The Effects of Socioeconomic Position and Neighborhood Characteristics on Mother's Engagement in Physical Activity

Background & Purpose.  Overweight and obesity are of major concern in the United States. Racial/ethnic and economic disparities persist, with minority and lower-income populations suffering the burden of these conditions.  Obesity is preventable through healthy diet and engagement in physical activity (PA). Many interventions exist to promote engagement in PA at the individual level.  However, engagement in PA is also influenced by social and physical neighborhood characteristics, including neighborhood safety, social cohesion, access to quality parks, and individual- and neighborhood-level socioeconomic position.  The purpose of this study was to examine how neighborhood-and individual-level factors affect mothers’ engagement in PA, and whether those relationships are mediated by social cohesion.

Methods. The Geographic Research on Wellbeing (GROW) study includes survey data from a diverse sample of 2750 California mothers ages 2057 with 4-10 year old children.  The GROW survey was conducted via mail and telephone in 2013 and 2014.

Structural equation modeling was used to test relationships between household and neighborhood level socioeconomic position (SEP), perceived neighborhood safety, availability of neighborhood parks, and mothers’ PA, mediated by social cohesion.   Each of these latent factors was composed of multiple indicators from the GROW survey, Census data and geocoded measurements.  Mplus was used for the SEM analyses to test the factorial validity of the latent constructs and to test the validity of the causal models. SPSS was used for descriptive analyses.  In the structural analyses, we controlled for family structure, mothers’ depressive symptoms, children’s age, mothers’ age, children’s gender, mothers’ race/ethnicity, and mothers’ obesity status.

Results. The measurement model fit well with minor specifications, which included co-varying error terms within the social cohesion factor and deleting an indicator measuring distance to the nearest park, which did not load significantly onto the factor measuring park access  [c² (df) = 1521.25(174), p<0.01, c²/df = 8.74, RMSEA=0.053 (0.051,0.056), CFI=0.91 and TLI = 0.89].  

In the structural analyses, social cohesion was significantly positively related with mothers’ engagement in PA (p=0.02) but was not significantly related to children’s PA (p=0.088) once the control variables were added to the model.  In the mothers’ PA model, paths from perceived neighborhood safety and neighborhood socioeconomic position were fully mediated by social cohesion.  Park access and individual socioeconomic position contributed significantly to social cohesion but not to mother’s PA directly.

Conclusions & Implications. Social cohesion was a significant predictor of PA in mothers, and fully mediated the paths from perceived neighborhood safety and neighborhood SEP to mother’s PA.  These results indicate interventions to improve engagement in PA should address neighborhood social cohesion.  Strong social cohesion may be able to reduce the effects of fear of crime and low neighborhood socioeconomic position on engagement in PA.  Understanding of these relationships could be enhanced in future research by including objective assessments of the built environment and other recreational facilities, and other indicators of individual health such as body mass index and nutrition information. Social workers and public health professionals should consider neighborhood-level interventions to build social cohesion, especially in economically depressed and/or high-crime neighborhoods.