increasing parents’ willingness to allow AST (16, 27), logistical factors such as car-sharing needs or time constraints in larger families could also contribute to a greater reliance on active travel. Overall, these findings underscore the crucial role of the family environment, not only in modeling PA but also in creating supportive conditions for children’s independent mobility.

Among the physical-environment-level correlates, perceived walking time to school and to nearby playgrounds was significantly negatively associated with AST. In contrast, subjective assessments of neighborhood density and quality of the neighborhood for PA were not related to AST in this sample. This suggests that functional aspects of the environment, such as the perceived travel time required, may be more influential in parental decision-making than more abstract evaluations of neighborhood quality or infrastructure. However, the lack of a significant association with neighborhood quality may also be due to limitations in how neighborhood quality was measured. In the present study, this variable was based on a composite score combining four dimensions - attractiveness, safety, amenities, and location - which were not analyzed separately. It is possible that only certain aspects, such as safety or connectivity, are relevant for AST decisions, and that aggregating them may have diluted their individual effects. The strong predictive value of perceived walking distance confirms previous findings identifying distance as one of the most consistent correlates of AST in children (26, 28, 31, 33, 34). The current findings underline that parental perceptions of walking distance are crucial in shaping school travel decisions. Future interventions may benefit from addressing parental distance perceptions through education, supportive communication strategies that inform parents about walkable routes, and urban design that supports visual connectivity and walkability (2, 4). 

Implications for Practice and Policy
The findings highlight the need for multi-level strategies to promote AST in young primary school children. Interventions should target parents, aiming to raise awareness of the need to actively organize their children’s school journey to support both health and environmentally friendly travel. Promoting active routines in daily family life and improving parental understanding of feasible walking distances are central components. Local authorities could support this by providing families with personalized maps showing the safest routes to school. Urban planning initiatives should focus on improving the perceived and actual accessibility of schools through traffic-calmed zones, safe crossings, and connected pedestrian networks, as recommended in walkability frameworks (4). Schools can further support AST by offering extracurricular programs that teach children to use various active transport modes safely. Engaging urban planners and school communities in participatory design processes could further enhance the local relevance of interventions. Finally, as early PA patterns tend to track into later childhood and adolescence (11, 38), interventions should ideally begin before school enrollment, for example, via preschools or informational materials provided at school registration, to establish long-term active travel habits.

Strengths and Limitations
This study has several strengths, including the application of a socio-ecological framework, the combination of objective anthropometric measures with parent-reported perceptions and behaviors, and the focus on young primary school children, an underrepresented group in AST research. However, some limitations should be acknowledged. This cross-sectional analysis precludes causal inferences and does not allow conclusions regarding the directionality of associations. Although parental perceptions provide valuable insights, they may be subject to social desirability or recall bias. Moreover, the generalizability of the findings may be limited to urban contexts. Future research should incorporate longitudinal designs and include children’s own perceptions to deepen the understanding of AST behaviors. While the present study assessed overall parental PA, future studies should separately examine maternal and paternal activity patterns to explore their potentially distinct roles in shaping children’s school travel behavior.

This study highlights the complex interplay of individual, family, and environmental correlates of AST in young primary school children. Parental perceptions of walking distance and family-level influences, such as parental PA and shared family values and routines related to PA, were particularly important. These findings emphasize the need for strategies that actively involve parents and address environmental barriers to promote AST from an early age. Future research should focus on longitudinal analyses and differentiate maternal and paternal influences to deepen the understanding of school travel behavior during early primary school years.

Conflict of Interest
All authors have completed the ICMJE Uniform Disclosure Form at www.icmje.org/coi_disclosure.pdf and declare no conflict of interests.

Funding
The research was funded by the European Social Fund (ESF) and co-financed by the Free State of Saxony, Germany (Grant Number: 100670471).

Ethical Approval 
An ethics vote was granted from the ethics committee of the Medical Faculty of Leipzig University (File number: 253-14-14072014).

This study examines socio-ecological correlates of active school travel (AST) in first-grade primary school children, an age group that has been underrepresented in previous research. By combining anthropometric data with parent-reported behaviors and  environmental perceptions, the study highlights the complex interplay of correlates at multiple levels to shape children’s school travel behavior.

The findings show that parental perceptions of walking distance, family health climate for physical activity, parental physical activity, maternal education, household size, and children’s body fat percentage are significantly associated with AST. These findings underline the need for interventions that not only improve infrastructure but also engage parents by promoting active family routines and raising awareness of walkable school distances.